Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2020 Jun 4;5(23):14086–14095. doi: 10.1021/acsomega.0c01597

Formation of Fe(III)-phosphonate Coatings on Barium Hexaferrite Nanoplatelets for Porous Nanomagnets

Darja Lisjak †,*, Patricija Hribar Boštjančič †,‡, Alenka Mertelj †, Andraž Mavrič §,∥, Matjaz Valant §,∥, Janez Kovač †, Hermina Hudelja †,‡, Andraž Kocjan †, Darko Makovec †
PMCID: PMC7301540  PMID: 32566875

Abstract

graphic file with name ao0c01597_0006.jpg

Amorphous coatings formed with mono-, di-, and tetra-phosphonic acids on barium hexaferrite (BHF) nanoplatelets using various synthesis conditions. The coatings, synthesized in water with di- or tetra-phosphonic acids, were thicker than that could be expected from the ligand size and the surface coverage, as determined by thermogravimetric analysis. Here, we propose a mechanism for coating formation based on direct evidence of the surface dissolution/precipitation of the BHF nanoplatelets. The partial dissolution of the nanoplatelets was observed with atomic-resolution scanning transmission electron microscopy, and the released Fe(III) ions were detected with energy-dispersive X-ray spectroscopy and electron energy loss spectroscopy in amorphous coating. The strong chemical interaction between the surface Fe(III) ions with phosphonic ligands induces the dissolution of BHF nanoplatelets and the consequent precipitation of the Fe(III)-phosphonates that assemble into a porous coating. The so-obtained porous nanomagnets are highly responsive to a very weak magnetic field (in the order of Earth’s magnetic field) at room temperature, which is a major advantage over the classic mesoporous nanomaterials and metal–organo-phosphonic frameworks with only a weak magnetic response at a few kelvins. The combination of porosity with the intrinsic magneto-crystalline anisotropy of BHF can be exploited, for example, as sorbents for heavy metals from contaminated water.

Introduction

Barium hexaferrite (BHF; BaFe12O19) is traditionally used for permanent magnet and microwave technology applications.1 However, the development of a hydrothermal synthesis for its production in nanoform2,3 led to a new material, ferromagnetic liquids,4,5 and opened up a variety of new application possibilities6−9 in magneto-optics, magneto-rheology, spin memory devices, bioimaging, catalysis, and magneto-mechanical cancer treatment. These new applications arise from the intrinsically anisotropic magnetic properties of the BHF nanoplatelets (NPLs) that allow for their alignment at a very low magnetic field, i.e., below 1 mT. This is also a main advatange over the other magnetic nanoparticles that are typically with isotropic properties or superparamagnetic, i.e., without a permanent magnetic moment. Other applications, such as magnetically driven chemical purification, chemical reactions, or drug delivery with a simultaneous monitoring (e.g., through changes of refraction index),10 would be made possible by combining BHF NPLs with a porous phase. Here, we propose the formation of a porous phase by coating the BHF NPLs with phosphonic acids.

Only a few studies on the surface modifications of BHF NPLs, without any study of the porous coatings, have been reported.6,9,11 One option would be to use the knowledge of various highly porous coatings (e.g., mesoporous silica, dopamine, liposome) that have been successfully made on iron oxide superparamagnetic nanoparticles intended for magnetically targeted drug delivery.12,13 However, the same synthetic strategies cannot be simply translated to the BHF NPLs due to the challenging colloidal stability during processing.6 BHF crystallizes in the shape of thin hexagonal platelets with limited growth in the direction of the hexagonal c-axis. At the same time, the magnetization direction coincides with the c-axis and results in uniaxial magneto-crystalline anisotropy. A highly anisotropic crystal structure (a, b = 0.589 nm and c = 2.318 nm) of the BHF can be described as a combination of two structural blocks, S and R, along the hexagonal c-axis as RSR*S*.14 The cubic S block, (Fe6O8)2+, is a slice of the spinel lattice with its diagonal parallel to the c-axis, and the R block (BaFe6O11)2– is hexagonal, while the asterisk (*) denotes the rotation of the block for 180° around the c-axis. The Fe(III) ions occupy five different lattice sites. However, some structural deviations of the hydrothermally synthesized BHF NPLs from bulk have been reported recently.15 All of the BHF NPLs terminate with the S block, in particular with the surface Fe(III) ions at octahedral 12k positions. The BHF NPLs evolve with distinct thicknesses of 1.6, 3.0, 4.1, or 5.3 nm, depending on the synthesis temperature. In contrast, the width distribution is broad but can be narrowed by a partial Sc(III) substitution for Fe(III).3 The Sc-substituted BHF NPLs retain the uniaxial magneto-crystalline anisotropy, high sensitivity to magnetic field, and surface chemistry, and they can be electro(-sterically) stabilized in water and alcohols.11,16,17

Phosphonic acids are known for the strong chemical interaction with metal ions. The strength of their interaction with metal oxide surfaces originates from various attachment modes, from mono- to bridging bi- and tridentate, and chelating modes.18 Phosphonic acids with long alkyl chains form robust self-assembled monolayers (SAMs) on metal(oxide) surfaces that cannot form via the carboxyl or hydroxyl groups.19,20 The phosphonates with two or more phosphonic groups exhibit especially strong complexing efficiency for the metal ions.21−24 For example, ethylenediaminetetra(methylenephosphonic acid) (EDTMP; Scheme 1) is a very strong complexing agent. It is used as an anticorrosion and antiscaling agent (e.g., EDTMP·Na5, CUBLEN E series) and in cancer treatment (Sm135-lexidronam, i.e., Quadramet). In addition to this, EDTMP, when coated onto nanoparticles surfaces, also provides a high surface charge. Consequently, EDTMP ensures the colloidal stability of nanoparticles in aqueous solutions with a high ionic strength and even reduces the toxicity of nanoparticles.25,26 None of these studies elucidated the formation mechanism or the microstructure of the EDTMP coatings. In general, only a few of these studies proposed the possible formation of metal phosphonate or phosphate precipitates on oxide particles/surfaces, while most of the studied phosphonic acids were supposed to form dense coatings.18,27

Scheme 1. Selected Phosphonic Acids.

Scheme 1

Open in a new tab

Ethylenediaminetetra(methylenephosphonic acid) (EDTMP), octadecylphosphonic acid (OPA), (12-phosphono)dodecyl phosphonic acid (DPA), and (12-phosphono)dodecyl sulfonic acid (PSA).

The aim of this work was to explain the formation mechanism of porous phosphonate coatings on magnetic BHF NPLs. We compared four phosphonic acids (Scheme 1) with different numbers of phosphonic groups, molecular structure, and polarity. EDTMP is a highly polar molecule with four phosphonic groups and has a relatively high solubility in water (second highest among the selected phosphonic acids). Octadecylphosphonic acid (OPA) has only one phosphonic group and a long nonpolar alkyl chain, and is not soluble in water. The chemical properties of (12-phosphono)dodecyl phosphonic acid (DPA) and (12-phosphono)dodecyl sulfonic acid (PSA) are in between those of OPA and EDTMP. DPA with two phosphonic groups is only slightly soluble in water, while PSA with one phosphonic and one sulfonic group is highly soluble in water. Finally, we present preliminary results on a possible application of the new porous nanomagnets for the extraction of heavy metals from water.

Experimental Section

Materials and Methods

Barium(II) nitrate (99.95%), iron(III) nitrate nonahydrate (>98%), scandium(III) nitrate hydrate (99.9%), bismuth(III) nitrate pentahydrate (98%), NaOH (98%), and 1-hexanol were obtained from Alfa Aesar. Octadecylphosphonic acid (OPA), (12-phosphono)dodecyl phosphonic acid (DPA), and HNO3 were purchased from Sigma-Aldrich. N,N,N,N-ethylenediaminetetra(methylenephosphonic acid) hydrate (EDTMP hydrate) was obtained from abcr GmbH, (12-phosphono)dodecyl sulfonic acid (PSA) from Sikemia, and toluene was obtained from VWR Chemicals. All chemicals were used as received.

