The synthesis of Fe₃O₄/polyaniline (Fe₃O₄/PANI) nanomaterials by a chemical method is presented in this paper. The X-ray diffraction (XRD) shows that the lattice constant a = 8.376 A⁰ and the particle size of is about 14.5 nm for all samples, since polymer cannot influence the crystal structure of Fe₃O₄. The transmission electron microscopy (TEM) images show that the Fe₃O₄ grain sizes vary from 13 nm to 20 nm. The results of Raman spectral analysis and thermal gravimetric analysis reveal that the PANI partly forms in the Fe₃O₄/PANI nanomaterials samples. Thus, the grain size of Fe₃O₄/PANI nanomaterials is about of 25–30 nm, which has been confirmed by a scanning electron microscope (SEM). The saturated magnetic moment of Fe₃O₄/PANI samples is decreased from 66 emu/g to 39.7 emu/g with PANI content varying from 5% to 15%. However, Fe₃O₄/PANI nanomaterials are stable on chemical-physical properties and lead to improve an arsenic adsorption ability. In addition, Fe₃O₄/PANI sample with PANI 5% content has the highest arsenic adsorption ability in pH 7. In strong acidic or basic media, the arsenic adsorption of magnetic nanoparticles is insignificant. The results suggest the desorb can be conducted at pH 14 then the materials could reabsorb in further trials.

Fig. 8 Determination of balance time of arsenic absorption. Inset: Remained arsenic content depended on the pH.

Transition metal doped Fe₃O₄ nanomaterials attract attention of the scientists due to their wide application, especially in water treatment, heavy metal ions removal^1–16). The magnetic nanoparticles exhibit special properties when their specific surface area increases. Recently, the Cd, Cr, As removals were observed using the metal oxide hetero-structures, as meso-porous MnO₂, Al₂O₃, TiO₂, CuO, ZrO₂^1,17–23). But, it is difficult to recover the residuals in the media when using these oxides. In this circumstance, magnetic nanoporous materials are considered as alternative candidate because it's easy to recover them from the solution by using external magnetic field.

However, Fe₃O₄ is easily oxidized into γ-Fe₂O₃ in air that leads to a decrease in its magnetization. Large numbers of studies have been conducted focusing on chemical stability of magnetic nanomaterials, that simultaneously improve the saturated magnetization as well as enhance its adsorption capacity^{10,19,24–29)}.

To maintain the chemical properties of material, Fe₃O₄ was doped with suitable transition metals. The unit cell of Fe₃O₄ consists of two kinds of interstitial sites denoted as tetrahedral A-site (Fe³⁺) and octahedral B-site (Fe²⁺:Fe³⁺ = 1:1)^19,25). These sites are occupied by metal ions depending on their radii, electrostatic energies of the lattice and the matching of the electronic configuration with the surrounding oxygen ions. The divalent metal ions like Mn²⁺, Cu²⁺ and Zn²⁺ usually substituted into A- site formed to the inverse spinel ferrite Fe_1−xMn_xFe₂O₄^27,30) (x = 0 ÷ 0.15 or 0.5), Fe_1−xCu_xFe₂O₄³⁰⁾ (x = 0 ÷ 0.15), Fe_1−xZn_xFe₂O₄¹⁹⁾ (x = 0; 0.2; 0.4; 0.5). Some works reported the partial substitutions of Ni²⁺ ions into B-sites of Ni_xFe_3-xO₄ samples (x = 0.04; 0.06; 0.11)²⁵⁾, and of Ni_1−xZn_xFe₂O₄ samples (x = 0, 0.5 and 1)²⁶⁾. In addition, the synthesized procedure also influence to the spinel type of the Cu doped Fe₃O₄ thin film samples²⁸⁾ prepared by cosputtering technique on Si(100) substrates. The study in Ref. 28) show that Cu²⁺ ions (x = 0 ÷ 0.35) substituted into tetrahedral A-site and octahedral B-site and formed mixed spinel $ (Cu_{y}^{2+} Fe_{1-y}^{3+}) [Cu_{\beta - y}^{2+} Fe_{1= \beta}^{2+}Fe_{1+y}^{3+}] O_{4}^{2-} $, where the ions $ Cu_{y}^{2 + } Fe_{1-y}^{3+} $ belong to the tetrahedral A-site and the ions $ Cu_{\beta - y}^{2+} Fe_{1-\beta}^{2+} Fe_{1+y}^{3+} $ belong to the tetrahedral B-site.

