Experimental Study Regarding the Synthesis of Iron Oxide Nanoparticles by Laser Pyrolysis Using Ethanol as Sensitizer; Morpho-Structural Alterations Using Thermal Treatments on the Synthesized Nanoparticles

Abstract

The laser pyrolysis technique was used in the synthesis of magnetic iron oxide nanopowders in the presence of ethanol vapors as a sensitizer. This technique uses the energy from a continuous-wave CO₂ laser operating at a 9.25 μm wavelength, which is transferred to the reactive precursors via the excited ethanol molecules, inducing a rapid heating of the argon-entrained Fe(CO)₅ vapors in the presence of oxygen. For a parametric study, different samples were prepared by changing the percentages of sensitizer in the reactive mixture. Moreover, the raw samples were thermally treated at different temperatures and their morpho-structural and magnetic properties were investigated. The results indicated a high degree of crystallinity (mean ordered dimension) and enhanced magnetic properties when high percentages of ethanol vapors were employed. On the contrary, at low ethanol concentrations, due to a decrease in the reaction temperature, nanoparticles with a very low size were synthesized. The raw particles have a dimension in the range of 2.5 to 10 nm (XRD and TEM). Most of them exhibited superparamagnetic behavior at room temperature, with saturation magnetization values up to 60 emu/g. The crystalline phase detected in samples is mainly maghemite, with a decreased carbon presence (up to 8 at%). In addition to the expected Fe-OH on the particles surfaces, C (and O) bearing functional groups such as C-OH or C=O that act as a supplementary hydrophilic agent in water-based suspension were detected. Using the as-synthesized and thermally treated nanopowders, water suspensions without or with hydrophilic agents (CMCNa, L-Dopa, chitosan) were prepared by means of a horn ultrasonic homogenizer at 0.5 mg/mL concentrations. DLS analyzes revealed that some powder suspensions maintained stable agglomerates over time, with a mean size of 100 nm, pH values between 4.8 and 5.3, and zeta-potential values exceeding 40 mV. All tested agents greatly improved the stability of 250–450 °C thermally treated NPs, with L-Dopa and Chitosan inducing smaller hydrodynamic sizes.

1. Introduction

Magnetic nanoparticles (NPs) offer a wide range of biomedical applications [1,2]. In oncologic therapy, hyperthermia [3] relies on inducing a differentiated thermal transfer at the site where the magnetic material is inserted, as opposed to the surrounding tissues, which are generally non-magnetic. Therefore, in order for the magnetic nanomaterial to meet these criteria, it must exhibit high magnetization (both saturation magnetization and magnetic susceptibility at low magnetic fields), satisfactory dispersibility in biological fluids, well-determined superparamagnetic properties (which involves a narrow dimensional and crystalline distribution, 2–10 nm), low cytotoxicity, and cellular permeability [4,5]. Another application is the manipulation of the cells surrounding the magnetic material by applying a magnetic gradient which imprints a force on the desired direction only on these cells [6]. Iron oxide nanoparticles with spinel structure (maghemite and magnetite) are frequently used as active magnetic materials, due to their unique properties, such as an enhanced biocompatibility and cellular penetration [7,8,9], chemical stability in physiological fluids [10,11], and high magnetization [12]. Furthermore, these properties can be optimized using thermal treatments and adequate, biocompatible stabilizers. Also, a vast field of biomedical use of magnetic iron-based NPs is dedicated to drug delivery, opening the possibility of organ-targeted local guiding/administration of bioactive molecules via external magnetic fields [13,14]. The magnetic and superparamagnetic NPs were also intensely studied for their application in theragnostics, such as for enhancing the contrast in magnetic resonance imaging (MRI) [15,16,17]. Fe₃O₄, γ-Fe₂O₃, and other magnetic oxides, such as ferrites type MFe₂O₄ (where M is a divalent cation) under nanometric form also have applications in life sciences, such as for biological macromolecules (nucleic acids, proteins) or bio-entities (eukaryotic cells, bacteria, viruses) extraction/separation/sequestration or detection, and tracking [1,18,19], as well as for environmental remediation—extraction of toxic heavy metal ions or organic pollutants from wastewater [20,21].

A wide variety of synthesis methods can be employed to synthesize magnetic NPs, which can be classified in function of the synthesis medium: liquid, gas (vapor) phase, or solid phase. From the first category, most employed are: sol-gel, coprecipitation, thermal decomposition in high boiling point solvents, polyol, hydrothermal or solvothermal, microemulsion, sonochemical decomposition, microwave decomposition, and pulsed laser ablation in liquids. From the second category: flame combustion, wire evaporation, chemical vapor deposition, spray pyrolysis, and laser ablation under gas medium have been reported, whereas the ball-milling and solid state combustion methods belong to the last class [20,21,22].

Laser pyrolysis is a gas phase method for nanoparticle synthesis based on the decomposition of precursors from vapors state which offers several advantages, making it a valuable technique for material synthesis and processing: (a) high precision and control over reaction conditions, such as temperature and energy input, leading to consistent and reproducible results; (b) rapid heating and cooling—providing intense, localized heating, enabling rapid thermal decomposition of precursors and fast cooling, which is crucial for producing NPs with controlled sizes and properties; (c) versatility in material synthesis—it is used to produce a wide range of nanomaterials, including oxides, carbides, nitrides, and metals; (d) purity and high-quality products—the method avoids the contact of the reactive zone with the reactor walls; and (e) scalability and efficiency—the NP production is continuous and industrial scalable [23,24,25].

Laser pyrolysis facilitates the obtaining of iron oxide nanoparticles with extremely low dimensions, a reduced dimensional dispersion, optimum crystallinity (frequently, the crystallite dimension can be equal with those of an individual particle—crystalline monodomain on every particle), and in quite large quantities: a couple of grams per hour with an experimental set-up that can be easily scaled for industrial needs. Using optimal experimental parameters, NPs with high magnetizations (>80 emu/g) can be attained [25].