Sc-substituted BHF nanoplatelets (BHF NPLs) were synthesized hydrothermally at 240 °C, as described previously,3 starting with a nominal molar ratio of metal nitrates (Fe + Sc)/Ba = 5 and Fe/Sc = 9. Stable aqueous suspensions of the BHF NPLs were obtained at pH = 2–2.2, which was adjusted with a 14.3 M solution of HNO3. The chemical composition of the NPLs was determined in average to be Ba1±0.08Fe14.9±0.12Sc0.87±0.10Ox, and their thickness was found to be ca. 3–5 nm.16 The width of the as-synthesized BHF NPLs was determined (see Materials Characterization) from transmission electron microscopy (TEM) images (as in Figure 1a). For the estimation of surface ligand fractions and coating densities (for details, see the Supporting Information), we considered an average width of 50 or 60 nm (Table 1) and a thickness of 4 nm.

Figure 1.

Figure 1

Open in a new tab

TEM images of the as-synthesized BHF NPLs with clean surfaces (a) and BHF–EDTMP NPLs (nominal fraction 8 EDTMP/nm2) (b). The enlargement in (b) shows the amorphous surface layer.

Table 1. Nominal and Bonded Ligand Fractions with Room-Temperature Magnetic Properties of the As-Synthesized and Coated BHF NPLs.

nominal ligand fraction
 bonded ligand fraction
    
(molecules/nm2) (%) (%) (molecules/nm2) Ms (A m2/kg) Hc (kA/m)
as-synthesizeda 48 ± 21 nm 0 0 0 38 85
as-synthesizedb 60 ± 25 nm 0 0 0 37 99
5 EDTMPb 28.4 14.0 2 27 126
8 EDTMPb 40.0 17.0 3 26 74
20 EDTMPb 61.3 18.1 3 26 134
5 OPAa 21.0 29.5 7 29 108
17 OPAa 47.6 35.9 9 24 98
48 OPAa 72.4 37.9 10 23 101
8 DPAb 34.0 29.6 7 23 118
20 PSAb 45.9 28.6 7 27 106
Open in a new tab
a

Denotes samples synthesized from the first batch of as-synthesized NPLs.

b

Denotes samples synthesized from the second batch of as-synthesized NPLs.

BHF–EDTMP NPLs

The effect of processing parameters on the formation of the EDTMP coating was studied (see the Supporting Information for details), but most of the samples were synthesized at 80 °C with a holding time of 3 h. The pH was set to 1–2 for EDTMP to be fully protonated.22,23 The coating procedure was as follows: the EDTMP was dissolved in water at the synthesis temperature, and the pH was adjusted to the selected value with 1 M HNO3. Then, the aqueous suspension of BHF NPLs was added. The concentration of the BHF NPLs in the reaction suspension was 0.1 mg/mL, while the EDTMP concentration varied with respect to the desired surface coverage (i.e., nominal fraction) of the BHF NPLs. During the whole process, the reaction slurry was mixed with a glass stirrer. The reaction was stopped after 3 h, and the reaction vessel was left to cool to room temperature for 1 h. The coated BHF NPLs were collected by a centrifuge (3047 rcf for 5 or 10 min) and washed with water. The washing procedure was repeated five times. The reaction product was a supernatant obtained after the last centrifugation (5 min). No (or only minor) sedimentation of the coated BFH NPLs was observed after a week or more. The NPLs were resuspended with an ultrasonic probe (Sonics Vibra-cell) at a power of 150 W using 1 s long on/off pulses with a total sonication time of 1 min.

BHF–DPA NPLs were synthesized at the same conditions as those of EDTMP–BHF NPLs, at 80 °C for 3 h and pH ∼ 1–2. At this pH, DPA should be fully protonated or with only one deprotonated −OH group in each of the two phosphonic groups.28 The samples were washed and sonicated as the BHF–EDTMP NPLs. However, the obtained suspensions sedimented completely overnight at neutral pH, while the sedimentation was partial at pH ≥ 10.

BHF–PSA NPLs were synthesized similarly as above, at 80 °C for 2 h and pH of 2.9. After washing, the BHF–PSA NPLs were ultrasonicated (as above) and stable suspensions were obtained in water.

BHF–OPA NPLs were synthesized by the functionalization at a phase boundary. OPA was dissolved in a mixture of toluene and 1-hexanol, with a volume ratio of 4:1. After that, the aqueous suspension of the BHF NPLs was added. The OPA solution in organic solvents and the suspension of the BHF NPLs in water were separated with a visible phase boundary: brownish aqueous suspension and transparent OPA solution. The reaction mixture was mixed with a glass stirrer for 2 h at room temperature. The transition of the BHF NPLs from aqueous to organic phase was observed visually from the brown coloring of the organic phase. The two phases were separated in a separating funnel. The BHF–OPA NPLs were collected by centrifugation at 3047 rcf for 5 min and washed five times with a mixture of toluene and 1-hexanol (volume ratio, 4:1) and four times with toluene. The collected BHF–OPA NPLs were immersed in toluene and ultrasonicated as above.

If there is need for the characterization, the samples were dried at 80 °C for a few hours in air or freeze-dried (Freeze Dryer, Kambič, LIO-5P) in vacuum for 2 days.

Materials Characterization

The morphologies of the as-synthesized and coated BHF NPLs were analyzed with a TEM (Jeol 2100) operated at 200 kV. A few drops of the suspension were deposited onto the TEM Cu grid support and left to dry. The width distribution of NPLs was determined from their surface area with DigitalMicrograph Gatan Inc. software and expressed as equivalent diameters. The BHF crystal structure was confirmed with electron diffraction. The chemical composition was inspected with energy-dispersive X-ray spectroscopy (EDXS). A more detailed investigation of the NPLs was performed with a CS-probe-corrected scanning transmission electron microscope (STEM, Jeol ARM 200 CF) operated at 80 kV. During the analysis, high-annular dark-field (HAADF) and annular bright-field (ABF) detectors were used simultaneously at 68–180 and 10–16 mrad collection semiangles, respectively. The chemical composition was analyzed using a Jeol Centurio EDXS system with a 100 mm2 silicon drift detector (SDD) and a Gatan GIF quantum ER dual electron energy loss spectroscopy (EELS) system.

The surface modification with phosphonic acids was analyzed by measuring the zeta-potentials of the diluted aqueous suspensions (0.1 mg/mL) and their dependence on pH with a ZetaPALS Zeta-potential analyzer (Brookhaven Instruments). pH was adjusted with 0.1 or 1 M solutions of HCl or NaOH. The measuring procedure was the same as that for all samples, allowing us to make a direct comparison between them.

The dried as-synthesized and coated NPLs and pure phosphonic acids were analyzed by Fourier transform infrared (FTIR) spectroscopy using a PerkinElmer Spectrum 400 FT-IR/FT-FIR spectrometer. The spectra of the dried samples were taken with a Universal attenuated total reflectance (ATR) sampling accessory in the range between 4000 and 650 cm–1. Additional analyses were performed with X-ray photoelectron spectroscopy (XPS) to analyze the surface chemistry using a PHI TFA XPS spectrometer (Physical Electronics). The instrument was equipped with an Al monochromatic X-ray source and a hemispherical electron energy analyzer. The analysis depth in the XPS spectroscopy was 3–5 nm.