Recently, the composites of iron oxide-coated sand, diatomite or multiwall carbon nanotubes were studied to investigate the effect of factors to the arsenic removal^31–34). In particular, some coated polymers are studied to protect the chemical and physical properties of magnetic nanoparticles, simultaneously to ensure the recovery and increased reusing of the absorbent materials. As for example, the role of poly(1-naphthylamine) and poly(1-naphthylamine)/ZnO nanomaterials were investigated with decomposition of blue dye³⁵⁾ or the arsenic adsorption of poly(1-naphthylamine)/Fe3O4 nanomaterials³⁶⁾.

In this work, the synthesis, magnetic properties, microstructure of Fe₃O₄/PANI nanomaterials, the role of PANI, the effect of pH level for arsenic adsorption ability will be discussed more clearly and into details. In addition, the possibilities of the desorb and reabsorb, the parameters of adsorption process were also investigated using Langmuir isotherm equation^6,10,17).

FeCl₃.6H₂O; Na₂SO₃; 25% NH₃, NaOH, HCl, acetone 99%, isopropanol (IPA) and AsO₃ were prepared, in which the As(III) content of 106 ppb (10 times higher than allowed levels by the World Health Organization WHO). All chemicals are at analytical grade.

Labconco Freeze Concentrator (USA) was used to dry the materials in vacuum. The structure and morphology of the samples were investigated by XRD patterns (D5005, Bruker), TEM (JEOL5410), SEM (S4800), Ramanlog 9I (USA) and thermal gravimetric analysis (DTG-60H). The magnetization were measured by Vibrating Sample Magnetometer (VSM) and the mesopores structure of samples was observed by TriStar 3000 V6.07A with TriStar 3000 V6.08 software. The Atomic Absorption Spectrophotometer (AAS 6300 Shimadzu) was used to determine the arsenic content in the solutions that before and after was adsorbed using magnetic nanomaterials.

Firstly, the FeCl₃.6H₂O was mixed with a Na₂SO₃ solution, then (magnetically) stirred till it turned in yellow. Secondly, the NH₃ solution was added dripping slowly, until the pH reached 10. Kept stirring the solution for 30 minutes until it turned in black. The processes were chemically described by following reactions.   

Thirdly, these magnetic nanomaterials were separated from mixed solution using external magnetic field, and then filtered, washed with distilled water. Finally, these materials were desiccated at 50℃ in 48 hours and finely grinded to form Fe₃O₄ nanoparticles.

The Fe₃O₄/PANI nanomaterials were synthesized by polymerization method as following: i) a calculated amount of Fe₃O₄ was put in 100 mL distilled water, followed by 40 ml IPA, aniline (mixed solution A) and well stirred in 60 minutes. ii) a stoichiometric amount of (NH₄)₂S₂O₈ solution, with the monomer and the oxidizing agent molar ratio 1:1:5, was added, drop by drop into solution A. This dark blue mixture (solution B) was stirred in 2 hours with exothermal reaction. iii) The precipitates were separated from this mixed solution using external magnetic field, then they were dried by Labconco Freeze concentrator in 5 hours at 1 mPa and −40℃. The nanocomposites with various ratios of Fe₃O₄ and PANI were coded T₁, T₂, T₃, corresponding to 5%, 10% and 15% PANI as shown in Table 1.

Table 1 Fe3O4,T₁, T₂, T₃ samples.

