Journal Pre-proofs

Continuous production of magnetic iron oxide nanocrystals by oxidative pre-

cipitation

T. Asimakidou, A. Makridis, S. Veintemillas-Verdaguer, M.P. Morales, I.

Kellartzis, M. Mitrakas, G. Vourlias, M. Angelakeris, K. Simeonidis

PII: S1385-8947(20)30584-2
DOI: https://doi.org/10.1016/j.cej.2020.124593
Reference: CEJ 124593

To appear in: Chemical Engineering Journal

Received Date: 25 October 2019

Revised Date: 21 January 2020
Accepted Date: 26 February 2020

Please cite this article as: T. Asimakidou, A. Makridis, S. Veintemillas-Verdaguer, M.P. Morales, I. Kellartzis, M.
Mitrakas, G. Vourlias, M. Angelakeris, K. Simeonidis, Continuous production of magnetic iron oxide nanocrystals
by oxidative precipitation, Chemical Engineering Journal (2020), doi: https://doi.org/10.1016/j.cej.2020.124593

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will
undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing
this version to give early visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2020 Published by Elsevier B.V.

 Continuous production of magnetic iron oxide nanocrystals by

oxidative precipitation

T. Asimakidou1, A. Makridis1, S. Veintemillas-Verdaguer2, M.P. Morales2, I. Kellartzis3, M. Mitrakas3, G.

Vourlias1, M. Angelakeris1, K. Simeonidis4,5

1Department of Physics, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece

2Instituto de Ciencia de Materiales de Madrid, CSIC, Sor Juana Inés de la Cruz 3, Cantoblanco, 28049

Madrid, Spain

3Department of Chemical Engineering, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece

4Hephaestus Advanced Laboratory, Eastern Macedonia and Thrace Institute of Technology, 65404

Kavala, Greece

5Ecoresources P.C., Giannitson-Santaroza Str. 15-17, 54627 Thessaloniki, Greece

Abstract

Continuous processes are always preferred over batch ones when reproducible and scalable industrial

procedures are needed. This work illustrates the production of magnetite nanoparticles by oxidative

precipitation in aqueous media, following a continuous approach that offers additional advantages.

Particularly, the developed reaction setup succeeds (i) the complete separation of the green rust’s

precipitation from Fe3O4 nucleation, (ii) the achievement of constant concentrations in all ionic and solid

forms throughout the production line when steady-state is reached, what means constant
supersaturation from both the formation of green rust and Fe3O4, and (iii) the possibility to control

critical parameters, such as OH- excess over the initial stoichiometric Fe(OH)2 precipitation, through on-

line regulation of synthesis parameters such as the reactor’s pH and redox potential. Importantly,

continuous flow synthesis of Fe3O4 nanoparticles enables high production capacities, low energy

consumption and proportional scale-up at any volume. As a proof of concept, obtained nanoparticles

were evaluated according to their magnetic response as potential magnetic hyperthermia agents

indicating significant improvement of heating efficiency that goes up to 1.5-2 kW/gFe3O4 for both smaller

(~40 nm) and larger (~200 nm) particles.

Keywords: iron oxide, nanoparticles, large-scale production, continuous-flow process, magnetic

hyperthermia

1. Introduction

Engineered nanoparticles consisting of inorganic phases (metals, alloys, oxides, composites) stand out as

an important class of nanomaterials presenting wide chemical diversity and interesting electronic,

optical, magnetic and mechanical properties while further opportunities arise by their size-induced

coupling effects. Their role in our everyday-life gradually expands as more technological sectors,

including manufactures, electronics, energy generation and storage, sensors, biomedicine, and

environment processes, incorporate nanoparticles in order to improve efficiency [1–3]. Compatibility of

the production of such materials with aqueous chemistry, which is the base of most industrial,

environmental and biological processes, as well as the development of large-scale production schemes

with low cost and good reproducibility, are key issues need to be addressed. Among inorganic

nanoparticles, magnetic ones represent one of the most important materials characterized by their
ability to respond to the application of an external magnetic field. This response enables their

localization by Magnetic Resonance Imaging (MRI) and their use in therapy as drug carriers, biomolecule

separation using a field gradient or as heat or mechanical damage sources using an alternating magnetic

field [4]. In addition, several applications take advantage of their biocompatibility and chemical affinity

(direct or induced by appropriate coatings) towards many molecules and ions resulting, for instance, in

antimicrobial properties [5].

Magnetic iron oxide nanoparticles usually refer to those consisting of magnetite (Fe3O4) or maghemite

(γ-Fe2O3) that combine facile and cheap preparation, chemical stability, biocompatibility, sufficient

magnetism and wide range of sizes and morphologies available. The synthesis routes for such systems

are based on the control of specific parameters (pH, temperature, pressure or redox potential), needed

to generate a high supersaturated solution where nanoparticles nucleate. In addition, to achieve well-

defined chemical composition, structural and colloidal stability, the nanoparticle synthesis usually

involves protective coating by organic or inorganic molecules. However, the production of high quality

nanoparticles rarely complies with green chemistry rules using organic solvents and expensive and toxic

reagents combined with high reaction temperatures for long times. Such methods enable the

production of small quantities of highly uniform nanoparticles with specific particle size and shape [6,7].

Besides, green synthetic methods using non-toxic reagents and low energy consumption, stand on the

aqueous precipitation of iron salts with very low cost and high productivity, but they are restricted to a

certain size range (below 15 nm) with wide size distribution. As a result, magnetite nanoparticles

present modest magnetic response, and often favor the particle agglomeration with the consequent

decrease in effective surface area and heating performance.