Comprehensive thermal analyses (TA) of the ligands and dried, as-synthesized, and coated NPLs were performed simultaneously with a thermogravimetry and differential scanning calorimetry (TGA/DSC 2, Mettler Toledo) thermal analyzer coupled with a mass spectrometer (MS, Thermostar300, Vacuum Pfeifer) for the analysis of the gas evolved. The samples were heated from 40 to 1100 °C at 20 °C/min in a static air atmosphere. The reported curves are buoyancy-corrected. The fractions of bonded ligands were determined by TGA, as described in the Supporting Information.

The specific surface area of the freeze-dried NPLs was determined with a nitrogen sorption analyzer (Quantachrome Instruments NOVA 2200E) operating at liquid nitrogen temperature. For the calculation, the Brunauer–Emmett–Teller (BET) equation, based on the nitrogen adsorption data in the P/P0 range between 0.05 and 0.3 (seven-point analysis), was used. The freeze-dried samples were degassed at 80 °C in vacuum overnight prior to the measurement.

The magnetic properties of the (freeze-)dried as-synthesized and coated BHF NPLs were measured with a vibrating-sample magnetometer (VSM, LakeShore 7404). The samples with known mass were sealed into a plastic sample container. To verify the coercivity (Hc) values, the dried NPLs (mass fraction, 3–11%) were mixed with sugar to minimize the interparticle interactions and compressed in a cube with a force of 3 kN to prevent the rotation of the NPLs. The magnetic properties of the cubes were measured in all three directions and averaged to eliminate the effect of any preferential alignment of the NPLs.

Water Purification

A preliminary test was done with BHF–EDTMP NPLs. We tested the adsorption of Bi(III) from the aqueous solution of bismuth nitrate. The stable suspension of BHF–EDTMP NPLs (0.2 mL, 1.5 mg/mL, nominal fraction 8 EDTMP/nm2) was diluted with 1.5 mL of the Bi solution (0.1 mg/mL). The final pH of the mixture was ∼5. The mixture was shaken overnight (∼16 h) and sedimented on a permanent magnet in an hour. The solution was removed from the sediment and acidified with conc. HNO3 to pH ≤ 2 to determine the concentration of the nonadsorbed Bi with an inductively coupled plasma-optical emission spectrometer (ICP-OES; Agilent 720).

Results and Discussion

EDTMP Coating

Amorphous EDTMP coatings (few nanometers thick) formed at the surfaces of BHF NPLs (Figure 1b) regardless of the synthesis parameters (see Materials and Methods and the Supporting Information). However, due to very long dissolution times (1–2 h) of EDTMP at room temperature, most of the coatings were synthesized at 80 °C from acidic solutions. Unless otherwise stated, we will here refer to the samples prepared under such synthesis conditions. The stability of metal complexes, as well as the coordination number of metal ions with EDTMP, was found to be higher at lower pH values due to lower or no deprotonation of EDTMP.22,23 Similar to this, the adsorption of phosphates and phosphonates on goethite (i.e., iron oxyhydroxide) surfaces was promoted at acidic pH.29 The amorphous coatings were observed with TEM for all of the nominal EDTMP fractions, even for values as low as 0.5 EDTMP/nm2, for which ∼1 nm thick coating was observed (Figure S1a, Supporting Information). For larger nominal fractions, 5–84 EDTMP/nm2, the coatings had similar thicknesses of 2–3 nm. This is consistent with the similar fractions of bonded EDTMP (ca. 2–3 EDTMP/nm2; Table 1). Due to the small molecular size, it was assumed that EDTMP formed multiple layers and/or a porous framework on the surfaces of BHF NPLs.

Indeed, the specific surface area of the coated NPLs increased in comparison to that of the as-synthesized NPLs (90 m2/g). For example, the BET specific surface area of the BHF–EDTMP NPLs with three bonded EDTMP/nm2 was for different batches 135–160 m2/g. However, a higher specific surface area can be assumed for the coating itself. Namely, some fraction of the NPLs may agglomerate via basal planes into thicker stacks, thus resulting in the apparently lower surface area. Note that the specific surface area is inversely proportional to the thickness of NPLs. In addition to this, the core BHF NPL represents more than 80% of the BHF–EDTMP NPL’s mass (see Table 1), but it does not contribute to the surface area. A rough estimation of the EDTMP coating surface area (eq S3, Supporting Information) gives us a value of 600–700 m2/g that is comparable to some known metal–organic frameworks (MOFs)30 intended for cleaning polluted air and water.

The EDTMP coating significantly affected the surface charge of NPLs (Figure 2a). Due to the increased density of acidic surface groups, provided by the EDTMP coating,24−26 the pH of isoelectric point (IEP) decreased from ∼7 to ≤3 for the as-synthesized and EDTMP-coated BHF NPLs, respectively. Such an increase of the negative surface charge was only possible if some of the EDTMP phosphonic groups remained unbonded at the BHF NPL surfaces. The pH of IEP and the zeta-potential value abruptly decreased for up to 5 EDTMP/nm2, whereas the decrease was much smaller above 5 EDTMP/nm2. This correlates well with the fraction of bonded EDTMP, determined experimentally (Table 1).

Figure 2.

Figure 2

Open in a new tab

Comparison between the as-synthesized and EDTMP-coated BHF NPLs: zeta potential vs pH (a), ATR-FTIR spectra, including the spectrum of EDTMP (b), and room-temperature magnetic hysteresis loops (c). The BHF–EDTMP NPLs are distinguished by nominal ligand fractions, as noted in the legends. The lines in (a) are applied only for a clearer presentation.

The ATR-FTIR spectra of EDTMP, as-synthesized, and EDTMP-coated NPLs are shown in Figure 2b. There is a clear difference between them. The as-synthesized BHF NPLs do not show many distinct bands since typical bands for Fe–O vibrations appear at ∼590 cm–1 24 that is below the measuring range of our ATR attachment (650 cm–1). The main very broad band at 3000–3500 cm–1 and a weak band at 1621 cm–1 can be attributed to the physisorbed and/or hydrogen-bonded water and a weak band at 1520 cm–1 to carbonate.21,31 The carbonate band is not present in the ATR-FTIR spectrum of BHF–EDTMP NPLs, suggesting the replacement of carbonate with EDTMP, while the bands at 3000–3500 and 1649 cm–1 indicate physisorbed water or/and hydrogen-bonded phosphonic groups. When coated on BHF NPLs, the characteristic phosphonic bands of EDTMP21,25,26,32−35 (Table S1, Supporting Information) merged into a broad band centering around 1050 cm–1, providing clear evidence of some structural modification coming from the interaction with BHF NPL.21,24,31 The broad phosphonate band overlaps with the bands previously36 assigned to P–O–Fe at 975 and 1047 cm–1.

The wide-energy-range XPS spectra of the as-synthesized BHF NPLs and EDTMP-coated NPLs are shown in Figure S2 (Supporting Information). In these spectra, the peaks of O 1s, C 1s, Fe 2p, Fe 3p, Ba 3d, Sc 2p, P 2p, P 2s, and N 1s were identified. Although the depth of NPLs obtained from the XPS analysis (i.e., 3–5 nm) coincides with their thickness, some difference in the surface composition between the bare and the EDTMP-coated NPLs was detected. The surface composition of BHF NPLs was 47.9 atom % O, 39.9 atom % C, 9 atom % Fe, 1.6 atom % Ba, and 1.7 atom % Sc, and the surface composition of BHF@EDTMP NPLs was 36.6 atom % O, 50.9 atom % C, 2.7 atom % Fe, 0.5 atom % Ba, 4.8 atom % P, and 4.6 atom % N. The presence of P and N in the BHF@EDTMP NPLs confirms the EDTMP coating. In addition, the high-energy resolution spectra P 2p (not shown here) with binding energy 132.9 eV confirms the presence of phosponates,37 thus confirming the results described above.