In Fig. 1, the diffraction peaks of (220), (311), (400), (442), (511), (440) of Fe₃O₄, sample are fit completely with the standard diffraction pattern, and show the single phase of samples with the centered face cubic structure. The XRD patterns of T₂, T₃ samples appeared also the diffraction peaks completely the same as that of Fe₃O₄ sample, this prove that the addition of polymer did not affect the crystal structure of Fe₃O₄. From diffraction conditions 2dsinθ = nλ, the lattice constants of Fe₃O₄ nanoparticles in the Fe₃O₄, T₂ and T₃ samples are same with a = 8.376 A⁰ determined by the relation: $ {\rm a} = {\rm d}_{\rm hkl} \sqrt{h^{2} + k^{2} + l^{2}} $; where d_hkl is the distance of the lattice planes. Thus, the average crystal particle sizes of Fe₃O₄ in all Fe₃O₄, T₂ and T₃ samples is about of 14.5 nm that calculated by formula D = 0.9λ/βcosθ (λ = 1.5416 A⁰; β: full width at haft maximum of diffraction line). However, in Fig. 2, the TEM image shows the grain sizes of about 13–20 nm by the agglomeration of some Fe₃O₄ grains in sample. Due to PANI polymer is an amorphous material, its presence does not affect the crystal structure of Fe₃O₄, thus the particle size of about 14.5 nm for T₂ and T₃ samples is the size of Fe₃O₄ core. Meanwhile, in SEM image Fig. 3(a), the grain size of T₂ is larger than that calculated from XRD patterns. This result is appropriate, and represents the core-shell structures of the samples by the polymerization of PANI in surface of Fe₃O₄ grains. However, for T₃ sample in SEM image, Fig. 3(b), the fibers formed. This is explained due to more aniline content, the polymerization was concentrated and formed polymeric fibers from surface of Fe₃O₄ grains. The small peaks of 360 and 666 cm⁻¹ appeared in Raman spectrum of Fe₃O₄ (inset of Fig. 4). However, Raman spectra of PANI and T₂ (Fig. 4(a), (b)) show the 1592.52 cm⁻¹, 1477.45 cm⁻¹ and 1352.97 cm⁻¹ characteristic peaks of PANI in samples, in which the 1601.5 cm⁻¹, 1509.8 cm⁻¹ and 1336 cm⁻¹ peaks of T₂ are strong intensity. These peaks are adjacent to 1592.52 cm⁻¹–1477.45 cm⁻¹ and 1352.97 cm⁻¹ peaks of PANI nanotubules³⁷⁾. The 1601.5 cm⁻¹, 1509.8 cm⁻¹ peaks characterize by the oscillations of C=C and C=N groups, while 1336 cm⁻¹ peak is characterized by the oscillation of C-N groups. In addition, the 1159 cm⁻¹, 820.8 cm⁻¹, 608.66 cm⁻¹ peaks of T₂ sample are also very near to the shift in the range of 413.43 cm⁻¹–1176.28 cm⁻¹ of PANI³⁷⁾ that are corresponding to the oscillation frequency of the C-N⁺, C-N, C-H groups. The obtained analysis confirmed the presence of the PANI in samples (Fig. 4(a), (b)) and indicated the oxidation state of PANI. The results also demonstrate the formation of nanocomposite grains by polymerization of monomers outside of Fe₃O₄ nanoparticles.

Fig. 1

XRD patterns of Fe₃O₄ and T₂, T₃.

Fig. 2

TEM image of Fe₃O₄.

Fig. 3

SEM images of Fe₃O₄/PANI (a) T₂ and (b) T₃.

Fig. 4

Ramman spectra: Fe₃O₄, PANI, T₂.

The stability over time of magnetic, microstructure property of Fe₃O₄ was investigated by by comparing two magnetization results measured at different time. One is measured just after the sample being synthesized and another one is after two months of synthesis. Figure 5(a), (b) showed that the saturated magnetization of freshly synthesized Fe₃O₄ sample is about 63.13 emu/g, but after 2 months of synthesis, its magnetization of this sample was decreased to 56 emu/g. This phenomenon related the oxidation of Fe₃O₄ into γ-Fe₂O₃ due to oxygen in air with equation:   

Fig. 5

Magnetization and Raman spectra of Fe₃O₄. Magnetization: (a) sample after 2 months (b) new synthesized sample. Raman spectra (inset): (a) sample was measured after 2 month and (b) new synthesized sample.

Figure 6 shows the super-paramagnetic properties of Fe₃O₄, T₁, T₂ and T₃ with small Hc about 3 Oe. In Table 2, the saturated magnetization of Fe₃O₄ is about of 66.05 emu/g, due to the double exchange interaction through Fe²⁺ and Fe³⁺ ions at the (B) site, while the (A) site Fe³⁺ ions form antiferromagnetic interactions with (B) site Fe ions. However, the saturated magnetizations of Fe₃O₄/PANI samples were decreased to 39.74 emu/g by increased nonmagnetic PANI content from 5% to 15%.

Fig. 6

Magnetization of samples.

Table 2 Maximum absorption capability and saturate magnetic moment of Fe₃O₄, T₁, T₂, T₃.

TGA curve of PANI in Fig. 7(a) showed that at temperatures below 80℃, the mass 13% reduction of the sample due to the evaporation of water, corresponding to with the sharp endothermic peak at 45℃ in the DTG curve. From 80–210℃ the sample mass is almost unchanged. However, from 210–320℃ in the TGA curve, the reduction of sample mass happened due to the decomposition of PANI to forming the monomers, oligomers, dimer and trimers with the sharp endothermic peak at 292℃ in the DTG curve. Due to increase temperature, the thermal decomposition of the oligomer, dimer, trimer increased and completely decomposes of PANI at above 600℃.