An interesting alternative for the production of iron oxide nanoparticles following a green route is the

oxidative hydrolysis of Fe2+ in aqueous media. Thus, through the intermediate formation of green rust,

uniform magnetite nanoparticles are obtained by completely separating the hydrolysis of iron (II) from
dehydroxylation and oxidation of green rust [8]. Green rust refers to a class of mixed ferrous and ferric

iron hydroxides based on the brucite Fe(OH)2 layers where part of Fe2+ is oxidized to Fe3+ and anions and

water molecules are in the interlaminar space following the general formula [Fe2+(6-

x)Fe
3+ (OH) ]x+·[A ·yH O]x−, being A an n-valent anion such as Cl-, CO32- or SO42- and y the number of
x 12 x/n 2

interlayered water molecules crystallized in a layered double hydroxide structure. Mimicking natural

processes during which green rust is a critical compound considered as a life-generation engine [9], the

accurate control of reaction parameters is known to result in high purity and advanced crystallinity

products with improved magnetic properties [10,11]. The role of oxidant’s nature, the Fe2+ salt, the

solvent, the pH and the temperature has been previously monitored in batch processes concluding

FeSO4, NaOH, NaNO3 and ethanol/water as the combination of low-cost reagents that brings the most

stable green rust precursor following a lower dissociation rate when supersaturation is achieved [12].

However, carrying out the two-step process in a batch reactor involves a number of challenges arising

from the time evolution of critical parameters such as the actual reagents and products concentration,

the temperature and the pH during both processes. This is important not only for the reproducibility of

the process but also for the energy consumption and the dimensions of the pilot plant needed for the

synthesis. For instance, a large-scale production line would require a high-volume batch reactor and

facilities to load the reaction mixture, significant power to heat the solution from room temperature up

to nearly boiling temperature and non-negligible loss of time to evacuate, clean and reload the system.

The long time needed for mixing and heating the reaction mixture in this pilot plant implicates the

variation of critical parameters previously mentioned, and all will end up in an enhanced polydispersity

of the product. On the contrary, continuous flow microfluidic reactors manage to overcome some of

these difficulties and optimize nanoparticles characteristics compared to batch synthesis. Moreover, it is

possible to receive unusual morphologies or phase combinations that cannot be achieved otherwise or

introduce lab-on-a-chip devices to enable on-site synthesis [13–16]. Such processes take advantage of
working in the steady state of a continuous flow mode, but their potential scale-up by assembling an

extremely high number of microreactors working in parallel will imply additional difficulties and costs.

In spite of the fact that the demand of magnetic nanoparticles for developing applications is

exponentially growing, not much effort has been devoted to investigate possibilities of transferring

laboratory knowledge into large-scale production schemes. In an attempt to improve the conditions of

synthesizing Fe3O4 nanoparticles and promote a process for their industrial-scale production, this work

presents the continuous-flow version of the oxidative precipitation process able to operate under time-

independent state in stationary condition. In order to fulfill this objective, we adapted the optimum

critical parameters of the process determined previously using the batch approach to a continuous flow

process. The design of the described process was implemented keeping in mind commonly adopted

processes for the industrial production of iron oxides i.e. precipitation from salts into continuous stirred

tank reactors (CSTR). More specifically, the oxidative precipitation of Fe2+ was performed in a sequence

of two stirring reactors, in which the formation of the hydroxide gel and its ageing under mild oxidizing

conditions are carried out separately. The role of the acidity during green rust production and the

ethanol’s presence were investigated, as well as the key parameters to define the geometrical

characteristics and the magnetic properties of obtained nanoparticles. In order to test the validity of

Fe3O4 nanoparticles as potential agents for biomedical practices, their efficiency in magnetically-

triggered hyperthermia was examined.

Figure 1. Schematic representation of the laboratory continuous-flow process for the synthesis of Fe3O4

nanoparticles.

2. Experimental

Product synthesis

Magnetite nanoparticles were prepared in a continuous-flow setup (Figure 1) adapted to the reagents’

proportions used for the synthesis of 22 nm Fe3O4 spherical nanoparticles in a batch [12], after dividing

all concentrations by a factor of 10. Two separate 5 L tanks were used to store the reagent mixtures: one

for the FeSO4·7H2O solution (0.03 M) and a second one for the mixture of NaNO3 (0.06 M), NaOH (0.125

M) and ethanol (30 % v/v when used). The tanks were periodically refilled by freshly dissolved reagents

to feed the reactor. The solutions were pumped at a constant rate, 0.3 L/h and 1.2 L/h respectively, in a

0.5 L reactor operating at room temperature under mild mechanical stirring and intense nitrogen

bubbling (residence period 20 min). Successful application of inert conditions in this part proved to be

critical to completely inhibit oxidation by air and avoid formation of red-colored Fe3+ oxyhydroxides. The

pH in this reactor was kept to a specific value for each experiment by the periodical addition of drops of
NaOH or H2SO4 solutions (one value in the range 8-12±0.2 as shown in Table 1). In this reactor, the rapid

formation of green rust takes place according to the reaction:

5 Fe(OH)2 + Fe2+ + SO42- + 0.25 NO3- + (n+1.5) H2O → [Fe2+4,Fe3+2,(OH)12]2+·[SO4·nH2O]2− + 0.25 NH3 + 0.25
OH-

The green rust slurry was then pumped into the ageing reactor, consisting of two consecutive stirring

tanks of 3 L each, placed into a water bath heated at 90 oC and surrounded by an insulating top to avoid

water evaporation. In this way, green rust immediately meets maximum temperature avoiding the usual

half-an-hour heating period in batch process. Relatively intense stirring and nitrogen bubbling is applied

to the first reactor to homogenize the mixture whereas slow stirring is applied to the second one to

allow better crystallization of the product. The total residence period in the ageing reactor was set at 4

h, equally split in each tank, and this value was used as the design parameter for the whole system.