The magnetization of the as-synthesized and coated BHF NPLs was measured with a magnetic field up to 796 kA/m (1 T). The magnetization increased only slightly above 590 kA/m. Therefore, the maximum magnetization at 796 kA/m was considered as the room-temperature saturation magnetization (Ms). As expected, the coating of BHF NPLs with EDTMP decreased the Ms values of the samples (Table 1) due to the dilution of the magnetic BHF NPLs with nonmagnetic coating. However, if the observed decrease is attributed solely to that dilution, the fraction of bonded EDTMP should be 28–32% (for the nominal 5–20 EDTMP/nm2), which is significantly higher than the values obtained from TGA, i.e., 14–18% (Table 1). This is another indication of some structural changes of the NPLs. The coating must have affected the magnetic exchange interaction between the Fe(III) ions (at the surface and surface interior), thus lowering the Ms value of BHF NPLs.38,39 Nevertheless, the BHF–EDTMP NPLs retain the high sensitivity to a very low magnetic field (Figure 2c). The coating seems to have some effect on the Hc values of the BHF NPLs, but this effect is relatively small in comparison to the Hc values of bulk Ba hexaferrite, ca. 255–530 kA/m.1 In general, the coated NPLs show higher Hc values than the as-synthesized NPLs. This may be a consequence of the elimination of all direct magnetic interparticle interactions by the nonmagnetic coating.40 In addition, Hc is an extrinsic property and is also affected by, e.g., size distribution that can change during the processing (e.g., during the multiple centrifugation/washing steps).

Detailed thermal analyses were performed to analyze the thermal stability of coatings (Figure 3) and to estimate the fraction of bonded ligands (Table 1). We were able to distinguish between three low-temperature processes by simultaneously using TGA and DSC coupled with MS (Table S2, Supporting Information), which are the release of crystal and/or physisorbed water (185 °C), melting of EDTMP, and condensation of the melted molecules (210 °C). The subsequent decomposition of EDTMP begins at ∼330 °C and continues up to ∼650 °C. Some additional mass loss is associated with the combustion of carbon black. EDTMP does not decompose completely (only 72%) up to 1100 °C, and no additional mass loss is observed above ∼1000 °C. The solid residue is supposed to be a mixture of oxides and phosphates.33

Figure 3.

Figure 3

Open in a new tab

Thermal decomposition of pure EDTMP and the as-synthesized, and EDTMP-coated BHF NPLs: DSC (top), TGA (middle), and MS (bottom).

No significant mass loss was observed with TGA, as well as no distinct process with DSC for the as-synthesized BHF NPLs (Figure 3). The minor mass loss can be attributed to the desorption of physisorbed water. Similar TGA and DSC curves were observed for the BHF–EDTMP NPLs but with a higher mass loss due to the decomposition of the EDTMP coating. The distinct difference between the TGA and DSC curves of the pure EDTMP and the EDTMP coated onto BHF NPLs indicates that the bonded EDTMP does not undergo either typical melting or condensation process. This suggests some structural differences that change the thermal decomposition behavior of bonded EDTMP. The significantly lower fraction of bonded EDTMP than the nominal fraction (Table 1) indicates that only a minor part of the added EDTMP was bonded to the NPL surfaces, i.e., with a maximum density of ∼3 EDTMP/nm2. Moreover, the fraction of the bonded EDTMP increased only slightly with the increase in the nominal fraction, which is in accordance with the similar coating thickness and zeta-potential behavior (Figure 2a) for the BHF–EDTMP NPLs with nominal ligand fractions between 5 and 84 EDTMP/nm2.

Other Coatings—A Comparison with EDTMP

Next, we compare how different types of phosphonic acids (Scheme 1) affected the properties of the coated BHF NPLs. The most significant difference was expected between the EDTMP and OPA. Indeed, the thicknesses of the OPA coatings were ∼1 nm regardless of the ligand fractions (Figure S1b,c, Supporting Information). The nominal fraction of 5 OPA/nm2 was not enough to coat all of the NPLs since the fraction of bonded OPA (7 OPA/nm2) exceeds the nominal one (Table 1). The uncoated BHF NPLs must have remained in the aqueous phase and were omitted from our analyses (see Materials and Methods). In contrast, the nominal fractions of ≥17 OPA/nm2 were excessive since the bonded fractions were lower than the nominal. The increasing fraction of bonded OPA must have only resulted in an increased density of the coating since it did not affect the coating thickness. Based on the maximum possible surface density of 4.2 OPA/nm2,42,41 OPA (bonded fraction, 7–10 OPA/nm2) seems to form a bilayer on the BHF NPLs. Similar to this, the bilayer could constitute a 1–2 nm thick PSA coating (Figure S1e, Supporting Information) with seven bonded PSA/nm2. The DPA coating (Figure S1d, Supporting Information) was a few nanometers thick, thicker than the OPA and PSA coatings, despite the smaller/comparable molecular size and comparable bonded ligand fraction.

The estimated density (eq S1, Supporting Information) of the 4 nm thick DPA and 2 nm thick EDTMP coatings was ∼0.7 g/cm2. For comparison, the densities of the OPA (thickness, 1 nm) and PSA (thickness, 1.5 nm) coatings with a similar fraction of bonded ligands (∼30%) were ∼4 and 2 g/cm3, respectively. We can assume that multiple DPA layers formed a porous framework, similar to EDTMP. The BET specific surface area of the BHF–DPA NPLs was 60 m2/g, meaning that the surface area of the DPA coating was ∼400 m2/g (eq S3, Supporting Information), which is lower that of the BHF–EDTMP NPLs (600–700 m2/g). The difference can be attributed to the different molecular structures of EDTMP and DPA (Scheme 1). However, a more systematic study is needed for a complete analysis of the surface properties of these porous nanomaterials.

The coating of the BHF NPLs with OPA made them hydrophobic and prone to flocculation in water. On the other hand, DPA made BHF NPLs more acidic (Figure S3, Supporting Information). The zeta-potential decreased and the IEP occurred at a lower pH (3–3.5) than that for the as-synthesized BHF NPLs (pH ∼ 7, Figure 2a), similar to that for EDTMP. The suspensions of the BHF–DPA NPLs were stable in water only at a high pH. There was no significant difference in the zeta-potential behavior between the different nominal surface fractions of the DPA in the range of 5–100 DPA/nm2. On the contrary, the coating with PSA resulted in a highly negative zeta-potential at the measured pH range of 2–11 suggesting a high density of acidic surface groups, higher than those for the DPA and EDTMP coatings. The different surface properties of the PSA coating can be explained with a selective bonding of phosphonic groups with the surface Fe(III) ions due to the higher interaction energy of the metal ions with phosphonic than with sulfonic groups.43 Consequently, (almost) all of the sulfonic groups remained free and provided for the high negative surface charge and excellent colloidal stability in water. This is also in accordance with the very low pKa values (<1) for sulfonic acids in comparison to pKa1 ∼ 2 for phosphonic acids.22,23,28

The Ms values of the BHF–OPA NPLs decreased in line with the increasing fraction of bonded OPA (Table 1 and Figure S4a). The most significant decrease was observed between the as-synthesized BHF and the BHF–OPA NPLs with a nominal fraction of 5 OPA/nm2. This coincides with the fact that for larger nominal fractions of OPA, only part of the added OPA molecules bonded to the BHF NPLs. The correlation between the decrease in Ms and a fraction of the bonded OPA seems in contradiction with the observed negative effect of the monophosphonate coating on the Ms values of the magnetite nanoparticles.39 However, for the magnetite (FeO·Fe2O3) nanoparticles, the decrease in Ms was attributed to an oxidized surface layer as a consequence of the coating with the monophosphonate. Such oxidation is not possible for the BHF NPLs since Ba hexaferrite is already a fully oxidized compound. The nano/surface effects (spin canting, surface anisotropy, shape)40 are already significant for the uncoated BHF NPLs, and the potential additional contribution due to the OPA coating is not significant.