Fig. 7

TGA curves of samples: PANI (a) and Fe₃O₄/15%PANI (b).

TGA curve of Fe₃O₄/PANI in Fig. 7(b) also showed that at temperatures below 80℃, the mass 17% reduction of the sample due to the evaporation of water corresponding to the sharp endothermic peak at 45.3℃ in the DTG curve. However, from 80–320℃ as observed in the TGA curve, the 6.2% reduction of sample mass may be happened due to the decomposition of the residual monomers and oligomers in sample. From 320–600℃, the thermal decomposition of the oligomer, dimer, trimers makes the sharp endothermic peak at 426.3℃ in the DTG curve. But, after that the mass of sample still remained about of 28% at 600℃.

The above analytic and TGA results showed that the polymer coating outside Fe₃O₄ nanoparticles is the optimal method to stabilize chemical properties and magnetization of Fe₃O₄.

The evaluation of arsenic adsorption ability of the samples was performed at room temperature. The initial As(III) content of 106 ppb was performed by adding 0.01 g of Fe₃O₄, T₁, T₂ and T₃. Then each mixture were stirred in 20 minutes for a complete adsorption. The arsenic contents before and after adsorption was determined by the Atomic Absorption Spectrophotometer.

In this work, the pH of the solution containing arsenic was prepared from 1 to 14. In Fig. 8 inset, one can see that the remaining arsenic content is a function of the pH. The arsenic adsorption was better in the range of pH 5–9. In both strong acidic and basic solution, the adsorption ability decreased. In the pH range from 7 to 9, the dependence of arsenic adsorption ability on the pH dominated by oxidation state of PANI (original υ_C-N⁺). That means the electrostatic affinity interactions appeared and contributed to As(III) anion adsorption. When the pH elevates, As(III) participation in H₂AsO₃⁻, HAsO₃²⁻anions, and the surface charge of the nanoparticles is negative. Thus they push the negative charges into solution and lead to the decrease of arsenic adsorption ability. At extreme high pH 14, due to strong electrostatic repulsive force, the material has no arsenic adsorption ability. When the pH is smaller than 5, the adsorption materials can be analyzed that also leads to a decrease of the adsorption ability. In this work, at low pH level (1–2), the nanocomposite was decomposed as iron ion. In neutral environment (pH = 7), the nanocomposite is stable and non-iron ion was detected in the mixed solution. Inset of Fig. 8 also shows that Fe₃O₄ is stable in strong alkaline meadium. It's suggested that the de-adsorption process should be conducted in solution at pH 14.

Fig. 8

Determination of balance time of arsenic absorption. Inset: Remained arsenic content depended on the pH.

As the above analysis, arsenic adsorption ability relates to the oxidation state of the nanomaterials in different pH environments. The adsorption is best in a neutral environment. However, in terms of dynamics, the adsorption process has the essence of the absolutely non-elastic collision phenomena between As and the nanoparticles. The adsorption kinetic of Fe₃O₄/polymer nanocoposites can be explained^30,36) by the relation of adsorption ability on the inelastic exchange interaction between specific surface area of nanoparticles and adsorption materials. The surface and mesopores structure of nanoparticles were studied by the nitrogen adsorption-desorption isotherms of 0.54 g for typical Fe₃O₄ sample at 77 K^30,36). By the BET (Brunauer, Emmett, and Taller) theory³⁸⁾, the pore size distribution was calculated at relative pressure P/P₀ ≈ 1, in which the pore size range is about 12–15 nm, corresponding to the maximum absorption pore volume.

Balance time of adsorption was determined by remaining arsenic content with different value of time (10 to 50, step 10 minutes) in solution pH 7. As shown in Fig. 8, the arsenic contents rest in balance state when the adsorption time is 20 minutes at room temperature. These observations were performed by the atom absorption spectrophotometer (AAS). The maximum arsenic adsorption ability q_max of unit volume of adsorbent (mg/g) is calculated by the Langmuir isotherm equation in pH = 7 and 300 K^6,10,17) with the first order linear relation:   

The calculated values of q_max for Fe₃O₄, T_1, T₂ and T₃ are presented in Table 2.