During ageing time, the reduction of nitrate ions caused the conversion of green rust to magnetite [17–

19]:

4[Fe2+4Fe3+2(OH)12]2+·[SO4·nH2O]2−+ NO3-+7OH-→ 8 Fe3O4 + NH3 + 4SO42- + (26+n)H2O

Finally, the black colored dispersion is collected in the outflow of the ageing reactor, cooled down to

room temperature and then, washed several times with distilled water until a conductivity below 1

mS/cm was obtained. The pH during the production of the intermediate iron hydroxyl-sulfate and the

presence of ethanol were used as the controlling parameter defining the reaction efficiency and the

characteristics of obtained nanoparticles. In all cases, a drop of the pH was observed at the end of the

ageing process as a consequence of the oxidation pathway [12]. An overview of the synthesis

parameters for the studied samples is presented in Table 1.

The relative decrease of the OH- concentration (Table 1) is considered representative of the green rust’s

oxidation extend and increases in the presence of ethanol and at lower pH values. It should be explained

that the OH- decrease was estimated by the pH and the ionic product decrease and it is related to the
weakening of hydrogen-bonding network and the reduction of dielectric constant in the case of ethanol,

or the weakening of the green rust’s stability in the case of the low starting pH.

Table 1. Experimental parameters used for the Fe3O4 nanoparticles synthesis in the continuous-flow

setup at 90 oC.

pH value [OH-]
Sample Solvent
Green rust Final decrease %

S1 12.5 12.0 H2O/EtOH 63

S2 11.6 10.9 H2O/EtOH 75

S3 10.8 10.1 H2O/EtOH 79

S4 8.4 6.7 H2O/EtOH 100

S5 12.3 12.1 H2O 38

S6 11.3 11.0 H2O 50

S7 10.1 9.8 H2O 50

S8 9.2 6.7 H2O 100

Characterization

Structural-phase identification on dried samples was performed by powder X-ray diffractometry (XRD)

using a water-cooled Rigaku Ultima+ diffractometer with CuKa radiation, a step size of 0.05° and a step

time of 3 s, operating at 40 kV and 30 mA. The diffraction patterns were compared to the Powder

Diffraction Files (PDF) database [20] and the crystallite sizes were determined using the Scherrer’s

equation. The morphology and the mean particle size distributions were determined through

transmission electron microscopy (TEM) images obtained using a JEOL JEM1010 microscope operating at

100 kV. Samples were prepared by placing a drop of the particles suspended in water onto a carbon
coated copper grid and allowing it to dry at room temperature. The specific surface area of the

nanoparticles was estimated by nitrogen gas adsorption at 77 K using a micropore surface area analyzer

according to the Brunauer-Emmett-Teller (BET) model.

The ratio of Fe2+/Fe3+ was determined by acid digestion of the nanoparticles and titration using KMnO4

solution [21,22]. A sample quantity of 0.1 g was dissolved under heating in 50 mL 7 M H2SO4 (pre-

treated with nitrogen bubbling) and titrated with 0.05 M KMnO4. The end point of the titration was

defined by the persisting weak pink color, indicating that the MnO4− ions were no longer being reduced.

Total iron content was measured after sample dissolution in HCl, by graphite furnace atomic absorption

spectrophotometry using a Perkin Elmer AAnalyst 800 instrument.

Colloidal properties

A Zetasizer nano ZS by Malvern Instruments was used for the determination of the hydrodynamic size

and the ζ-potential. The dynamic light scattering (DLS) allows the measurements of the aggregate size in

dispersion among 0.6 and 6000 nm. A log-normal distribution function in intensity was used to fit the

size data obtained. The Zetasizer nano ZS calculates the ζ-potential by determining the electrophoretic

mobility using laser Doppler velocimetry and applying the Smoluchowski approach. All the

measurements were performed at room temperature using 0.001 M KNO3 as background electrolyte

and HNO3 and KOH to change the suspension pH.

Magnetic measurements

A vibrating sample magnetometer MagLab VSM (Oxford Instrument) was used for the measurement of

the magnetic properties up to a maximum field 1 T (~800 kA/m). Coercive field and saturation

magnetization values were obtained from the hysteresis loops recorded at room temperature. The

saturation magnetization values (Ms) expressed in Am2/kgFe3O4 were determined by extrapolating to

infinite field the magnetization values in the linear region of M versus 1/H. The corresponding minor
loops measured after demagnetization at a maximum DC field 24 kA/m were also recorded as an

indication of the AC field heating potential introduced by hysteresis losses.

Induction heating efficiency

Calorimetric measurements on the heating efficiency of nanoparticles under an AC magnetic field were

acquired using a commercial converted 4.5 kW inductive heater at 765 kHz frequency and 24 kA/m

magnetic field intensity. The specific loss power (SLP), or specific absorption rate (SAR) used to evaluate

the heating performance was derived from the slope of the temperature versus time curve after

subtracting water background signal and heat losses to the environment. Data are given in Watts per

gram of Fe3O4. In addition, corresponding intrinsic loss power (ILP) was calculated after the expression

ILP=SLP/(f·H2) with f and H being the AC frequency and the field intensity, respectively. Temperature

was monitored by using an Opsens PicoM device with a GaAs-based fiber optic probe.