The Ms value also decreased after the coating of BHF NPLs with DPA and PSA (Table 1 and Figure S4b). While the decrease of the Ms values of the BHF–OPA and BHF–PSA NPLs was in accordance with the mass fraction of bonded ligands, it was not so for the BHF–DPA NPLs. The fraction of bonded DPA suggested from the Ms value (38%) was higher (similar to the BHF–EDTMP NPLs) than that determined from TGA (29.6%) and even higher than the nominal fraction (34.0%). We can conclude that the binding of DPA and EDTMP onto the surfaces of BHF NPLs significantly affected the (surface) structure of BHF NPLs, but this was not so for the OPA and PSA. The Hc values of all of the coated BHF NPLs were slightly higher than those of the as-synthesized NPLs, similar to BHF–EDTMP NPLs.

The thermal stability of the coatings was analyzed with TGA, DSC, and MS (Figure S5, Supporting Information). If we compare all four ligands from this study, the total mass loss was found to be the largest for OPA and smallest for EDTMP, which is in line with an increasing length of the alkyl segments.44 Exactly the opposite was true for the low-temperature mass loss at ∼200 °C, associated with the condensation of phosphonic groups and the mass fraction of solid residue that both increased in line with the increasing number of phosphonic groups, from OPA and DPA to EDTMP. As with EDTMP (Figure 3), differences were observed between the pure phosphonic acids and the respective coatings, suggesting the structural changes of bonded phosphonic acids: no melting (as in pure OPA and DPA), no condensation (as in pure DPA), and a two-step decomposition of the coatings as opposed to a single-step decomposition in OPA, DPA, and PSA. The two-step decomposition of the OPA-coated Fe oxide nanoparticles was observed previously, but the steps were identified at lower temperatures, 340 and 479 °C.19 Consequently, different binding energies of the OPA with the nanoparticles’ surfaces were assumed based on two different binding modes: a monodentate with lower binding energy and a bidentate with higher binding energy. However, our TGA and MS results clearly show that pure OPA starts to decompose at ∼420 °C (Figure S5, Supporting Information).

The chemisorption of OPA, DPA, and PSA onto the BHF NPLs surface can be assumed from their ATR-FTIR spectra (Figure S6, Supporting Information). Similarly to the EDTMP coating, distinct phosphonic bands between 740 and 1260 cm–1 merged into one broad band at ∼1000 cm–1. Another difference between the pure and coated ligands is an additional broad band above 3000 cm–1 associated with the OH groups from physisorbed and/or hydrogen-bonded water.31,33 In addition, characteristic bands at ∼2850 and 2900 cm–1 can be associated with alkyl chains and a band at ∼1066 cm–1 in the PSA spectrum can be associated with the SO32– group that is slightly shifted for BHF–PSA NPLs to 1036 cm–1, most likely a consequence of hydrogen bonding.19,34

STEM Investigation

A detailed atomic-resolution STEM investigation revealed a distinct difference between the EDTMP and OPA coatings on BHF NPLs. While no structural changes were observed for the BHF–OPA NPLs (Figure S7, Supporting Information), the coating with EDTMP showed signs of a partial dissolution on the surfaces of all inspected BHF–EDTMP NPLs. In contrast to this, the uncoated BHF NPLs exhibited a uniform crystalline structure with the basal surfaces always terminated with a complete Fe-only cubic S-structural block of the hexaferrite structure (Figure 4a)15,16 Therefore, any significant change in the structure at the surfaces of NPLs can easily be observed with STEM. As an example, Figure 4b,c illustrates the effect of dissolution on the structure of the BHF–EDTMP NPL. It is evident that the complete columns of the Fe(III) ions are missing at the surfaces. In fact, the entire surface S block is absent on the left-hand side of the NPL, as presented in Figure 4a,b. The dissolved Fe(III) ions from the NPL were detected in the amorphous coating with EDXS and EELS analyses (Figures S8 and S9, Supporting Information, respectively).

Figure 4.

Figure 4

Open in a new tab

HAADF-STEM image of the as-synthesized BHF NPL (a), and HAADF (b) and BF (c) STEM images of the EDTMP-coated BHF NPL. The projected structural model is superimposed over the images (a) and (b) to illustrate the positions of the Ba(II) (green) and Fe(III) (brown) ions across the complete NPL structure.

It has already been suggested that a surface reaction of phosphonic/phosphate ligands with oxide surfaces competes with a surface dissolution–precipitation process and subsequently promotes the formation of multilayers.27,38,41,45,46 The only evidence for the surface dissolution–precipitation of Al(III) phosphonates from the previous studies is based on the NMR analyses, while the surface precipitation of Fe(III) phosphate after the functionalization of magnetic Fe oxide nanoparticles with bisphosphonates was only assumed from their magnetic behavior.27,38,41 According to our knowledge, the results from our study (Figure 4, and Figures S8 and S9, Supporting Information) are the first direct evidence of the surface dissolution of oxide nanoparticles as a consequence of their interaction with phosphonic ligands.

Mechanism of the Coating Formation

First, we discuss the dissolution of BHF NPLs with EDTMP (Figure 4) that was completely unexpected since barium hexaferrites are considered to be insoluble in water. No such dissolution has ever been observed for the as-synthesized, citrate, or sulfonate-coated BHF NPLs in aqueous suspensions that were in use for months. BHF NPLs can be dissolved in boiling concentrated HF, followed by an additional treatment in concentrated HCl or HNO3. The most probable explanation for the observed dissolution of the BHF–EDTMP NPLs is the interaction of the surface Fe(III) ions with the EDTMP. The dissolution is possible when metal ions are hydrated and can be enhanced when a stable solid product is formed from the dissolved ions. Consequently, the formation of a stable Fe(III)–EDTMP complex promotes the surface dissolution of BHF NPLs as in

graphic file with name ao0c01597_m001.jpg 1

The dissolution–precipitation of metal phosphonates/phosphates can be controlled with reaction conditions (temperature, pH, solvent, oxide nature). For example, the dissolution–precipitation of phosphonates on alumina surfaces is high at pH = 4 and low at pH = 4.5–8.5.27 This coincides with the higher and poorer solubility of alumina in water, respectively. The dissolution–precipitation can be prevented or significantly decreased using phosphonic ligands insoluble in water,18,27 as the OPA in our case.

The partial dissolution of BHF NPLs explains the negative effect on the room-temperature Ms of the BHF–EDTMP NPLs (Table 1 and Figure 2c), which cannot be attributed solely to the dilution of the BHF NPLs with a nonmagnetic EDTMP. The observed dissolution of the surface as well as some subsurface Fe(III) ions (Figure 4) represents a significant change of the overall structure of the NPLs. A similar negative effect of the DPA coating on the Ms indicates that DPA can also induce some dissolution of BHF NPLs. On the other hand, the decrease in the Ms value was in accordance with the fraction of the bonded OPA, for which no dissolution was observed (Figure S7, Supporting Information). This could be expected since the hydrophobic OPA coating was not formed in water, as in the hydrophilic EDTMP and DPA coatings. As with OPA, the decrease in the Ms value of BHF–PSA NPLs correlates with the fraction of bonded PSA and does not suggest any significant modifications of the BHF structure despite the high solubility of PSA in water. PSA and OPA are both monophosphonic acids with lower interaction energies toward the surface Fe(III) ions than EDTMP with four phosphonic groups.21−23

The different formation mechanism of the studied coating significantly affects their microstructures. SAMs readily form from long alkyl-phosphonic acids on oxide surfaces.19 Direct evidence for the formation of 1.8 nm thick SAMs made of OPA on mica surfaces was provided with atomic force microscopy.20 The OPA coating in our samples forms a dense bilayer, as suggested from the high fraction of bonded 7–10 OPA/nm2 (Table 1). The OPA molecules from the second layer were adsorbed onto the first, chemisorbed, layer via hydrogen bonds, as suggested by the broad IR band above 3000 cm–1 (Figure S6a, Supporting Information). Moreover, an ordered assembly of the OPA molecules on the BHF NPLs can be assumed from the low wavenumbers (2912 cm–1) for the C–H stretching.41 Similar to this, PSA must have also formed a dense self-assembled bilayer as indicated by the fraction of bonded ligand (7 PSA/nm2) and typical IR bands (Figure S6c, Supporting Information) for hydrogen bonds and ordered aliphatic chains. It has to be noted that the bilayer does not necessarily form in the solvent where there could be an equilibrium between the bonded and dissolved ligands. The dissolved ligands can adsorb and assemble on the coated BHF NPLs during the drying process.