The results in Table 2 and Fig. 8 shows that the arsenic absorption ability of Fe₃O₄/PANI sample is best. With PANI content from 5 to 15%, the arsenic absorption ability of Fe₃O₄/PANI samples is better than Fe₃O₄, althought, their magnetization are smaller than Fe₃O₄.

The Fe₃O₄/PANI nanocomposite materials were successfully synthesized with different polymer content. PANI is difficult to dissolve in solvents and is stable in environments containing O₂ and H₂O, so the physical, chemical properties of Fe₃O₄/PANI are stable and lead to improve an arsenic adsorption ability in the aqueous solution. The pH 7 of aqueous medium and 20 minute adsorption time are suitable factors for effective adsorption process. This work suggests the desorb can be conducted at pH 14 then the materials could reabsorb in further trials.

This work was supported by the project of Hanoi National University of Education with code SPHN-15-421.

• 1)   S.  Prasanna,  S.  Mitsugu,  K.  Kenji and  Y.  Iwao: J. Magn. Magn. Mater. 247 (2002) 1–5.
• 2)   J.  Liu,  Y.  Bin and  M.  Matsuo: J. Phys. Chem. C 116 (2012) 134–143.
• 3)   J.F.  Bertone,  J.  Cizeron,  R.K.  Wahi,  J.K.  Bosworth and  V.L.  Colvin: Nano Lett. 3 (2003) 655–659.
• 4)   A.I.  Rusanov: Surf. Sci. Rep. 58 (2005) 111–239.
• 5)   H.  Gu,  Y.  Huang,  X.  Zhang,  Q.  Wang,  J.  Zhu,  L.  Shao,  N.  Haldolaarachchige,  D.P.  Young,  S.  Wei and  Z.  Guo: Polymer 53 (2012) 801–809.
• 6)   A.  Khodabakhshi,  M.M.  Amin and  M.  Mozaffari: Iran. J. Environ. Health Sci. Eng. 8 (2011) 189–200.
• 7)   M.  Auffan,  J.  Rose,  O.  Proux,  D.  Borschneck,  A.  Masion,  P.  Chaurand,  J.-L.  Hazemann,  C.  Chaneac,  J.-P.  Jolivet,  M.R.  Wiesner,  A.  Van Geen and  J.-Y.  Bottero: Langmuir 24 (2008) 3215–3222.
• 8)   A.I.  Zouboulis and  I.A.  Katsoyiannis: Environ. Int. 31 (2005) 213–219.
• 9)   G.R.  Chaudhary,  P.  Saharan,  A.  Kumar,  S.K.  Mehta,  S.  Mor and  A.  Umar: J. Nanosci. Nanotechnol. 13 (2013) 3240–3245.
• 10)   R.  Liu,  Y.  Lu,  X.  Shen,  X.  Yang,  X.  Cui and  Y.  Gao: J. Nanosci. Nanotechnol. 13 (2013) 2835–2841.
• 11)   X.B.  Fang,  Z.Q.  Fang,  P.K.E.  Tsang,  W.  Cheng,  X.M.  Yan and  L.C.  Zheng: Appl. Surf. Sci. 314 (2014) 655–662.
• 12)   T.  Hao,  C.  Yang,  J.  Wang,  C.  Niu,  X.  Su and  X.  Rao: Appl. Surf. Sci. 292 (2014) 174–180.
• 13)   G.  Yang,  L.  Tang,  X.  Lei,  G.  Zeng,  Y.  Cai,  X.  Wei,  Y.  Zhou,  S.  Li,  Y.  Fang and  Y.  Zhang: Appl. Surf. Sci. 292 (2014) 710–716.
• 14)   Q.  Chen,  Q.  He,  M.  Lv,  X.  Liu,  J.  Wang and  J.  Lv: Appl. Surf. Sci. 311 (2014) 230–238.
• 15)   Q.L.  Jiang,  S.W.  Zheng,  R.Y.  Hong,  S.M.  Deng,  L.  Guo,  R.L.  Hu,  B.  Gao,  M.  Huang,  L.F.  Cheng,  G.H.  Liu and  Y.Q.  Wang: Appl. Surf. Sci. 307 (2014) 224–233.
• 16)   M.J.  Chen,  H.  Shen,  X.  Li and  H.F.  Liu: Appl. Surf. Sci. 307 (2014) 306–310.