The following study will explore a possible alternative to the conventional laser pyrolysis technique, which usually employs common sensitizers, such as ethylene (C₂H₄), silane (SiH₄), or sulfur hexafluoride (SF₆), for producing the synthesis of iron oxide NPs with a low carbon content, which will further lead to an increased dispersibility of the NPs. Details regarding the experimental study conducted using ethanol (C₂H₆O) vapors as a sensitizer will be presented and discussed.

2. Materials and Methods

Iron oxide nanoparticles have been synthesized using the laser pyrolysis technique with the experimental set-up presented in Figure 1. Briefly, the method is based on the cross-flow configuration of the reaction chamber that allows the interaction between the emission line of the IR laser and the absorption line of the sensitizer.

An IR CO₂ laser (Puri Laser Technology CO., Nantong, China) having λ = 9.245 µm and 80 W nominal power was used, with iron pentacarbonyl Fe(CO)₅, Merk KGaA, Darmstadt, Germany) as the iron precursor and absolute ethanol (C₂H₆O, Chimreactiv S.R.L., Bucharest, Romania) as the sensitizer. A coaxial 4.0 argon (Ar, Messer SE & Co. KGaA, Bad Soden, Germany) flow was used with a dual purpose: firstly, to ensure the flow of the reactive gas mix towards the reaction chamber, and secondly, to transfer the freshly nucleated particles towards the collection chamber. The process has been described in detail elsewhere [12].

The samples NP1 through NP7 were synthesized using a distance between the focal lens (F = 33 cm) and a reaction zone of 28 cm, which corresponds to a focal spot of φ = 2 mm. A gas/vapor injection system based on two concentric tubes was used. The pressure and laser power were kept constant throughout the experiments, at 300 mbar and 66 W, respectively. A series of gas flows were also kept constant in order to highlight the dependence of the sensitizer flow: confinement Ar flow, introduced through the outer admission nozzle—2000 sccm; Ar flow keeping the windows clean—300 sccm; and those from the reactive mix (via the inner nozzle): O₂—50 sccm; Ar for Fe(CO)₅ entraining from a bubbler—66 sccm; a second Ar flow of 50 sccm, which is either fully (NP1) or partially (NP2-7) bubbled through liquid ethanol (see

Table 1. Experimental parameters.

Sample	D-Ar Dilution	D-Ar Through Ethanol	D-Ethanol	Laser Power Ar/Gas Mixture	Flame Temperature
	[sccm]	[sccm]	[sccm]	[W]/[W]	[°C]
NP1	0	50	17.2	66/64	607
NP2	25	25	8.6	66/65	588
NP3	37.5	12.5	4.3	66/65	510
NP4	40	10	3.4	66/65	500
NP6	42	8	2.7	67/65	500
NP7	30	20	6.9	66/64	550

, row 3).

The obtained NPs (nanoparticles) were subjected to thermal treatment (150°, 250°, 350°, 450°, 500°, and 600°) in order to enhance their magnetic properties and nano-crystallinity. The treatments were conducted in a normal-pressure atmosphere for 3 h, with a heating rate of 3 degrees/min. Since exposure to air was unavoidable, the powders were subjected, during the treatment, to an oxidation process which could supplement the synthesis process, resulting in nanopowders exclusively containing only Fe species having the maximum oxidation state, Fe³⁺ (either as γ-Fe₂O₃ or as α-Fe₂O₃).

Water-based suspensions with the raw or annealed NPs were used in order to determine their hydrodynamic diameter and zeta-potential via dynamic light scattering (DLS) measurements employing a Nanoparticle Analyzer SZ-100V2 (Horiba, Kyoto, Japan), which operated a diode-pumped solid-state (DPSS) laser emitting at 532 nm, with a power of 100–240 V AC ± 10% at 50/60 Hz. These suspensions were attained using the synthesized NPs and distilled water and/or a stabilizing agent (chitosan, with low molecular weight (75–85% deacetylated) L-3,4-dihidroxyphenylalanine (L-Dopa) and low viscosity sodium carboxymethyl cellulose (CMCNa), all from Merk, Darmstadt, Germany) at a concentration of 0.5 g/L and employing an ultrasound probe (500 W Ultrasonic Processor, with a 3 mm diameter probe) for 20 min in a cooling bath, in order to maintain the suspensions at a room temperature. The measurements were collected at 5 min and 2 h after the suspension preparation. The stabilized suspension followed the following protocols: (i) for chitosan: a stock solution of 0.5 g/L chitosan prepared in an ultrasound bath for 2 h, 59 kHz, 20 °C with a cooling system, to which the appropriate amount of NPs was added at the same concentration and the suspension homogenized with the use of a ultrasound probe, 20 min, with a cooling system; (ii) for L-Dopa: a stock solution of 0.5 g/L L-Dopa prepared in an ultrasound bath for 12 h, 59 kHz, 70 °C, protected from light; NPs were added over the solution and the resulted suspension (0.5 g/L NPs) was homogenized with the ultrasound probe, 20 min, with a cooling system; (iii) for CMC-Na: CMCNa and NP powders were added in small portions in distilled water under the action of an ultrasonic horn and cooling bath, continuing for 20 min in order to achieve homogeneous suspensions containing 0.5 g/L CMCNa and 0.5 g/L NPs.

The infrared (IR) absorption study on the composition of the reactive gases was performed using a Fourier transform infrared (FTIR) spectrometer (Thermo Fisher Scientific, Waltham, MA, USA), model Nicolet iS50, resolution 4 cm⁻¹. The required gasses were collected by attaching a gas cell at the end of the experimental set-up.

The resulting nanoparticles were analyzed for their phase composition and crystallinity by X-ray diffraction (XRD) with a PANalytical X`Pert MPD theta-theta X-ray diffraction apparatus (Malvern Panalytical Ltd., Malvern, UK) with a Cu K α source (1.5418 Å). The Sherrer equation was used to perform evaluation of crystallite sizes of the NPs; in this case, a value of k = 0.9 for the shape factor was employed [26].