As opposed to OPA and PSA, much smaller EDTMP molecules form thicker coatings on the BHF NPLs at lower surface fractions (Table 1, Figures 1b and S1a). This can be explained, as previously,42,46 by the formation of multilayered coatings. Namely, a kind of multilayered coating can also be formed by the dissolution of Fe(III) from the surface of BHF–EDTMP NPL (Figure 4b). The most inner EDTMP layer is chemisorbed via surface Fe(III) ions onto BHF NPLs surfaces, while subsequent layer(s) can be bridged via the dissolved Fe(III) (eq 1) or/and physisorbed onto the first (previous) layer via hydrogen bonds between unbonded phosphonic groups and/or physisorbed water (see the ATR-FTIR spectrum Figure 2b and related discussion in the EDTMP Coating section). The layers, however, are not homogeneous and densely packed as with OPA and PSA coatings, but they form a porous network. In line with the above, multilayered porous coating can also be made of DPA. DPA preferentially binds to the BHF surface with only one rather than with both phosphonic groups (negative zeta-potential, Figure S3), i.e., similar to PSA. However, DPA can bond the released Fe(III) ions with the unbonded phosphonic group, thus forming a framework of Fe(III)–DPA complexes (similar as in eq 1). In contrast to the DPA, the unbonded sulfonic group of the PSA coating has much lower interaction energy with Fe(III) than the phosphonic group41 and does not bind with Fe(III) ions under the studied conditions. Another difference between the DPA and PSA is in their solubility in water. Due to its relatively low solubility, DPA also homogeneously precipitates on the NPLs surfaces during cooling from the synthesis temperature (80 °C) and so forms very thick coatings and homogeneous spherical precipitates (Figure S1d, Supporting Information). This coincides with the relatively high fraction of bonded DPA, i.e., ∼90% of the nominal fraction (Table 1).

Water Purification

To demonstrate a potential of the new porous nanomagnets for the remediation of water contaminated with heavy metals, we performed a preliminary test of Bi(III) adsorption from an aqueous solution using the BHF–EDTMP NPLs. After 16 h treatment, the initial Bi concentration of 0.1 mg/mL decreased below 0.01 mg/mL, i.e., below the detection limit of the ICP-OES analysis. This means that at least 0.135 mg (i.e., ≥90%) of Bi was removed with 0.3 mg of BHF–EDTMP NPLs. A more comprehensive study, also involving a recycling procedure, is planned in the near future. Considering the great affinity of phosphonates for complexing of the multivalent metal ions, we expect that the new porous nanomagnets will be efficient for water purification of heavy metals in general.

Conclusions

Amorphous coatings formed on the barium hexaferrite nanoplatelets (BHF NPLs) from different phosphonic acids: ethylenediaminetetra(methylenephosphonic acid) (EDTMP), octadecylphosphonic acid (OPA), (12-phosphono)dodecyl phosphonic acid (DPA), and (12-phosphono)dodecyl sulfonic acid (PSA). Dense, ca. 1–2 nm thick, coatings were obtained with monophosphonic acids, OPA and PSA, that assembled in a bilayer on the BHF NPLs surfaces. Thicker and porous coatings were obtained from EDTMP and DPA with four and two phosphonic groups, respectively. The proposed mechanism for their formation is based on the strong interaction between the surface Fe(III) ions and the phosphonic groups, which induces the dissolution of the BHF NPLs in water and the consequent precipitation of stable Fe(III)-phosphonates on the NPLs. In this way, some fraction of the phosphonic acid is chemisorbed directly on the NPLs surfaces, while the other fraction is bridged by the dissolved Fe(III) and/or physisorbed via hydrogen bonds, forming a multilayered porous coating. The conditions for the dissolution of BHF NPLs were identified. The phosphonic acid must be (at least partly) soluble in water and should contain at least two phosphonic groups, while the coating process should be done in an aqueous medium. The porosity can be coupled in the same way with any other functional oxide particles and surfaces using the same approach.

All of the coated BHF NPLs showed typical magnetic behavior at room temperature and high sensitivity to the magnetic field of few millitesla but the EDTMP and DPA coatings also had a high specific surface area (several hundred m2/g). We conclude that the BHF NPLs with porous EDTMP coatings represent a model example of porous nanomagnets to be used for magnetically directed catalysis and remotely triggered chemical reactions (e.g., under extreme conditions), drug delivery and release, and magnetically retractable sorbents. Our preliminary experiment showed that such porous nanomagnets can be especially useful for the purification of waste or contaminated water from heavy metals.

Acknowledgments

The authors acknowledge the financial support from the Slovenian Research Agency (D.L. and D.M. research core funding no. P2-0089; P.H.B. and A.M. research core funding P1-0192; M.V. research core funding P2-0412; J.K. research core funding P2-0082; A.K. research core funding P2-0087; D.L., A.M., and M.V. project J7-8276 and P.H.B. project PR-08415). They also acknowledge CENN Nanocenter for the use of the VSM and TEM (JEOL 2100).

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.0c01597.

  • Experimental details on the variation of the synthesis parameters for BHF–EDTMP NPLs and on the determination of bonded phosphonic acids with TGA; calculations of the coating density and specific surface area; identified ATR-FTIR bands of EDTMP (Table S1); MS data for EDTMP (Table S2); and additional results for the coated NPLs: TEM images (Figure S1), XPS spectra (Figure S2), zeta-potential behavior (Figure S3), magnetic hysteresis loops (Figure S4), ATR-FTIR spectra (Figure S6), and the STEM images of the BHF–OPA NPL (Figure S7) (PDF)

Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

The authors declare no competing financial interest.