• 17)   C.M.  Babu,  B.  Palanisamy,  B.  Sundaravel,  M.  Palanichamy and  V.  Murugesan: J. Nanosci. Nanotechnol. 13 (2013) 2517–2527.
• 18)   Lei  Chen,  Hongchuan  Xin,  Yuan  Fang, Cong Zhang, Feng Zhang, Xing Cao, Chunhui Zhang, and Xuebing Li: J. Nanomaterials Vol. 2014, Article ID 793610, 10 pages.
• 19)   J.W.  Park,  A.N.  Jang,  J.H.  Song,  C.Y.  Park and  Y.S.  Lee: J. Nanosci. Nanotechnol. 13 (2013) 1895–1898.
• 20)   X.  Li,  F.  Zhang,  C.  Ma,  E.  Saul and  N.  He: J. Nanosci. Nanotechnol. 12 (2012) 2939–2942.
• 21)   B.  Zapotoczny,  M.R.  Dudek,  N.  Guskos,  JJ.  Kozioł,  BV.  Padlyak,  M.  Ko'smider and  E.  Rysiakiewicz-Pasek: J. Nanomaterials 2012, 1–7.
• 22)   L.  Méndez-Rodríguez,  T.  Zenteno-Savín,  B.  Acosta-Vargas,  J.  Wurl and  M.  Imaz-Lamadrid: Nat. Sci. 5 (2013) 238–243.
• 23)   K.-S.  Lin,  K.  Dehvari,  Y.-J.  Liu,  H.  Kuo and  P.-J.  Hsu: J. Nanosci. Nanotechnol. 13 (2013) 2675–2681.
• 24)   H.M.  Zaki,  S.  Al-Heniti,  A.  Umar,  F.  Al-Marzouki,  A.  Abdel-Daiem,  T.A.  Elmosalami,  H.A.  Dawoud,  F.S.  Al-Hazmi, and  S.S.  Ata-Allah: J. Nanosci. Nanotechnol. 13 (2013) 4056–4065.
• 25)   S.  Larumbe,  C.  Gómez-Polo,  J.I.  Pérez-Landazábal,  A.  García-Prieto,  J.  Alonso,  M.L.  Fdez-Gubieda,  D.  Cordero and  J.  Gómez: J. Nanosci. Nanotechnol. 12 (2012) 2652–2660.
• 26)   D.  Rathore,  R.  Kurchania and  R.K.  Pandey: J. Nanosci. Nanotechnol. 13 (2013) 812–1819.
• 27)   M.  Ishikawa,  H.  Tanaka and  T.  Kawai: Appl. Phys. Lett. 86 (2005) 222504.
• 28)   D.  Tripathy and  A.O.  Adeyeye: J. Appl. Phys. 103 (2008) 07F701
• 29)   F.  Genç,  E.  Turhan,  H.  Kavas,  U.  Topal,  A.  Baykal and  H.  Sözeri: J. Supercond. Nov. Magn. 28 (2015) 1047–1050.
• 30)   M.T.  Tran,  T.H.T.  Nguyen,  T.V.A.  Nguyen: Appl. Surf. Sci. 340 (2015) 166–172.
• 31)   V.  Uwamariya,  B.  Petrusevski,  Y.M.  Slokar,  C.  Aubry,  P.N.L.  Lens and  G.L.  Amy: Water Air Soil Pollut. 226 (2015) 184.
• 32)   H.  Fakour,  Y.F.  Pan and  T.F.  Lin: Water Air Soil Pollut. 226 (2015) 14.
• 33)   D.  Budimirović,  Z.S.  Veličković,  V.R.  Djokić,  M.  Milosavljević,  J.  Markovski,  S.  Lević and  A.D.  Marinković: Chem. Eng. Res. Des. 13 (2017) 75–86.
• 34)   K.D.  Hristovski and  J.  Markovski: Sci. Total Environ. 598 (2017) 258–271.
• 35)   S.  Ameen,  M.S.  Akhtar,  Y.S.  Kim,  O-.B.  Yang and  H-.S.  Shin: Colloid Polym. Sci. 288 (2010) 1633–1638.
• 36)   M.T.  Tran,  T.H.T.  Nguyen,  Q.T.  Vu and  M.V.  Nguyen: Front. Mater. Sci. 10 (2016) 56–65.
• 37)   M.  Tagowska,  B.  Pałys and  K.  Jackowska: Synth. Met. 142 (2004) 223–229.
• 38)   P.A.  Webb,  C.  Orr,  R.W.  Camp,  J.P.  Olivier,  Y.  Simon Yunes: Analytical methods in fine particle technology, (Micromeritics Instrument Corporation, Norcross, GA USA, 1997) pp. 60–62.