Elemental evaluation was performed using X-ray photoelectron spectroscopy (XPS) and energy-dispersive X-ray spectroscopy (EDS) analysis. X-ray photoelectron spectroscopy (XPS) employed an ESCALAB Xi+ X-ray Photoelectron Spectrometer Microprobe from Thermo Fisher Scientific Inc (Waltham, MA, USA) apparatus with an Al Kα radiation source (hν = 1486.2 eV). The energy reference used was C 1s level (284.4 eV). ‘Average’ software, version 5.978, was employed for the identification of superficial chemical compositions and oxidation states. For the EDS investigations, the powders were hand-pressed and fixed onto copper foils, and the analysis was performed using the EDAX detector attached on a FEI Quanta Inspect S Scanning Electron Microscope (SEM) at 10 kV accelerating voltage (Thermo Fisher Scientific Inc., Waltham, MA, USA). Bright-field TEM images were obtained using a high-resolution Titan Themis transmission electron microscope from Thermo Fisher Scientific Inc (Waltham, MA, USA). The microscope operated in transmission mode at a voltage of 200 kV.

Specific surface area was measured by a Brunauer-Emmett-Teller (BET) using a SA-9600 flowing gas surface area analyzer from Horiba Kyoto, Japan, with a 30% N₂/70% He gas mixture.

The magnetization of the samples was measured with Vibrating Sample Magnetometry (VSM) at room temperature using an ADE Technologies VSM880 magnetometer (Lowell, MA, USA).

Thermogravimetry analysis was performed employing a Thermogravimetric Analysis/Single Differential Thermal Analysis (TGA/SDTA) Mettler Toledo apparatus, Greifensee, Switzerland. The measurements were attained in air, with a heating rate of 5 °C/min, from 25 °C up to 1000 °C. Differential scanning calorimetry (DSC) analysis was obtained using a DSC Mettler Toledo instrument, Star1 model, Switzerland and were performed in air with a 5 °C/min step within a 25–450 °C temperature range, using indium as a calibration reference.

3. Results and Discussion

The synthesis process was realized, in all cases, in a stable regime, and the experiments with lower flow rates required a 5 to 10 min waiting time in order to allow the proper, uniform mix of the precursors. The 4th and 6th (NP) experiments required a higher starting power, 68–72 W; once the pyrolytic flame appeared, the absorbed laser power could be lowered to the nominal value found in the table. The flame has a smoking appearance with a visible area that exceeds the irradiated volume, which decreases with a decrease of the sensitizer flow rate, and the luminous course (flame height) is of a maximum 7–8 mm in the case of NP1. The main reason for this phenomenon is the secondary burning process of ethanol [27] after the initiation of the pyrolytic flame, an exothermal reaction which contributes to the increase in temperature in the area of nucleation and growth of iron oxide NPs. This process has been optimized due to the beneficial contribution related to the improvement of NPs crystallinity, but also an enhancement of the inter-particle welding process or C doping [12]. The optimization process involved size and zeta-potential analysis on the collected NPs in water-based suspensions, the acceptable limit being an increase in the aggregate size of maximum 20% over the course of 2 h and zeta-potentials over 35 mV.

A comparative analysis of the reactive gases before and after the NP synthesis (NP2) was attained and the resulting spectra can be observed in Figure 2.

In Figure 2, the absorption maxima for Fe(CO)₅ and ethanol in the precursor/reactive mix are identified, and in the gas spectrum resulting from the pyrolytic reaction, an additional CO peak can be detected, a result of the Fe(CO)₅ decomposition. For the absorption attenuation analysis in the 645 and 1066 cm⁻¹ maxima area, it can be calculated that 3.7% Fe(CO)₅ and 31.5% ethanol proceed unreacted.

The decomposition of ethanol with the formation of C condensable species is a phenomenon that should preferably be avoided, which is why the experimental parameters that do not highlight the reaction production along the ethanol decomposition path or its partial burning must be identified. In the abundant presence of O₂ and at high temperatures, ethanol can completely burn, releasing C as CO₂, which is why it is used as an alternative fuel. The process of ethanol combustion, at least the complete one, must be avoided because it induces a parasitic reaction alongside that of laser pyrolysis with iron oxide production, which is why gas mixtures and synthesis conditions for insufficient ethanol combustion are sought. In the case of incomplete burning, the chemical reaction would probably be [28]:

C₂H₅OH + 3⁄2 O₂ → 3 CO₂ + 3 H₂

(1)

The main product of ethanol decomposition in the absence of oxygen would be in the line of C formation, leading, initially, to the formation of ethene, in accordance with Equation (2) [29]:

C₂H₅OH → C₂H₄ + H₂O

(2)

The presence of ethene in the reactive product mix should be easily revealed by its specific absorption peak at 948 cm⁻¹. By focusing on the wavelength area where ethene is active, a small absorption peak can be observed, suggesting that a reduced amount of ethanol leads to ethene production. The next step in the generation of C condensable species would be the formation of acetylene from ethene, a common aspect encountered when ethene is used as a sensitizer. The reaction has a significant share at high temperatures (over 1200 °C) [30]; however, it is possible that, in the pyrolytic flame, the freshly nucleated and hot iron NPs have a catalytic role in the decomposition of hydrocarbons.

C₂H₄ → C₂H₂ + H₂

(3)

As can be observed in the right-hand insert of Figure 2, there are no traces of C₂H₂ in the reaction gaseous mix; therefore, the thermodynamic conditions are favorable for the formation of iron oxide nanoparticles at a high yield of conversion: over 90 at.% Fe from Fe(CO)₅ can be found in the iron oxide NPs. It is worth mentioning that the production of NPs in the majority of the experiments in this study was between 1.5 and 2 g/h.

The obtained nanopowders have been analyzed regarding their crystalline structure and their diffractograms, as illustrated in Figure 3.

In order to facilitate the comparative study of the nanometric samples obtained, the intensity signal has been normalized for all the diffractograms, this being achieved by the absolute maximum peak, identified as (311) at 2θ~35.5°. The analyzed material is a nanometric one, the peaks having a broad appearance and being regularly fitted with the pseudo-Voigt function, which could be due to the narrow crystallite size distribution. In the optimized conditions under which this study has been conducted, the nanoparticles contain only the spinel phase, and the presence of metallic iron is not confirmed with certainty, even in the experiments with a low flow rate of the sensitizer (NP3, 4, and 6). A degree of damping can be observed for the analyzed material (the base curve presents significant noise, especially at lower angles), a trait that can be correlated with the decrease of sensitizer flow and, implicitly, of the synthesis temperature.

Figure 4 highlights the dependence of the sensitizer flow on the crystallite dimensions, evaluated under the accepted theory that, after bubbling, the carrier gas flow is mixed with the ethanol vapors at the saturated vapor pressure (76.65 mbar/25 °C).

Evidently, the size of the particle depends greatly on other important parameters, such as the temperature in the pyrolytic flame, pressure, laser power, iron pentacarbonyl and oxygen vapor flow rates, and the nozzle diameter. However, in this case, it is assumed that all these parameters are constant, with the exception of the temperature in the pyrolytic flame, which is directly dependent on the percentage of the sensitizer in the reactive mix. It can be observed that by controlling the carrier gas flow, Ar through the bubbler with ethanol, one can select a wide range of NPs sized exactly within the area where the magnetic properties are in interdependence with the crystallite dimension (2–10 nm).

The thermally treated nanopowders were also analyzed using XRD, and the overlapped diffractograms for two reference samples (NP1 and NP2) at different temperatures are presented in Figure 5.

Moreover, the mean crystallite dimensions of the freshly synthesized NPs and those thermally treated in experimental sets NP1, 2, and 6 are presented in

Table 2. Mean crystallite dimensions of NP1, 2, and 6, freshly nucleated and thermally treated at different temperatures.

Sample	γ-Fe2O3-Fe3O4			α-Fe2O3
	2θ-440 [o]	a [Å]	Dmean [nm]	2θ-104 [o]	Dmean [nm]
NP1 as syn	62.74	8.377	9.9	-	-
NP1-150 °C	62.90	8.358	10.2	-	-
NP1-250 °C	62.92	8.356	10.4	-	-
NP1-350 °C	62.95	8.352	10.3	-	-
NP1-450 °C	63.02	8.344	11.0
NP1-500 °C	62.81	8.369	7.8	33.06	22.5
NP1-600 °C	-	-	-	33.08	33.3
NP2 as syn	62.84	8.365	5.6	-	-
NP2-150 °C	62.88	8.360	6.2	-	-
NP2-350 °C	62.97	8.350	6.3	-	-
NP2-600 °C	-	-	-	33.19	33.2
NP6 as syn	62.82	8.368	3.0
NP6-150 °C	63.07	8.337	3.0
NP6-350 °C	62.93	8.354	3.2	-	-
NP6-450 °C	43.51	8.319	3.5	33.09	20.8

.

Evaluation of the lattice constant ‘a’ was carried out by following the hypothesis that the searched lattice is the cubic one, similar to γ-Fe₂O₃ and Fe₃O₄, small differences being exactly the value of this constant. In the case of γ-Fe₂O₃—the theoretic value for γ-Fe₂O₃ is 8.352Å (PDF file: 00-039-1346), and for Fe₃O₄ is 8.396 Å (PDF file: 00-019-0629). It can be observed that for the freshly synthesized samples, the lattice constant is found between the two theoretical values, with thermal treatments correcting this constant and placing it closer every time to the theoretical value of maghemite. X-ray diffraction confirms that by subjecting the sample to thermal treatments, the nanostructure orders itself towards maghemite, even if it has very low crystallite dimensions. This aspect is further highlighted due to the appearance of the specific peaks for ordered maghemite: (210) and (211). The crystalline structure of maghemite borrows the complex architecture of magnetite, replacing the orthogonal positions of the iron divalent ions with a combination of trivalent ions and vacancies on a well-defined succession alongside the cubic elemental cells [31]. Up until the transition temperature, the crystallite dimension of the maghemite slowly increases with the rise in treatment temperature (see column 4, Table 2). Moreover, it can be observed that at high temperatures, at least 450 °C for NP1 and 350 °C for NP6, the maghemite is stable, with the transition towards hematite beginning at 450 °C and becoming fully complete at temperatures above 600 °C for all analyzed powders. The transition temperatures tend to slowly decrease with the decrease in crystallite size. The phase transition is attained with a significant alteration of the crystallite dimension, similar to the process of synthesizing ceramics. Another notable aspect is that the fresh powder presents a quite low background noise, which is further reduced by increasing the temperature. This phenomenon is substantial between 150 °C and 350 °C; as it reaches 600 °C, the noise is barely noticeable—basically, the baseline closely follows the zero-intensity line, meaning that the nanomaterial is almost entirely crystalline.

Both the as-synthesized samples and those thermally treated have been investigated for their elemental composition and the most relevant inter-atomic bonds in these nanoparticles (Figure 6).

Table 3. Elemental composition evaluation by XPS for synthesized NPs with ethanol as sensitizer.

Sample	Fe (at.%)	C (at.%)	O (at.%)	Fe-O (at.O%) /Peak Energy (eV)	Fe-OH + C-O (at.O%)/Peack Energy (eV)
NP1	29.15	22.58	48.27	73.1/530.0	26.9/531.6
NP1 T-350 °C	21.35	25.00	53.65	76.0/529.9	24.0/531.4
NP2 as synth.	34.14	14.29	51.56	73.8/529.9	26.2/531.6
NP2 T-150 °C	24.61	18.68	56.71	75.0/529.8	25.0/531.4
NP2 T-350 °C	33.75	16.53	49.72	81.2/529.8	18.8/531.3
NP2 T-600 °C	33.73	14.52	51.75	77.6/529.7	22.4/531.4
NP4	20.37	26.39	53.24	74.3/529.9	25.7/531.6
NP6	28.73	23.67	47.6	68.0/529.9	32.0/531.5
NP6 T-350 °C	19.96	28.8	51.24	70.9/529.9	29.1/531.5

presents the ratios of the elements as well as the two O formed bonds: Fe-O and Fe-OH. Although the C content is higher compared with the EDS evaluation, it is well below the most common iron oxide NPs synthesized with ethylene and comparable with the ones using isopropanol (~20 at.% C). The lowest C concentration can be observed in the NP2 sample, and this is likely one of the reasons for its facile dispersibility in water. C content slowly decreases with thermal treatment.

Figure 6 presents the stacked XPS spectra for O 1s active zone (534–527 eV) from the NP1 sample. Three bonds are identified: the main bond, ionic, metal-oxide, centered at a low bonding energy of 529.9 eV; the hydroxide Fe-OH-specific bond, located in the 531.5 eV region; and, where appropriate, C-O bonds such as O=C-O (carboxylic), C-O-C (aromatic), and C-OH (aliphatic) at bonding energies above 532 eV. By closely analyzing the deconvolution of the O 1s peak, it can be observed that fitting with two peaks centered on the first two types of bonds is appropriate for the experimental spectra. Table 3 presents the weighting of the two types of oxygen bonds with iron, and a reduction of the Fe-OH bond with the increase of treatment temperature distinguished. For the sample treated at 600 °C, a slight diminishment of the ionic bonding energy Fe-O (529.7 eV) can be observed, which could indicate the transition from maghemite, where the Fe-O energy was 529.9 eV, to hematite. For the untreated sample, an extremely low contribution from the C-O bond in the area 533.3 eV can be observed in the experimental curve. The main peak, which characterizes the intensity of the Fe-O ionic bond, exhibits slight variations in value with crystallite size or synthesis temperature, specifically around 529.9 eV and values commonly encountered in the literature for both nanometric materials and thin films [32,33,34]. The substantial presence of the functional bond Fe-OH, as well as the fact that it maintains its significance regardless of treatment temperature, leads to the conclusion that the treated NPs can be attached to different stabilizers, or easily functionalized, which is beneficial for biomedical applications.

Comparative analysis of the HR-XPS spectra of Fe 2p (Figure 7a) indicates the maxima of Fe 2p_3/2, Fe 2p_1/2, and their satellite zones. The position of the main maximum (~710.6 eV), as well as the general trend of Fe 2p_3/2 peaks, with the lack of clear shoulders and the absence of the Fe⁰ signal (black vertical line), leads to the conclusion that all the samples contain iron oxide with Fe³⁺ as a majority ionization state. Moreover, the satellite peak centered at 719.3 eV was attributed to Fe³⁺, serving as a specific signature of maghemite or hematite, while the magnetite is characterized by the absence of a such peak due to satellite signal compensation between Fe²⁺ (~715 eV) and Fe³⁺ [35]. Figure 7b shows the HR-XPS spectrum in the Fe 2p zone for sample NP1, which was fitted according to reference [33] for γ-Fe₂O₃. The fitted envelope curve nearly (but not exactly) matches the experimental curve, confirming the presence of maghemite as the majority phase. However, the presence of magnetite as a minor phase, or more specifically, of a majority phase of maghemite with defects in which a low concentration of Fe²⁺ is present, cannot be excluded. Corroborating these results with those from X-ray diffraction, it can be concluded that, for the analyzed samples, maghemite is the dominant crystalline phase.

Table 4. Elemental composition obtained from EDS analysis and BET-specific surface area measurements.

Sample	C (at.%)	O (at.%)	Fe (at.%)	BET (m2/g)
NP1	1.86	60.64	37.71	91.47
NP1 T-150 °C	1.17	59.86	38.97	90.73
NP1 T-350 °C	1.3	58.34	40.36	84.04
NP1 T-450 °C	1.38	58.56	40.06	77.20
NP1 T-600 °C	0.77	57.8	41.42	15.46
NP2	1.26	64.33	34.41	135.11
NP2 T-150 °C	1.11	63.69	35.20	-
NP2 T-350 °C	1.10	64.42	34.48	-
NP2 T-600 °C	0.64	61.56	37.80	-
NP3	2.35	64.6	33.04	193.52
NP4	2.61	59.42	37.97	205.13
NP4 T-350 °C	1.52	58.96	39.51	-
NP6	2.36	59.8	37.84	210.1
NP6 T-350 °C	4.13	56.46	39.40	146.50
NP6 T-450 °C	1.49	53.43	34.42	139.49
NP7	1.71	59.85	38.43	148.12

presents the elemental composition evaluated through EDS, a technique that allows averaging the results.

It can be observed that the laser pyrolysis synthesis using ethanol as a sensitizer leads to low C doping. Iron oxide NPs previously synthesized [36] using SF₆, air, and the same Fe(CO)₅ precursor have led to similar or even higher C doping (1–3 at.%). In that case, the only explanation is that the iron precursor releases CO, and this can generate C as a consequence of a disproportionation reaction catalyzed by the freshly formed Fe clusters (before they start to oxidize):

2CO → CO₂ + C

(4)

In our case, this chemical mechanism could be the main source of C presence, whereas ethanol as a C source is much less probable because all ethanol decomposition products are totally volatile.

By analyzing the data in Table 4 and the Figure 8, it can be concluded that, for diminishing C concentration, there is an optimum relative to the ethanol flow introduced in the synthesis: experiment NP2, characterized by a 25/25 dispersion of Ar for ethanol dilution and entrainment, and a pyrolytic flame temperature of 585 °C. Thermal treatments usually lead to a diminishing of C presence, so the samples with NPs restructured to hematite end up with C content under 1 at.%, likely due to the 600 °C temperature, which contributes to the partial elimination of C through oxidation to CO₂. However, the XPS evaluation indicated a value of 14.5 at.% C in sample NP2-treated at 600 °C. The C1s peak for this sample was analyzed revealing the following characteristic bonds and proportions: C-C (284.8 eV)—68 at. C%, C-O-C (285.85 eV)—19.9 at.% C and O-C=O (288.53 eV)—11.9 at.% C.

It can therefore be concluded that ethanol as a sensitizer is adequate for the synthesis of iron oxide NPs with an extremely low C content. Moreover, there is a diminishing tendency of C with thermal treatments, as the superficial functional groups, or those graphene C layers covering the nanoparticles’ surface, are likely eliminated through combustion at sufficiently high temperatures (~350 °C). The Fe/O ratio shows an excess of oxygen relative to the most oxidative iron phase (Fe₂O₃: Fe/O = 0.667); an explanation might be that on the surface, Fe-O-OH or COOH functional groups are formed, contributing additional oxygen, while chemisorbed water vapor-derived species may also contribute to the evaluation over the γ-Fe₂O₃ stoichiometry of O₂.

The as-synthesized and treated samples have also been analyzed regarding their specific surface areas, as shown in the last column of Table 4. The specific surface area is inversely proportional to the particle size, approximately equal to the one attained from XPS as particle’s order dimension, and increases with the decrease in synthesis temperature (which determines the growth speed of the crystal). The values range from 91.5 m²/g for sample NP1 with a 9.9 nm d_mean from XRD, to 210 m²/g for sample NP6 with a 2.4 nm d_mean. The values for the treated samples slowly decrease with the increase in the treatment temperature. For example, NP1, which has 77.2 m²/g at 450 °C, drops significantly after maghemite-hematite phase transition and reaches 15.5 m²/g at 600 °C.

TEM images of NPs from NP1-350 °C and raw NP4 samples are presented in Figure 9, TEM (a) and (b) as examples for the as-synthesized and the annealed powders, respectively. In both samples, the individual NPs are easy to discern and are joined together, forming interconnected, chain-like, non-compacted aggregates, a characteristic feature exhibited by laser pyrolysis-synthesized NPs [25]. NPs structured in loose agglomerates, as in this case, are very useful due to their ability to be easily divided under the action of ultrasounds in desired fluids. The NP4 individual particles have a spheroidal morphology and narrow size distribution, with a mean dimension of ~4.2 nm, whereas those from NP1-350 °C present a larger size distribution with a median value around 9.6 nm, having similar rounded shapes with few exceptions for polyhedral or elongated forms. For the annealed sample, different welding stages between NPs were revealed, which is detrimental to their disaggregation in fluids. Thus, the time and temperature treatment of the NPs should be correlated with the Z-average value from DLS analysis of their suspensions.

The samples have been analyzed regarding their magnetic behavior with a VSM magnetometer at room temperature.

Figure 10a shows the hysteresis loops for NP1 (9.9 nm average dim.) treated at different temperatures. At low fields, a slow increase in magnetization and susceptibility can be observed up to the 450 °C treatment (with saturation magnetization rising from 51 to 58 emu/g), before dramatically dropping in the sample transformed into hematite (NP1—600 °C). The samples present a very low hysteresis; one possible explanation is that the larger nanoparticles have breached the room temperature blocking barrier and act as particles with magnetically stable, unique domains [37]. The rest of the material remains in the superparamagnetic zone (H_c =0 Oe). Figure 10b shows the same hysteresis dependency for the treated sample NP6—the sample with the smallest crystal dimension, ~2.4 nm. The freshly synthesized sample presents the lowest value of the coercive field: 0.12 Oe, which indicates a totally superparamagnetic material. However, the saturation magnetization is achieved with difficulty at large fields, with a value of 7 emu/g. With thermal treatments, the saturation magnetization of these powders is significantly enhanced and can reach a maximum of M_s = 15.7 emu/g and a H_c = 4.8 Oe.

A reference sample, NP2, was used for thermogravimetric analysis. The TGA analysis indicates a continual weight loss process along the entire analyzed area from 30–1000 °C, highlighting specific areas/sections with different drop speeds. TGA, DTG, and DSC thermograms are presented in Figure 11.

The first area associated with the loss of the volatile component absorbed on the nanoparticles surface (especially water) can be found between 30 °C and 110 °C, which generates a final weight loss of 0.7 wt %. A second representative drop area is between 175 °C and 280 °C, with a weight loss of 2.4 wt %, and this can be related to the restructuring processes in the nanoparticles, which lead to the significant decrease of the 531.5 eV peak of O 1s associated with the Fe-OH functional groups, and, to a lesser extent, the aromatic O-(C=O)-C.

Referring to Table 3 and the analysis of the HR XPS O 1a spectra (Figure 7) for the untreated NP2 powder, a 26.2 at. O% is determined for the peak centered at 531.6 eV, which then drops to 25.0 at. O% after treatment at 150 °C, and continues until 18.8 at. O% when treated at 350 °C. Simultaneously, by looking at the diffractograms for the same sample (Figure 5b), there is a clear increase in crystallinity with the increase in temperature, a process explained in detail in the XRD section. The DSC analysis confirms the restructuring process similar to the decomposition processes, with an exothermal maximum at 282 °C. Moreover, on the thermal range presented, there is no weight gain (which could have been explained by the oxidation of the possible magnetite crystalline domains). This further strengthens the hypothesis that the dominant phase, after synthesis, is maghemite. Over 350 °C, the weight drop continues slowly and can be associated with the loss of C and functional groups from the NP surface. There are two more areas that require highlighting: 466 °C and 647 °C. The first can be associated with the γ-Fe₂O₃—α-Fe₂O₃ phase transition, while the second is likely due to the restructuring of the superficial C layers.

Both the as-synthesized and thermally treated NPs were used in nanofluid preparation, with or without different stabilizers, and the resulting samples were analyzed by DLS and zeta-potential evaluation. In dynamic light scattering (DLS), the Z-average is a key parameter used to describe the size distribution of particles in a suspension or colloid. The Z-average represents the average size of NP agglomerations based on their scattering intensity, which accounts for both their size and concentration in a suspension [38].

Table 5. DLS analysis for NP-based suspensions in distilled water.

Sample	Z-Average (5 min)	PDI	Zeta	Z-Average (2 h)	PDI
	nm		mV	nm
NP1	111.5	0.377	44.8	125.1	0.408
NP1 T-150 °C	87.4	0.23	48.6	101.2	0.292
NP1 T-250 °C	4048.8	0.824	−4.5	decanted	-
NP1 T-350 °C	8488.9	4.208	−4.1	decanted	-
NP1 T-450 °C	4482.1	1.007	−1.0	decanted	-
NP1 T-600 °C	2571.5	0.744	−39.6	decanted	-
NP2	81.5	0.331	60.1	88.6	0.376
NP2 T-150 °C	78.7	0.335	35.5	81.4	0.291
NP2 T-350 °C	2941.2	0.638	−1.5	decanted	-
NP2 T-600 °C	290.1	0.425	34.9	325.6	0.437
NP3	81.6	0.381	0.3	100.8	0.411
NP4	98.5	0.389	22.0	106.1	0.416
NP4 T-150 °C	95.6	0.353	48.6	105.3	0.438
NP4 T-250 °C	2673.0	0.675	12.7	decanted	-
NP4 T-350 °C	5871.6	0.749	6.0	decanted	-
NP4 T-450 °C	1347.8	0.560	17.2	decanted	-
NP5	80.7	0.360	39.1	116.7	0.436
NP6	85.5	0.372	50.1	99.3	0.430
NP6 T-150 °C	95.2	0.439	54.5	87.5	0.357
NP6 T-250 °C	4987.7	0.823	−29.6	decanted	-
NP6 T-350 °C	2984.6	0.641	8.3	decanted	-
NP6 T-450 °C	2217.5	0.482	8.9	decanted	-
NP7	81.3	0.323	50.1	92.1	0.401

presents the results of the DLS measurement for the samples in water-based suspensions. The PDI (polydispersity index) can provide a quantitative measure relative to the dimensional range of aggregates in a suspension. Ideally, the values should be as close to 0 as possible: values below 0.1 indicate the suspensions are monodisperse, values between 0.1 and 0.4 indicate the polydispersity is moderate, and values above 0.4 indicate the suspensions are broadly polydisperse [38,39].

Table 6. DLS analysis for stabilized NP-based suspensions in distilled water.

Sample	Z-Average (5min)	PDI	Z-Average (2 h)	PDI
	nm		nm
CHITOSAN
NP1	169.8	0.379	183.9	0.396
NP1 T-150 °C	143.7	0.427	150.8	0.382
NP1 T-250 °C	145.0	0.396	192.0	0.386
NP1 T-350 °C	142.4	0.414	177.7	0.397
NP1 T-450 °C	134.2	0.402	153.1	0.441
NP1 T-600 °C	277.1	0.357	301.2	0.328
NP2	145.5	0.484	182.3	0.390
NP3	156.6	0.400	171.3	0.491
NP4	138.3	0.399	136.7	0.504
NP4 T-350 °C	154.6	0.430	164.0	0.509
NP5	123.5	0.411	153.2	0.486
NP6	153.7	0.438	158.9	0.410
NP6 T-350 °C	131.7	0.450	160.9	0.380
NP6 T-450 °C	140.1	0.388	197.4	0.492
NP7	158.4	0.456	145.5	0.424
L-DOPA
NP1	170.2	0.334	312.1	0.484
NP1 T-150 °C	147.9	0.359	483.6	0.393
NP1 T-250 °C	152.8	0.312	149.0	0.353
NP1 T-350 °C	99.1	0.379	98.7	0.317
NP1 T-450 °C	87.0	0.330	111.3	0.378
NP1 T-600 °C	178.0	0.290	193.8	0.268
NP2	84.7	0.302	129.3	0.386
NP3	103.7	0.332	112.9	0.409
NP4	98.9	0.379	122.1	0.336
NP4 T-350 °C	101.2	0.378	117.8	0.414
NP5	98.4	0.369	117.4	0.405
NP6	106.0	0.314	115.8	0.375
NP6 T-350 °C	93.4	0.429	124.9	0.430
NP6 T-450 °C	99	0.304	124.0	0.422
NP7	104.1	0.372	95.3	0.342
CMCNa
NP1	332.5	0.460	395.1	0.424
NP1 T-150 °C	240.8	0.383	321.8	0.427
NP1 T-250 °C	287.1	0.275	331.3	0.433
NP1 T-350 °C	189.7	0.437	212.9	0.515
NP1 T-450 °C	242.9	0.358	271.8	0.492
NP1 T-600 °C	370.5	0.324	402.0	0.288
NP2	373.8	0.455	364.1	0.485
NP3	300.4	0.401	298.7	0.435
NP4	220.6	0.463	307.6	0.442
NP4 T-350 °C	345.8	0.435	371.3	0.526
NP6	319.8	0.433	362.6	0.411
NP6 T-350 °C	231.9	0.525	279.3	0.481
NP6 T-450 °C	306.7	0.428	340.0	0.452
NP7	331.6	0.382	277.7	0.408

presents the DLS measurements for the same samples but with different stabilizers in water-based suspensions (chitosan, L-Dopa, CMCNa).

Figure 12 illustrates how the mean aggregate size of CMCNa-stabilized nanoparticles depends on both the carrier gas through ethanol and the temperature of the thermal treatment. Measurements were taken at two time points, 5 min and 2 h after suspension, with the sensitizer ratio correlated to the Ar flow rate through ethanol (ranging from 50 to 8 sccm) (Figure 12a). Regarding the freshly synthesized nanoparticles, it can be observed that an optimum is taking shape for the samples synthesized with 12.5 and 20 sccm Ar/ethanol, with Z-average values varying around 300 nm. Sample NP4 (10 sccm) exhibits the lowest aggregate dimension after 5 min; however, the rapid increase towards 300 nm indicates that there might be a tendency to re-agglomerate. NP3 is a satisfactory example of the suspension remaining stable over time, as opposed to NP4. Figure 12b presents the same evolution of the aggregate size, but targeting the influence of treatment temperature for the NP1 nanoparticles (50 sccm Ar through ethanol). The optimum in this case appears to be the nanoparticles treated at 350 °C. The same conclusion can be drawn for sample NP6, and relatively to in-time stability for NP4.

Table 5 presents the average size values (nm) of the iron oxide NP agglomerates using samples synthesized by laser pyrolysis (raw or thermally treated) and dispersed in distilled water. The Z-average measured for as-synthesized NPs either at 5 min or 2 h after the dispersion preparation have remarkably consistent values between 80 and 95 nm. The same feature was observed for those NPs treated at 150 °C, but at higher temperatures all the suspensions (with the exception of NP2 T-600 °C) became unstable, with very high average Z values (>1000 nm) and high PDI values. The main explanation for this is a lower NP stability in water, due to the NP surface restructuration accompanied by the loss of hydrophilic groups at higher temperatures (≥250 °C). This assumption is sustained by the very high zeta-potential values (>40 mV) for the suspensions with raw NPs and NPs treated at 150 °C, and low zeta-potential values (around 0 mV) for the suspensions containing the other thermally treated samples. Moreover, the water suspensions of both the as-synthesized NPs and those treated at 150 °C exhibit polydispersity index (PDI) values of 0.33–0.38, which indicate dispersions with narrower agglomerate size distributions, with a trend towards a monodisperse character (both after 5 min and after 2 h following preparation).

In order to improve the NP suspension stability, various stabilizers were used: chitosan, L-Dopa, and CMCNa, all of which are widely used in biomedical applications. Table 6 presents the values measured by DLS for suspensions stabilized with these three biocompatible molecules, assessing the stability of both as-synthesized and thermally treated NPs in distilled water. Slightly larger agglomerates were observed compared to nanoparticles simply dispersed in distilled water, although they remained small, with constant polydispersity indices over time.

These three stabilizers induced for all the NP suspensions (see Table 6) a satisfactory stability in time, regardless of the thermal-treatment temperature. Thus, for instance, those treated NPs can be used with optimal magnetic properties in nanofluids preparation for biomedical applications. Moreover, when chitosan or L-Dopa were used as a stabilizer, the Z-average exhibited lower values than CMCNa. An explanation might be that CMCNa polyanion seems to promote, in this case, the incipient coalescence of original oxidic NP aggregates.

Due to the particularities of the gas-phase synthesis methods [40], the laser pyrolysis under oxygen and ethanol vapors is characterized by the rapid formation (homogeneous nucleation) of nearby hot/reactive nanoparticles which stick together upon collisions, rapidly forming ramified chain-like/fractal aggregates of tens to hundreds of nm suspended in the gas flow, followed by their cooling and agglomeration [40] as fine powders collected on filter and collection chamber walls. These agglomerates are easily disrupted into the original aggregates or even in smaller ones upon strong horn ultrasonication in water [41], where their distribution of hydrodynamic diameters can be observed by DLS. As can be seen from Table 5, for all raw uncoated NPs the mean hydrodynamic size is situated between 80 and 110 nm.

A comparison study regarding our samples and other previously reported iron oxide NPs preferentially used for magnetic suspensions was performed [42,43,44,45,46,47,48,49,50,51] (see Supplementary Information, Table S1). Either as morpho-structural and magnetic properties or as components in aqueous suspensions, our samples present promising features for future applications, such as hyperthermia, drug delivery, and heat transfer agents.

4. Conclusions

The synthesis of iron oxide nanoparticles using ethanol as an energy transfer agent (sensitizer) is hereby reported. A CO₂ laser with an emission at 9.3 µm was employed, the iron pentacarbonyl and ethanol vapors are entrained by independent Ar flows, and the oxidative atmosphere is assured by a molecular oxygen flow. The experimental study investigates the influence of the ethanol flow rates on the morpho-structural features of the synthesized nanoparticles. The nanoparticles are mainly composed from disordered maghemite, with crystallite dimensions ranging from 9.9 nm to 2.4 nm based on the decrease of carrier gas flow through ethanol, from 50 to 8 sccm. For the samples analyzed by EDS and XPS, a low C content (1–2 at.%) can be observed, mainly positioned on the surface (~20 at.%). The powders were treated at different temperatures, and the influence this has on their structural and elemental composition was also analyzed. With the increase in temperature, the crystallinity of the nanoparticles is also enhanced, and are structured as ordered maghemite, the crystallite dimension also slightly increasing. BET measurements indicate significant variations in the superficial surface area based on chosen experimental conditions: 90–210 m²/g. Their magnetic features were evaluated at room temperature showing their mainly superparamagnetic character. The saturation magnetization and coercive field (very low) drop with the crystallite size: from approximately 60 emu/g to under 10 emu/g. Thermogravimetric analysis indicates a continuous weight drop between 25 and 1000 °C, with an area between 260 and 290 °C where the nanoparticles intensify their crystalline restructuring towards ordered maghemite (an exothermal process). XPS analysis reveals the presence of functional groups on the nanoparticle surface (especially Fe-OH), these groups facilitating the nanoparticles dispersibility in water. DLS measurements show aggregate sizes of 80 nm in water-based suspensions and zeta-potentials >30 mV. The stabilization protocols are highlighted and three different stabilizers are used: chitosan, L-Dopa, and CMCNa, all in distilled water. The main advantage of using ethanol vapors as a sensitizer in the obtaining of iron oxide NPs through laser pyrolysis is the low concentration of C in NP composition. This promotes excellent long-term NP suspension stability. Moreover, thermal treatment at optimal temperatures (in the range of 150–450 °C) further improves their saturation magnetization.