Supplementary Material

ao0c01597_si_001.pdf (1.7MB, pdf)

References

  1. Pullar R. C. Hexagonal Ferrites: A Review of the Synthesis, Properties and Applications of Hexaferrite Ceramics. Prog. Mater. Sci. 2012, 57, 1191–1334. 10.1016/j.pmatsci.2012.04.001. [DOI] [Google Scholar]
  2. Drofenik M.; Kristl M.; Žnidaršič A.; Hanžel D.; Lisjak D. Hydrothermal Synthesis of Ba-Hexaferrite Nanoparticles. J. Am. Ceram. Soc. 2007, 90, 2057–2061. 10.1111/j.1551-2916.2007.01740.x. [DOI] [Google Scholar]
  3. Lisjak D.; Drofenik M. Chemical Substitution-an Alternative Strategy for Controlling the Particle Size of Barium Ferrite. Cryst. Growth Des. 2012, 12, 5174–5079. 10.1021/cg301227r. [DOI] [Google Scholar]
  4. Mertelj A.; Lisjak D.; Drofenik M.; Čopič M. Ferromagnetism in Suspensions of Magnetic Platelets in Liquid Crystal. Nature 2013, 504, 237–241. 10.1038/nature12863. [DOI] [PubMed] [Google Scholar]
  5. Shuai M.; Klittnick A.; Shen Y.; Smith G. P.; Tuchband M. R.; Zhu C.; Petschek R. G.; Mertelj A.; Lisjak D.; Čopič M.; et al. Spontaneous Liquid Crystal and Ferromagnetic Ordering of Colloidal Magnetic Nanoplates. Nat. Commun. 2016, 7, 10394 10.1038/ncomms10394. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Lisjak D.; Mertelj A. Anisotropic Magnetic Nanoparticles: A Review of Their Properties, Syntheses and Potential Applications. Prog. Mater. Sci. 2018, 95, 286–328. 10.1016/j.pmatsci.2018.03.003. [DOI] [Google Scholar]
  7. Hu J.; Gorsak T.; Martín Rodríguez E.; Calle D.; Muñoz-Ortiz T.; Jaque D.; Fernández N.; Cussó L.; Rivero F.; Aguilar Torres R.; et al. Magnetic Nanoplatelets for High Contrast Cardiovascular Imaging by Magnetically Modulated Optical Coherence Tomography. ChemPhotoChem 2019, 3, 529–539. 10.1002/cptc.201900071. [DOI] [Google Scholar]
  8. Koplovitz G.; Primc D.; Ben Dor O.; Yochelis S.; Rotem D.; Porath D.; Paltiel Y. Magnetic Nanoplatelet-Based Spin Memory Device Operating at Ambient Temperatures. Adv. Mater. 2017, 29, 1606748 10.1002/adma.201606748. [DOI] [PubMed] [Google Scholar]
  9. Zeng M.; Huang D.; Wang P.; King D.; Peng B.; Luo J.; Lei Q.; Zhang L.; Wang L.; Shinde A.; et al. Autonomous Catalytic Nanomotors Based on 2D Magnetic Nanoplates. ACS Appl. Nano Mater. 2019, 2, 1267–1273. 10.1021/acsanm.8b02153. [DOI] [Google Scholar]
  10. Pevec S.; Donlagić D. Multiparameter Fiber-Optic Sensors: A Review. Opt. Eng. 2019, 58, 1–26. 10.1117/1.OE.58.7.072009. [DOI] [Google Scholar]
  11. Goršak T.; Makovec D.; Javornik U.; Belec B.; Kralj S.; Lisjak D. A Functionalization Strategy for the Dispersion of Permanently Magnetic Barium-Hexaferrite Nanoplatelets in Complex Biological Media. Colloids Surf., A 2019, 573, 119–127. 10.1016/j.colsurfa.2019.04.051. [DOI] [Google Scholar]
  12. Kralj S.; Potrc T.; Kocbek P.; Marchesan S.; Makovec D. Design and Fabrication of Magnetically Responsive Nanocarriers for Drug Delivery. Curr. Med. Chem. 2017, 24, 454–469. 10.2174/0929867323666160813211736. [DOI] [PubMed] [Google Scholar]
  13. Mehmood A.; Ghafar H.; Yaqoob S.; Gohar U. F.; Ahmad B. Mesoporous Silica Nanoparticles: A Review. J. Dev. Drugs 2017, 06, 100174 10.4172/2329-6631.1000174. [DOI] [Google Scholar]
  14. Went J. J.; Rathenau G. W.; Gorter E. W.; van Oosterhout G. W. Ferroxdure a Class of New Permanent Magnet Materials. Philips Tech. Rev. 1952, 13, 194–208. [Google Scholar]
  15. Makovec D.; Belec B.; Goršak T.; Lisjak D.; Komelj M.; Dražić G.; Gyergyek S. Discrete Evolution of the Crystal Structure during the Growth of Ba-Hexaferrite Nanoplatelets. Nanoscale 2018, 10, 14480–14491. 10.1039/C8NR03815E. [DOI] [PubMed] [Google Scholar]
  16. Makovec D.; Komelj M.; Dražić G.; Belec B.; Goršak T.; Gyergyek S.; Lisjak D. Incorporation of Sc into the Structure of Barium-Hexaferrite Nanoplatelets and Its Extraordinary Finite-Size Effect on the Magnetic Properties. Acta Mater. 2019, 172, 84–91. 10.1016/j.actamat.2019.04.050. [DOI] [Google Scholar]
  17. Hribar Boštjančič P.; Tomšič M.; Jamnik A.; Lisjak D.; Mertelj A. Electrostatic Interactions between Barium Hexaferrite Nanoplatelets in Alcohol Suspensions. J. Phys. Chem. C 2019, 123, 23272–23279. 10.1021/acs.jpcc.9b07455. [DOI] [Google Scholar]
  18. Queffélec C.; Petit M.; Janvier P.; Knight D. A.; Bujoli B. Surface Modification Using Phosphonic Acids and Esters. Chem. Rev. 2012, 112, 3777–3807. 10.1021/cr2004212. [DOI] [PubMed] [Google Scholar]
  19. Yee C.; Kataby G.; Ulman A.; Prozorov T.; White H.; King A.; Rafailovich M.; Sokolov J.; Gedanken A. Self-Assembled Monolayers of Alkanesulfonic and -Phosphonic Acids on Amorphous Iron Oxide Nanoparticles. Langmuir 1999, 15, 7111–7115. 10.1021/la990663y. [DOI] [Google Scholar]
  20. Woodward J. T.; Ulman A.; Schwartz D. K. Self-Assembled Monolayer Growth of Octadecylphosphonic Acid on Mica. Langmuir 1996, 12, 3626–3629. 10.1021/la9510689. [DOI] [Google Scholar]
  21. Lalatonne Y.; Paris C.; Serfaty J. M.; Weinmann P.; Lecouvey M.; Motte L. Bis-Phosphonates-Ultra Small Superparamagnetic Iron Oxide Nanoparticles: A Platform towards Diagnosis and Therapy. Chem. Commun. 2008, 2553–2555. 10.1039/b801911h. [DOI] [PubMed] [Google Scholar]
  22. Rizkalla E. N.; Zaki M. T. M. Metal Chelates of Phosphonate-Coating Ligands-I″, Stability of Some N’,N’,N’N’-Ethylenediaminetetra(methylendphosphonic) Acid Metal Chelates. Talanta 1979, 26, 507–510. 10.1016/0039-9140(79)80120-4. [DOI] [PubMed] [Google Scholar]
  23. Westerback S.; Rajan K. S.; Martell A. E. New Multidentate Ligands. III. Amino Acids Containing Methylenephosphonate Groups. J. Am. Chem. Soc. 1965, 87, 2567–2572. 10.1021/ja01090a008. [DOI] [Google Scholar]
  24. Das M.; Mishra D.; Dhak P.; Gupta S.; Maiti T. K.; Basak A.; Pramanik P. Biofunctionalized, Phosphonate-Grafted, Ultrasmall Iron Oxide Nanoparticles for Combined Targeted Cancer Therapy and Multimodal Imaging. Small 2009, 5, 2883–2893. 10.1002/smll.200901219. [DOI] [PubMed] [Google Scholar]