3. Results and discussion

All samples prepared in pure water and water/ethanol mixtures under different pH conditions (Table 1)

by the continuous flow process present a ferrite iron oxide structure as shown in the XRD diagrams

(Figure 1). Chemical analysis identifies them as solid solutions of magnetite-maghemite but, considering

the high Fe2+/Fe3+ ratio, they will be called for short according to the dominant crystal phase which is

magnetite (Table 2). It can be observed that below pH 10 (Samples 3, 4, 7, 8) the formation of iron oxy-

hydroxides such as α-FeOOH was also favored. This suggests that under these conditions the oxidation

of green rust takes place more rapidly than its dehydroxylation and therefore, the complete

transformation of intermediate products to magnetite does not occur. These intermediates and residual

green rust are oxidized to α-FeOOH when the sample is exposed to the atmosphere [23]. Consequently,

a reduction of the Fe2+/Fe3+ ratio is observed in the samples prepared in pure water, as it is the case

sample 8 (Table 2). The average size of crystal domains for each sample was evaluated by applying the
Scherrer’s equation [24], on the broadening of (311) Fe3O4 peak (Table 2). These values were also

compared to the actual particle size from microscopy observations to define the degree of

polycrystallinity of the particles in each case. An improvement of the crystallinity is observed when the

pH is around 10-11 reaching a maximum value of around 50 nm that falls off to less than half at

synthesis pH around 7. Such finding shall be attributed to the modification of crystal growth mechanism

as defined by the lower reaction rate of green rust at less alkaline environment.

Figure 2. XRD diagrams of samples prepared at pH 8.4-12.5 in H2O/EtOH (S1-S4) and samples prepared

at pH 9.2-12.3 in H2O (S5-S8).

On the other hand, the effect of reaction conditions during growth becomes more obvious in the

geometry of the obtained nanoparticles. TEM images in Figure 3 provide evidence about the different

shape and dimensions of magnetite nanoparticles synthesized in continuous flow reactor. The most

significant effect comes by the presence of ethanol which generally results in smaller nanocrystals in the

range of 40-53 nm whereas pure water favors much larger particles in the range of 150-310 nm with

higher polydispersity. Working at intense alkaline conditions in water/ethanol mixture enables the
formation of almost spherical nanoparticles reaching a mean diameter of 53 nm at pH 12 and 43 nm at

pH 11. At lower pH values, cubic-shaped nanocrystals dominate with a size of around 40 nm (diagonal

dimension). On the contrary, nanoparticles grown in pure water achieve almost one order of magnitude

larger dimensions forming well-defined nanocubes with their size strongly dependent on the synthesis

pH. Particularly, at pH 12 the cube diagonal slightly overcomes 300 nm but gradually decreases below

200 nm at pH 10 [25,26]. It has been reported that magnetite nanocubes are favorable under small

excess of OH-, i.e. when all Fe2+ nucleates in the form of Fe(OH)2. However, no OH- excess gives larger

octahedral nanocrystals and Fe2+ excess produces larger spherical (even submicronic) particles [10].

Under neutral conditions (pH 7) close to the isoelectric point of magnetite, nanoparticles tend to

agglomerate and a very polydisperse sample of spherical nanoparticles is obtained. It should be noted

that rod-shaped formations observed in samples 4, 6 and 7 correspond to the α-FeOOH that appears as

a synthesis byproduct, as confirmed by XRD [27].

In comparison to similar studies on the oxidative precipitation of magnetite nanocrystals in a batch

reactor [12], particle sizes of the continuous-flow process appear to be significantly enlarged. For

instance, magnetite formation using Na+ and SO42- as counterions in a 25 % ethanol/water mixture was

reported to result in monodisperse nanoparticles of 21.2 nm. The difference can be attributed to the

nucleation process at iron concentrations ten times higher in the case of the batch reactor leading to a

huge number of nuclei unable to grow further. The nucleation step is the initial precipitation of ferrous

ions in green rust that serves as substrate for the magnetite nanoparticles growth by means of

oxidation, dehydration and co-precipitation. The ten times reduction of concentration of iron with

respect to previous batch experiments caused a smaller nucleation rate. The smaller number of nuclei

formed leads in larger particles [28]. As a consequence, the corresponding dimensions of nanocrystals

prepared by the continuous-flow system are almost double in size (40-50 nm). Compared to similar

batch synthesis, the polydispersity index for continuous-flow system appears slightly smaller lying in the
range of 0.10-0.15. It should be underlined that the described low-concentration effect only partially

explains the observed size variation since batch synthesis at the same concentrations than the

continuous-flow procedure, indicated a small drop of the size for S1, ~43 nm (Supporting information).

This is an evidence about the significant role of setting the reaction parameters as time-independent in

the continuous-flow setup, in the nucleation/growth mechanism.

Table 2. Overview of the physical, chemical and morphological properties of the samples produced in

the continuous-flow system.

Crystal size BET Ms

Sample Solvent Shape Size (nm) Fe2+/Fe3+
(nm) (m2/g) (Am2/kg)

S1 H2O/EtOH Sphere 53±7.7 29.2 30 0.45 79.5

S2 H2O/EtOH Sphere 43±6.7 32.1 32 0.47 85.6

S3 H2O/EtOH Cubic 39±6.1 42.7 34 0.44 82.7

S4 H2O/EtOH Cubic 40±6.6 31.0 33 0.37 82.3

S5 H2O Cubic 310±28.6 45.6 16 0.42 76.6

S6 H2O Cubic 260±24.5 54.8 12 0.42 90.0

S7 H2O Cubic 190±25.9 55.5 15 0.37 90.2

150±21.8/
S8 H2O Sphere 24.1 14 0.28 56.3
280±28.9

The differences get even higher in pure water synthesis when the nanoparticles size goes from 36.9 nm

in a batch reactor at ten times higher iron concentrations or 55 nm at similar concentrations, up to 190

nm in the continuous-flow system. Here, key parameters related to the growth step dominate, i.e. the

transformation of green rust to magnetite. Thus, the use of water affects the speciation of precipitating
ionic forms and, as a result, oxidation and dehydroxylation rate. When the pH increases, larger

quantities of Na+ and OH- are included in the layered double hydroxide structure of the green rust

leading to larger magnetite particles [29]. It should be mentioned that under less alkaline conditions,

there is a tendency to produce intermediate products such as lepidocrocite (γ-FeOOH) or goethite (α-

FeOOH), especially for samples synthesized in pure water [30]. On the other hand, oxidation-

dehydroxylation of the initially formed green rust proceeds under very low Fe2+ and green rust

concentrations, very high residual sulfate concentrations and in the presence of formed magnetite

nanocrystals. Such parameters allow a slow reaction rate probably heterogeneously-assisted on the

surface of existing nanoparticles and therefore, larger particles are attained compared to a similar batch

reaction.

In accordance to the size of synthesized magnetite nanocrystals in the continuous-flow system, the

specific surface area of the samples prepared in the water/ethanol mixture is in the range of 30-35 m2/g

but decreases to around half of this value for the samples grown in pure water (12-16 m2/g). When

samples are dispersed in water forming a colloidal suspension, the actual organization of the

nanoparticles in aggregates is reflected in the DLS measurements concerning the hydrodynamic

diameter (Figure 4). Samples grown in pure water generally form larger aggregates with a mean

hydrodynamic size of 2400 nm, in particular for the sample synthesized at pH 11. The corresponding

sample produced in water-ethanol mixture show smaller hydrodynamic diameters with a bimodal

distribution with peaks located at 170 and 780 nm.

Figure 3. TEM images of samples prepared at pH 8.4-12.5 in H2O/EtOH (S1-S4) and samples prepared at

pH 9.2-12.3 in H2O (S5-S8) by the continuous-flow system. Note the differences in scale bars in the two

rows.

The effective surface charge of the obtained nanoparticles in aqueous colloidal suspensions was

signified by the electrophoretic mobility measurements at different pH values. Figure 5 presents the ζ-

potential curves for nanoparticles prepared at pH 11 in water-ethanol mixture (sample 2) and pure

water (sample 6). Isoelectric points for both samples were found to be shifted below pH 5.5 indicating

the important role of OH- and SO42- excess in the determination of a negatively-charged particle surface

through their favorable adsorption. In the absence of ethanol, the surface modification effect is more

intense and the nanoparticles show a lower isoelectric point (pH 4.7) compared to the smaller

nanocrystals produced in water-ethanol (pH 5.3). Similar results concerning the contribution of ethanol

in the stabilization of surface charge were observed for magnetite nanoparticles prepared by the
oxidative precipitation of Fe(II) sulfate in a batch [10]. Interestingly, a typical nanoparticles dispersion (1

g/L) has a time window of colloidal stability of up to 1 h.

The magnetic properties of the obtained nanoparticles are consistent with the presence of a well-

defined magnetite structure and the stoichiometric Fe2+/Fe3+ratio. All samples follow a ferrimagnetic

behavior at room temperature which is reasonable for nanoparticles lying well-above the

superparamagnetic limit for magnetite (>20 nm). More specifically, the saturation magnetization

estimated from the hysteresis loops (Supporting information) is equal to the reported values for bulk

magnetite for many samples [31]. An important deviation to lower magnetization values is observed for

nanoparticles grown in pure water and at low pH values (sample 8). As mentioned, the crystal growth

mechanism is deteriorated when working at neutral pH values resulting in a polydispersed system

containing iron oxy-hydroxides. Coercive field shows non-zero values varying within a narrow range 6.4-

8 kA/m.

Figure 4. Hydrodynamic size calculated by DLS measurements of the colloidal suspensions of magnetic

nanoparticles prepared at pH 11 in the continuous-flow system.

 Figure 5. Zeta-potential measurements of samples prepared at pH 11 in the continuous-flow system.

Figure 6 shows a comparison of the minor loops obtained for samples synthesized in water/ethanol and

pure water at a maximum field equal to the field applied in the magnetic hyperthermia experiments of

this work (24 kA/m). Although measurements were taken in DC field, it has been shown that the

integration of the area included in the loops is in general proportional to the expected heating efficiency

when hysteresis losses are the main heating mechanism at the radiofrequency AC fields used in

magnetic hyperthermia. In this case, samples prepared under strong alkaline conditions provide the

larger hysteresis area. In particular, sample 5 (pH 12 in water) brings the maximum value approaching

170 J/g whereas samples 1 and 2 (pH 12 and 11 in water/ethanol) show similar large hysteresis areas,

reaching 110 and 150 J/g, respectively. For the rest of the samples, hysteresis areas within the range 50-

80 J/g were measured.

Figure 6. Minor magnetic hysteresis loops at maximum field 24 kA/m for the studied samples prepared

in the continuous-flow system.

The comparison of the hysteresis areas of the minor loops measured in powders and the SLP values

estimated by temperature versus time curves in dispersions of the magnetite nanocrystals are shown in

Figure 7. It is very interesting to see that SLP values obtained under an AC magnetic field are directly

proportional to the hysteresis areas obtained from a DC magnetic measurement suggesting hysteresis

losses as the dominant mechanism of heat generation. Such finding is reasonable considering that the

dimensions are well above the superparamagnetic limit, i.e. within the monodomain range 20-70 nm
(water/ethanol) or in the multi-domain range, >70 nm (pure water) and their behavior is clearly

ferrimagnetic.

Importantly, the heating efficiency for the nanoparticles prepared under strong alkaline conditions is

among the highest ones reported for single-phase magnetite particles developed by the oxidative

precipitation method, with SLP values overcoming 1 kW/g or even higher, as it is the case of sample 5

whose SLP approaches 2 kW/g with respect to Fe3O4 mass. For comparison, the corresponding samples

prepared in a batch resulted in SLP values lower than 0.5 kW/g (708 kHz, 24 mT) [12]. These differences

are less significant when measuring the SLP at lower frequencies and field strengths, going down by one

order of magnitude [10,32] (check Supporting Information). It should be noted that much higher heating

efficiencies, even overcoming 10 kW/g, have been reported for combinations of Co, Zn, Mn prepared by

organic synthesis [33–35]. However, those synthesis methods cannot be defined as green and low-cost,

because of the presence of potentially toxic elements for biomedical applications [36]. Lower but not

negligible values in the range 0.25-0.75 kW/g were measured for samples prepared at less alkaline

environments. The ILP index for the most efficient samples was found between 3-4.4 nH·m2/kg whereas

samples prepared at less alkaline conditions presented values in the range 0.6-1.4 nH·m2/kg.

Therefore, regarding magnetic hyperthermia as an application of major interest for magnetite

nanoparticles, optimum conditions for its synthesis in water/ethanol mixture is the use of a pH above

11, while when working in pure water, the pH should be adjusted to 12. It must be underlined that the

obtained nanoparticles in pure water are almost one order of magnitude larger than those for

water/ethanol, limiting the potential applicability in biomedical uses. Other important applications of

these nanoparticles are now under study such as catalysis or water remediation [37–39].
Figure 7. Specific loss power dependence with synthesis pH and solvent for magnetite nanoparticle

colloidal suspensions prepared by continuous-flow system, measured at 24 kA/m and 765 kHz (a).

Corresponding magnetic hysteresis area estimated by minor loops up to 24 kA/m at room temperature

(b).

4. Conclusions

The continuous flow process developed in this work allows the preparation of uniform magnetite

nanoparticles with sizes between 40 and 300 nm and very interesting magnetic properties. The

continuous operation using the proposed concentrations and flowrates resulted in narrow dispersion of
properties, using reagents that are non-toxic or substantially less toxic and an important reduction in

waste generation in comparison to other synthesis methods in batches. Importantly, the system may

operate uninterruptedly providing reproducible nanoparticles since all reaction parameters are fixed at

any stage and complete homogenization is achieved. Furthermore, the “extracted” parameters by the

continuous flow synthesis of 0.7 g Fe3O4/h enable the design of a full-scale unit for the production of

huge quantities by just multiplying the tanks’ volumes and the flowrates. New industrial applications of

magnetite nanoparticles in catalysis, biotechnology, water remediation, etc. or in general, in the area of

environmental remediation, which is a subject of growing interest due to the superior performance of

nanomaterials compared to conventional technologies, could benefit from this work.

Acknowledgements

Scientific work was financially supported through an IKY scholarship (MIS 5033021) co-financed by E.U.

(European Social Fund - ESF) and Greek national funds through the action "Reinforcement of

Postdoctoral Researchers-2nd call" of the Operational Programme "Human Resources Development

Program, Education and Lifelong Learning" of the National Strategic Reference Framework (NSRF) 2014

– 2020. Part of the study was implemented within the frame of MagnoTher project funded by Stavros

Niarchos Foundation and Eastern Macedonia and Thrace Institute of Technology fellowships for assisting

young scientists in prototyping innovative products by using cutting-edge technology. Funding by the

Spanish Ministry of Economy and Competitiveness, COMANCHE project, No MAT2017-88148-R and the

EU project H2020-FETOPEN- RIA 829162, HOTZYMES is also acknowledged.

References
[1] M. Colombo, S. Carregal-Romero, M.F. Casula, L. Gutiérrez, M.P. Morales, I.B. Böhm, J.T.

Heverhagen, D. Prosperi, W.J. Parak, Biological applications of magnetic nanoparticles, Chem.

Soc. Rev. 41 (2012) 4306. https://doi.org/10.1039/c2cs15337h.

[2] K. Simeonidis, C. Martinez-Boubeta, P. Zamora-Pérez, P. Rivera-Gil, E. Kaprara, E. Kokkinos, M.

Mitrakas, Implementing nanoparticles for competitive drinking water purification, Environ.

Chem. Lett. 17 (2019) 705–719. https://doi.org/10.1007/s10311-018-00821-5.

[3] D. V. Talapin, J.S. Lee, M. V. Kovalenko, E. V. Shevchenko, Prospects of colloidal nanocrystals for

electronic and optoelectronic applications, Chem. Rev. 110 (2010) 389–458.

https://doi.org/10.1021/cr900137k.

[4] Nguyen TK Thanh, ed., Clinical applications of magnetic nanoparticles : design to diagnosis

manufacturing to medicine, 1st Editio, CRC Press, Boca Raton, 2018.

[5] H. Maleki, A. Rai, S. Pinto, M. Evangelista, R.M.S. Cardoso, C. Paulo, T. Carvalheiro, A. Paiva, M.

Imani, A. Simchi, L. Durães, A. Portugal, L. Ferreira, High Antimicrobial Activity and Low Human

Cell Cytotoxicity of Core–Shell Magnetic Nanoparticles Functionalized with an Antimicrobial

Peptide, ACS Appl. Mater. Interfaces. 8 (2016) 11366–11378.

https://doi.org/10.1021/acsami.6b03355.

[6] G. Salas, C. Casado, F.J. Teran, R. Miranda, C.J. Serna, M.P. Morales, Controlled synthesis of

uniform magnetite nanocrystals with high-quality properties for biomedical applications, J.

Mater. Chem. 22 (2012) 21065. https://doi.org/10.1039/c2jm34402e.

[7] B.H. Kim, N. Lee, H. Kim, K. An, Y. Il Park, Y. Choi, K. Shin, Y. Lee, S.G. Kwon, H. Bin Na, J.-G. Park,

T.-Y. Ahn, Y.-W. Kim, W.K. Moon, S.H. Choi, T. Hyeon, Large-Scale Synthesis of Uniform and

Extremely Small-Sized Iron Oxide Nanoparticles for High-Resolution T1 Magnetic Resonance

 Imaging Contrast Agents, J. Am. Chem. Soc. 133 (2011) 12624–12631.

https://doi.org/10.1021/ja203340u.

[8] T. Sugimoto, E. Matijević, Formation of uniform spherical magnetite particles by crystallization

from ferrous hydroxide gels, J. Colloid Interface Sci. 74 (1980) 227. https://doi.org/10.1016/0021-

9797(80)90187-3.

[9] M.J. Russell, Green rust: The simple organizing ‘seed’ of all life?, Life. 8 (2018).

https://doi.org/10.3390/life8030035.

[10] M.A. Verges, R. Costo, A.G. Roca, J.F. Marco, G.F. Goya, C.J. Serna, M.P. Morales, M. Andrés

Vergés, R. Costo, A.G. Roca, J.F. Marco, G.F. Goya, C.J. Serna, M.P. Morales, Uniform and water

stable magnetite nanoparticles with diameters around the monodomain-multidomain limit, J.

Phys. D-Applied Phys. 41 (2008) 134003. https://doi.org/Artn 134003\rDoi 10.1088/0022-

3727/41/13/134003.

[11] M.A. Gonzalez-Fernandez, T.E. Torres, M. Andrés-Vergés, R. Costo, P. de la Presa, C.J. Serna, M.P.

Morales, C. Marquina, M.R. Ibarra, G.F. Goya, Magnetic nanoparticles for power absorption:

Optimizing size, shape and magnetic properties, J. Solid State Chem. 182 (2009) 2779–2784.

https://doi.org/10.1016/j.jssc.2009.07.047.

[12] Y. Luengo, M.P. Morales, L. Gutiérrez, S. Veintemillas-Verdaguer, Counterion and solvent effects

on the size of magnetite nanocrystals obtained by oxidative precipitation, J. Mater. Chem. C. 4

(2016) 9482–9488. https://doi.org/10.1039/C6TC03567A.

[13] C.D. Ahrberg, J.W. Choi, B.G. Chung, Droplet-based synthesis of homogeneous magnetic iron

oxide nanoparticles, Beilstein J. Nanotechnol. 9 (2018) 2413–2420.

https://doi.org/10.3762/bjnano.9.226.
[14] G. Salazar-Alvarez, M. Muhammed, A.A. Zagorodni, Novel flow injection synthesis of iron oxide

nanoparticles with narrow size distribution, Chem. Eng. Sci. 61 (2006) 4625–4633.

https://doi.org/10.1016/j.ces.2006.02.032.

[15] W. Glasgow, B. Fellows, B. Qi, T. Darroudi, C. Kitchens, L. Ye, T.M. Crawford, O.T. Mefford,

Continuous synthesis of iron oxide (Fe3O4) nanoparticles via thermal decomposition,

Particuology. 26 (2016) 47–53. https://doi.org/10.1016/j.partic.2015.09.011.

[16] L. Uson, M. Arruebo, V. Sebastian, J. Santamaria, Single phase microreactor for the continuous,

high-temperature synthesis of <4 nm superparamagnetic iron oxide nanoparticles, Chem. Eng. J.

340 (2018) 66–72. https://doi.org/10.1016/j.cej.2017.12.024.

[17] J.T. Moraghan, R.J. Buresh, Chemical Reduction of Nitrite and Nitrous Oxide by Ferrous Iron, Soil

Sci. Soc. Am. J. 41 (1977) 47. https://doi.org/10.2136/sssaj1977.03615995004100010017x.

[18] H.C.B. Hansen, O.K. Borggaard, J. Sørensen, Evaluation of the free energy of formation of Fe(II)-

Fe(III) hydroxide-sulphate (green rust) and its reduction of nitrite, Geochim. Cosmochim. Acta. 58

(1994) 2599–2608. https://doi.org/10.1016/0016-7037(94)90131-7.

[19] H.C.B. Hansen, C.B. Koch, H. Nancke-Krogh, O.K. Borggaard, J. Sørensen, Abiotic nitrate reduction

to ammonium: Key role of green rust, Environ. Sci. Technol. 30 (1996) 2053–2056.

https://doi.org/10.1021/es950844w.

[20] Joint Center for Powder Diffraction Studies, ed., Powder Diffraction File, 2004t ed., International

Centre for Diffraction Data, Newtown Square, PA, n.d.

[21] S.J. Kemp, R.M. Ferguson, A.P. Khandhar, K.M. Krishnan, Monodisperse magnetite nanoparticles

with nearly ideal saturation magnetization, RSC Adv. 6 (2016) 77452–77464.

https://doi.org/10.1039/c6ra12072e.
[22] G.M. da Costa, C. Blanco-Andujar, E. De Grave, Q.A. Pankhurst, Magnetic Nanoparticles for in

Vivo Use: A Critical Assessment of Their Composition, J. Phys. Chem. B. 118 (2014) 11738–11746.

https://doi.org/10.1021/jp5055765.

[23] Y. Tamaura, P. V. Buduan, T. Katsura, Studies on the oxidation of iron(II) ion during the formation

of Fe3O4 and α-FeO(OH) by air oxidation of Fe[OH]2 suspensions, J. Chem. Soc., Dalt. Trans.

(1981) 1807–1811. https://doi.org/10.1039/DT9810001807.

[24] A.L. Patterson, The Scherrer formula for X-ray particle size determination, Phys. Rev. 56 (1939)

978–982. https://doi.org/10.1103/PhysRev.56.978.

[25] A. Géhin, C. Ruby, M. Abdelmoula, O. Benali, J. Ghanbaja, P. Refait, J.M.R. Génin, Synthesis of

Fe(II-III) hydroxysulphate green rust by coprecipitation, Solid State Sci. 4 (2002) 61–66.

https://doi.org/10.1016/S1293-2558(01)01219-5.

[26] C. Ruby, C. Upadhyay, A. Géhin, G. Ona-Nguema, J.-M.R. GÉnin, In Situ Redox Flexibility of FeII-III

Oxyhydroxycarbonate Green Rust and Fougerite, Environ. Sci. Technol. 40 (2006) 4696–4702.

https://doi.org/10.1021/es0606834.

[27] C. Pantke, M. Obst, K. Benzerara, G. Morin, G. Ona-Nguema, U. Dippon, A. Kappler, Green Rust

Formation during Fe(II) Oxidation by the Nitrate-Reducing Acidovorax sp. Strain BoFeN1, Environ.

Sci. Technol. 46 (2012) 1439–1446. https://doi.org/10.1021/es2016457.

[28] J.W. Mullin, Crystallization, 4th Editio, Butterworth-Heinemann, Oxford, UK, 2001.

[29] A. Al Mamun, A. Onoguchi, G. Granata, C. Tokoro, Role of pH in green rust preparation and

chromate removal from water, Appl. Clay Sci. 165 (2018) 205–213.

https://doi.org/10.1016/j.clay.2018.08.022.

[30] M. Usman, J.M. Byrne, A. Chaudhary, S. Orsetti, K. Hanna, C. Ruby, A. Kappler, S.B. Haderlein,
 Magnetite and Green Rust: Synthesis, Properties, and Environmental Applications of Mixed-

Valent Iron Minerals, Chem. Rev. 118 (2018) 3251–3304.

https://doi.org/10.1021/acs.chemrev.7b00224.

[31] B.D. Cullity, C.D. Graham, Introduction to Magnetic Materials, John Wiley & Sons, Inc., Hoboken,

NJ, USA, 2008. https://doi.org/10.1002/9780470386323.

[32] M. Marciello, V. Connord, S. Veintemillas-Verdaguer, M.A. Vergés, J. Carrey, M. Respaud, C.J.

Serna, M.P. Morales, Large scale production of biocompatible magnetite nanocrystals with high

saturation magnetization values through green aqueous synthesis, J. Mater. Chem. B. 1 (2013)

5995. https://doi.org/10.1039/c3tb20949k.

[33] S.H. Moon, S.H. Noh, J.H. Lee, T.H. Shin, Y. Lim, J. Cheon, Ultrathin Interface Regime of Core-Shell

Magnetic Nanoparticles for Effective Magnetism Tailoring, Nano Lett. (2017).

https://doi.org/10.1021/acs.nanolett.6b04016.

[34] S.H. Noh, W. Na, J.T. Jang, J.H. Lee, E.J. Lee, S.H. Moon, Y. Lim, J.S. Shin, J. Cheon, Nanoscale

magnetism control via surface and exchange anisotropy for optimized ferrimagnetic hysteresis,

Nano Lett. 12 (2012) 3716–3721. https://doi.org/10.1021/nl301499u.

[35] J.-H. Lee, J.-T. Jang, J.-S. Choi, S.H. Moon, S.-H. Noh, J.-W. Kim, J.-G. Kim, I.-S. Kim, K.I. Park, J.

Cheon, Exchange-coupled magnetic nanoparticles for efficient heat induction., Nat. Nanotechnol.

6 (2011) 418–22. https://doi.org/10.1038/nnano.2011.95.

[36] S.H. Moon, S. Noh, J.-H. Lee, T.-H. Shin, Y. Lim, J. Cheon, Correction to Ultrathin Interface Regime

of Core–Shell Magnetic Nanoparticles for Effective Magnetism Tailoring, Nano Lett. 17 (2017)

3989–3989. https://doi.org/10.1021/acs.nanolett.7b01759.

[37] A. Pratt, Environmental applications of magnetic nanoparticles, Front. Nanosci. 6 (2014) 259–
 307. https://doi.org/10.1016/B978-0-08-098353-0.00007-5.

[38] S.C.N. Tang, I.M.C. Lo, Magnetic nanoparticles: Essential factors for sustainable environmental

applications, Water Res. 47 (2013) 2613–2632. https://doi.org/10.1016/j.watres.2013.02.039.

[39] A.-H. Lu, E.L. Salabas, F. Schüth, Magnetic Nanoparticles: Synthesis, Protection, Functionalization,

and Application, Angew. Chemie Int. Ed. 46 (2007) 1222–1244.

https://doi.org/10.1002/anie.200602866.
Declaration of interests

☒ The authors declare that they have no known competing financial interests or personal relationships
that could have appeared to influence the work reported in this paper.

☐The authors declare the following financial interests/personal relationships which may be considered
as potential competing interests:

On behalf of all authors

Konstantinos Simeonidis
 Highlights

 Industrial-scale preparation of low-cost magnetite nanoparticles

 Transfer of laboratory synthesis to a continuous-flow process
 Method based on oxidative precipitation of Fe2+ controlled by OH- excess
 Nanoparticles’ dimensions range from 40-300 nm
 Improved magnetic hyperthermia efficiency validated for developed nanoparticles