  25. Li R.; Ji Z.; Dong J.; Chang C. H.; Wang X.; Sun B.; Wang M.; Liao Y.-P.; Zink J. I.; Nel A. E.; et al. Enhancing the Imaging and Biosafety of Upconversion Nanoparticles through Phosphonate Coating. ACS Nano 2015, 9, 3293–3306. 10.1021/acsnano.5b00439. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Cai X.; Lee A.; Ji Z.; Huang C.; Chang C. H.; Wang X.; Liao Y. P.; Xia T.; Li R. Reduction of Pulmonary Toxicity of Metal Oxide Nanoparticles by Phosphonate-Based Surface Passivation. Part. Fibre Toxicol. 2017, 14, 1–11. 10.1186/s12989-017-0193-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Guerrero G.; Mutin P. H.; Vioux A. Organically Modified Aluminas by Grafting and Sol-Gel Processes Involving Phosphonate Derivatives. J. Mater. Chem. 2001, 11, 3161–3165. 10.1039/b104411g. [DOI] [Google Scholar]
  28. Freedman L. D.; Doak G. O. The Preparation and Properties of Phosphonic Acids. Chem. Rev. 1957, 57, 479–523. 10.1021/cr50015a003. [DOI] [Google Scholar]
  29. Nowack B.; Stone A. T. Adsorption of Phosphonates onto the Goethite-Water Interface. J. Colloid Interface Sci. 1999, 214, 20–30. 10.1006/jcis.1999.6111. [DOI] [PubMed] [Google Scholar]
  30. Ma X.; Chai Y.; Li P.; Wang B. Metal-Organic Framework Films and Their Potential Applications in Environmental Pollution Control. Acc. Chem. Res. 2019, 52, 1461–1470. 10.1021/acs.accounts.9b00113. [DOI] [PubMed] [Google Scholar]
  31. Gao Q.; Guillou N.; Nogues M.; Cheetham A. K.; Férey G. Structure and Magnetism of VSB-2, -3, and -4 or Ni4(O3P-(CH2)-PO3)2·(H2O)n (n = 3, 2, 0), the First Ferromagnetic Nickel(II) Diphosphonates: Increase of Dimensionality and Multiple Coordination Changes during a Quasi Topotactic Dehydration. Chem. Mater. 1999, 11, 2937–2947. 10.1021/cm9910714. [DOI] [Google Scholar]
  32. Demadis K. D.; Armakola E.; Papathanasiou K. E.; Mezei G.; Kirillov A. M. Structural Systematics and Topological Analysis of Coordination Polymers with Divalent Metals and a Glycine-Derived Tripodal Phosphonocarboxylate. Cryst. Growth Des. 2014, 14, 5234–5243. 10.1021/cg501029b. [DOI] [Google Scholar]
  33. Cunha-Silva L.; Lima S.; Ananias D.; Silva P.; Mafra L.; Carlos L. D.; Pillinger M.; Valente A. A.; Paz F. A. A.; Rocha J. Multi-Functional Rare-Earth Hybrid Layered Networks: Photoluminescence and Catalysis Studies. J. Mater. Chem. 2009, 19, 2618–2632. 10.1039/b817381h. [DOI] [Google Scholar]
  34. Socrates G.Infrared Characteristic Group Frequencies; John Wiley & Sons Ltd.: Chichester, 1994. [Google Scholar]
  35. Princz E.; Szilágyi I.; Mogyorósi K.; Labádi I. Lanthanide Complexes of Ethylenediaminotetramethylene-phosphonic acid. J. Therm. Anal. Calorim. 2002, 69, 427–439. 10.1023/A:1019943319341. [DOI] [Google Scholar]
  36. Basly B.; Popa G.; Fleutot S.; Pichon B. P.; Garofalo A.; Ghobril C.; Billotey C.; Berniard A.; Bonazza P.; Martinez H.; et al. Effect of the Nanoparticle Synthesis Method on Dendronized Iron Oxides as MRI Contrast Agents. Dalton Trans. 2013, 42, 2146–2157. 10.1039/C2DT31788E. [DOI] [PubMed] [Google Scholar]
  37. Walter A.; Garofalo A.; Bonazza P.; Meyer F.; Martinez H.; Fleutot S.; Billotey C.; Taleb J.; Felder-Flesch D.; Begin-Colin S. Effect of the Functionalization Process on the Colloidal, Magnetic Resonance Imaging, and Bioelimination Properties of Mono- or Bisphosphonate-Anchored Dendronized Iron Oxide Nanoparticles. ChemPlusChem 2017, 82, 647–659. 10.1002/cplu.201700049. [DOI] [PubMed] [Google Scholar]
  38. Karimi A.; Denizot B.; Hindré F.; Filmon R.; Greneche J. M.; Laurent S.; Daou T. J.; Begin-Colin S.; Le Jeune J. J. Effect of Chain Length and Electrical Charge on Properties of Ammonium-Bearing Bisphosphonate-Coated Superparamagnetic Iron Oxide Nanoparticles: Formulation and Physicochemical Studies. J. Nanopart. Res. 2010, 12, 1239–1248. 10.1007/s11051-009-9815-7. [DOI] [Google Scholar]
  39. Derory A.; Donnio B.; Daou T. J.; Grenèche J. M.; Begin-Colin S.; Guillon D.; Buathong S.; Pourroy G.; Ulhaq-Bouillet C. Coupling Agent Effect on Magnetic Properties of Functionalized Magnetite-Based Nanoparticles. Chem. Mater. 2008, 20, 5869–5875. 10.1021/cm801405n. [DOI] [Google Scholar]
  40. Cebollada F.; González J. M.; De Frutos J.; González A. M. Reversal Mechanisms and Interactions in Magnetic Systems: Coercivity versus Switching Field and Thermally Assisted Demagnetization. Bol. Soc. Esp. Ceram. Vidrio 2005, 44, 169–176. 10.3989/cyv.2005.v44.i3.385. [DOI] [Google Scholar]
  41. Gao W.; Dickinson L.; Grozinger C.; Morin F. G.; Reven L. Self-Assembled Monolayers of Alkylphosphonic Acids on Metal Oxides. Langmuir 1996, 12, 6429–6435. 10.1021/la9607621. [DOI] [Google Scholar]
  42. Sahoo Y.; Pizem H.; Fried T.; Golodnitsky D.; Burstein L.; Sukenik C. N.; Markovich G. Alkyl Phosphonate/Phosphate Coating on Magnetite Nanoparticles: A Comparison with Fatty Acids. Langmuir 2001, 17, 7907–7911. 10.1021/la010703+. [DOI] [Google Scholar]
  43. Portet D.; Denizot B.; Rump E.; Lejeune J. J.; Jallet P. Nonpolymeric Coatings of Iron Oxide Colloids for Biological Use as Magnetic Resonance Imaging Contrast Agents. J. Colloid Interface Sci. 2001, 238, 37–42. 10.1006/jcis.2001.7500. [DOI] [PubMed] [Google Scholar]
  44. Pauly C. S.; Genix A. C.; Alauzun J. G.; Sztucki M.; Oberdisse J.; Hubert Mutin P. Surface Modification of Alumina-Coated Silica Nanoparticles in Aqueous Sols with Phosphonic Acids and Impact on Nanoparticle Interactions. Phys. Chem. Chem. Phys. 2015, 17, 19173–19182. 10.1039/C5CP01925G. [DOI] [PubMed] [Google Scholar]
  45. Li L.; Stanforth R. Distinguishing Adsorption and Surface Precipitation of Phosphate on Goethite (α-FeOOH). J. Colloid Interface Sci. 2000, 230, 12–21. 10.1006/jcis.2000.7072. [DOI] [PubMed] [Google Scholar]
  46. Řehoř I.; Kubíček V.; Kotek J.; Hermann P.; Száková J.; Lukeš I. Modification of Nanocrystalline TiO2 with Phosphonate- and Bis(Phosphonate)-Bearing Macrocyclic Complexes: Sorption and Stability Studies. Eur. J. Inorg. Chem. 2011, 2011, 1981–1989. 10.1002/ejic.201001100. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ao0c01597_si_001.pdf (1.7MB, pdf)

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES