Article Cite This: J. Agric. Food Chem. 2017, 65, 11157−11169
pubs.acs.org/JAFC
Key Roles of Size and Crystallinity of Nanosized Iron Hydr(oxides) Stabilized by Humic Substances in Iron Bioavailability to Plants Natalia A. Kulikova,†,‡,§ Alexander Yu. Polyakov,∥ Vasily A. Lebedev,‡,∥ Dmitry P. Abroskin,† Dmitry S. Volkov,‡ Denis A. Pankratov,‡ Olga I. Klein,‡,§ Svetlana V. Senik,⊥ Tatiana A. Sorkina,‡,# Alexey V. Garshev,‡,∥ Alexey A. Veligzhanin,∇ Jose M. Garcia Mina,○ and Irina V. Perminova*,‡ †
Department of Soil Science, Lomonosov Moscow State University, Leninskie gory 1-12, 119991 Moscow, Russia Department of Chemistry, Lomonosov Moscow State University, Leninskie gory 1-3, 119991 Moscow, Russia § Bach Institute of Biochemistry, Fundamentals of Biotechnology Federal Research Center, Russian Academy of Sciences, pr. Leninskii 33, 119071 Moscow, Russia ∥ Department of Materials Science, Lomonosov Moscow State University, Leninskie gory 1-73, 119991 Moscow, Russia ⊥ Komarov Botanical Institute, Russian Academy of Sciences, ul. Professora Popova 2, 197376 St. Petersburg, Russia # Science & Technology Department, Rusnano LLC., 10A, prospect 60-letia Oktyabrya, 117036 Moscow, Russia ∇ National Research Center “Kurchatov Institute”, 1, Akademika Kurchatova pl., 123182 Moscow, Russia ○ Department of Environmental Biology, BACh group, Sciences School, University of Navarra, C/Irunlarrea 1, 31008 na, Pamplona, Spain ‡
S Supporting Information *
ABSTRACT: Availability of Fe in soil to plants is closely related to the presence of humic substances (HS). Still, the systematic data on applicability of iron-based nanomaterials stabilized with HS as a source for plant nutrition are missing. The goal of our study was to establish a connection between properties of iron-based materials stabilized by HS and their bioavailability to plants. We have prepared two samples of leonardite HS-stabilized iron-based materials with substantially different properties using the reported protocols and studied their physical chemical state in relation to iron uptake and other biological effects. We used Mössbauer spectroscopy, XRD, SAXS, and TEM to conclude on iron speciation, size, and crystallinity. One material (Fe-HA) consisted of polynuclear iron(III) (hydr)oxide complexes, so-called ferric polymers, distributed in HS matrix. These complexes are composed of predominantly amorphous small-size components (90%).11 At the same time, Zhu and coauthors discovered significant uptake, translocation, and accumulation of 20 nm magnetite (Fe3O4) NPs both in the roots and leaves of pumpkin Cucurbita maxima Duch.12 The study showed that about 45.5% of fed NPs were accumulated in roots and 0.6% in leaves. However, the same study did not observe these effects for lima bean Phaseolus limensis L. Hematite (Fe2O3) NPs of 20 nm size were tested as Fe source for peanut (Arachis hypogaea L.) in the pot experiments and showed higher Fe contents in peanut plants than in the control group.13 Still, the study on hematite NPs of 40.9 nm has shown a lack of uptake by the Arabidopsis thaliana (L.) Heynh. plant during the entire growth period of 56 days.14 The above results indicate that iron-based engineered NPs uptake by plants is dependent on both plant physiology and the sizes of NPs. It was reported that any particle of >5 nm diameter has a hindered or zero passage through the plant cell wall.15 At the same time, the recent report on the uptake of carbon-bound iron oxide NPs of 30 mV have sufficient electrostatic repulsion and are stable
A negative zeta potential was observed for HA (−33 mV) and Fe-HA (−44 mV), whereas δ’-FeOOH-HS was characterized with a positive potential (33 mV). The negative charge of HA is provided by deprotonation of their carboxyl functionalities at neutral pH, the negative charge of Fe-HA particles is consistent with the dominant contribution of leonardite humic acids (HA) 11160
DOI: 10.1021/acs.jafc.7b03955 J. Agric. Food Chem. 2017, 65, 11157−11169
Article
Journal of Agricultural and Food Chemistry Table 2. Parameters of the Mössbauer Spectra Recorded at 78 and 298 K for δ’-FeOOH-HS δda temperature, K
mode
78
1 2 1 2 3
298
Δd (2εd)
Γ
mm/s 0.48 0.49 0.37 0.36 0.34
± ± ± ± ±
0.01 0.01 0.01 0.01 0.01
−0.04 ± 0.06 0.14 ± 0.02 0.76 ± 0.01 −0.52 ± 0.02 −0.07 ± 0.02
0.38 ± 0.01
Hd
Sd
kOe
%
491 ± 9 518 ± 2
0.62 ± 0.02 163 ± 13 394 ± 8
49 51 29 10 62
± ± ± ± ±
5 5 2 3 3
a δd, isomer shift; Δd (2εd), quadrupole splitting; Hd, hyperfine magnetic field; Sd, relative area for maximum of mode distribution functions; Γ, line width.
Figure 3. XRD patterns of the samples under study: Fe-HA sample as measured (a), Fe-HA sample with subtracted crystalline phases of Na2SO4 (b); the humic acid (HA) sample as measured (c); δ’-FeOOH-HS sample as measured (d).
in suspension.39 This is indicative of high colloidal stability of both Fe-HA and δ’-FeOOH-HS suspensions. 3.2. Iron Speciation and Phase Identification. Mössbauer spectroscopy was applied to study iron speciation in both samples used in this study. The Mössbauer spectra of the FeHA sample at 78 and 298 K (Figure 1) can be interpreted as the sum of two symmetric doublets with close isomer shift values (δ ≈ 0.5 mm s−1) corresponding to octahedral iron(III) compounds40,41 but with different quadrupole splitting values (0.35 and 0.58 mm s−1) and line widths (Table 1). The relative area of the doublet with smaller quadrupole splitting value (Table 1, subspectrum 1) accounts for 25%. It means that only 25% of iron atoms behave as crystalline state at the temperature of liquid nitrogen, which is indicative of low ordering degree (low crystallinity) of the Fe-HA sample. A lack of the temperature dependence of the quadrupole splitting and line widths (Table 1) may be indicative of the paramagnetic nature of this sample. It should be noted that the Mössbauer parameters for the Fe-HA sample obtained in this study are very similar to those reported by Johnston and Lewis38 for nanosized polynuclear complexes composed of [Fe2(OH)2]4+ dimers, which were formed under hydrolysis of iron nitrate.
This might indicate that the Fe-HA sample is comprised of the largely disordered noncrystalline ferric polymers, which are distributed in the HA matrix. The Mössbauer spectra of δ’-FeOOH-HS are shown in Figure 2. The spectra are typical for nanosized iron-containing particles of superparamagnetic materials with the blocking temperature laying between 78 and 298 K (Figure 2a). The spectral shape suggests that the magnetic fluctuations are influenced by interparticle interactions in their polydispersed ensemble. At liquid nitrogen temperature, such a spectrum might be described by a superposition of sextets (components). The quantitative parameters were calculated by fitting experimental data to the multiple-state superparamagnetic relaxation distribution model.43 The obtained data (Table 2) are indicative of oxidation state (+3) for all iron atoms in the δ’FeOOH-HS sample.44 The distribution functions of the parameters calculated from δ’-FeOOH-HS spectra at the both temperatures are consistent with the data for feroxyhyte.24,43 A lower value of the hyperfine magnetic field may be associated with the absence of the magnetic field saturation and the small particle size. Hence, it can be concluded that the δ’-FeOOH 11161
DOI: 10.1021/acs.jafc.7b03955 J. Agric. Food Chem. 2017, 65, 11157−11169
Article
Journal of Agricultural and Food Chemistry sample contains feroxyte NPs in the magnetically ordered crystalline state. The phase analysis using XRD (Figure 3) shows that the FeHA sample contains a very low amount of crystalline fraction other than traces of Na2SO4 which can be seen as sharp reflexes in Figure 3a. Subtraction of these reflexes from the measured XRD using appropriate software yields XRD which is very similar to initial humic material which does not contain any crystalline phases in it (Figure 3b,c). At the same time the δ’FeOOH-HS sample contains a crystalline phase, which can be assigned to feroxyhyte (ICDD PDF2 No. 13-87) and traces of halite (ICDD PDF2 No. 5-628; Figure 3d). The XRD data obtained for the crystalline δ’-FeOOH-HS sample were analyzed by the leBail method to refine lattice parameters and the size of coherent scattering regions (the details are described in the SI). The halite was used as an internal standard in the calculations. The refined cell parameters were as follows: a = b = 2.9452(7)Å, c = 4.561(6)Å, the cell volume −34.26(5)Å3. The calculated coherent scattering region (CSR) sizes were 5 ± 0.5 nm in the [0 0 2] direction (platelet thickness) and 40 ± 10 nm in the orthogonal direction (transversal plane). Poor crystallinity of the Fe-HA sample, which was demonstrated both by XRD and Mössbauer spectroscopy, did not allow size assignment to iron species present in this sample. For resolving this issue we have used SAXS, which is applicable both to amorphous and crystalline particles. The experimental scattering curves are given in Figure S4 in the SI along with the details of data treatment and fitting parameters. Given the unknown morphology of the Fe-HA complexes, the fitting of the scattering data of the Fe-HA sample was performed using Beaucage unified function, which describes a material over a wide range of sizes in terms of structural levels.45 This fitting yielded the gyration radius value (Rg) of (3.82 ± 0.02) nm and the P value (power law slope) of (3.3 ± 0.1). The latter is indicative of the surface fractal properties of the particles present in the Fe-HA sample with surface fractal dimension of DS = 6 − P = 2.7 ± 0.1. This means that the scattering objects have dense inner core and rough developed surface, which might reflect aggregation of humic molecules on polymeric iron (hydr)oxides. The upper size of such objects is constrained by the obtained Rg value of 3.82 nm. The SAXS data from the platelet-like NPs of the δ’-FeOOHHS sample were fitted to the shape-anisotropy model, which yielded the particle thickness of (4.6 ± 0.1) nm and the transverse size of >30 nm (beyond the upper size-limit of the used SAXS setup). The obtained thickness value is in good agreement with the CSR calculations, while the transverse size is larger than the XRD-derived one. It can be due to the polycrystalline nature of feroxyhyte NPs. 3.3. Particle Size, Crystallinity, and Iron Distribution Visualization Using TEM. TEM was used for visual characterization of crystallinity and size of the iron-containing species present in the samples under study. The typical TEM images for the δ’-FeOOH-HS sample are shown in Figure 4 (more images are shown in the SI, Figure S5). The sample contains large amounts of platelet-like particles oriented either parallel or perpendicular to the plane of the TEM grid. A large portion of these particles is hexagonally faceted. The mean transversal size of the particles is (35 ± 20) nm, and the mean thickness is (3 ± 1) nm, which is in good agreement with XRD and SAXS data. The further details on the nature of the δ’FeOOH-HS sample were deduced from the data of high-
Figure 4. TEM and HRTEM images of the δ’-FeOOH-HS sample: (a, b) TEM images with inset of SAED collected from a; (c, d) the size distributions of the platelet-like feroxyhyte NPs calculated for a and b, respectively; (e, f) HRTEM images with insets of fast Fourier transforms; (g) depicts radial integrated SAED patterns for the large area of δ’-FeOOH-HS sample (bottom), as well as small areas with NPs oriented vertically (middle) and horizontally (top); the corresponding selected areas are given on the insets in panel g.
resolution TEM (Figure 4c,d) images and their Fourier transformations, which correspond to the array of (100) planes of feroxyhyte. The electron diffraction pattern is shown in the inset in Figure 4a, and SAED is given in the SI (Figure S6). The results of radial integrations of SAED patterns are plotted in Figure 4g with the corresponding areas of signal generation: the feroxyhyte (δ’-FeOOH) lines (ICDD PDF2 No. 13-87) are labeled. This is a strong confirmation that the synthesized sample is composed of feroxyhyte NPs. The platelets orientation was determined from the SAED patterns of the small amount of particles which were oriented in parallel or perpendicular to the grid surface, as it is demonstrated in the inset of Figure 4g and in Figure S7. It was concluded that (002) 11162
DOI: 10.1021/acs.jafc.7b03955 J. Agric. Food Chem. 2017, 65, 11157−11169
Article
Journal of Agricultural and Food Chemistry
Figure 5. TEM images of the Fe-HA sample measured with different magnifications (a, b); size distribution of the observed inclusions (c); phase contrast microscopy: the bright-field (d) and the dark-field (e) images. The bright inclusions in the dark-field image (d) are crystalline. For the details of SAED reflexes used for dark-field imaging see Figure S10.
sample, and with the estimates of the Mössbauer spectroscopy that only 25% of Fe atoms in this sample are included in crystalline structures, whereas much larger amount (75%) are incorporated into the amorphous structures. 3.4. Suggested Modes of Fe-HS Interactions for the Two Humics-Based Materials under Study. Based on the conducted physical-chemical investigation we suggest that ironcontaining compounds in the Fe-HA sample are rather comprised of nanosized polynuclear iron-oxyhydroxide complexes of different molecular weight (ferric polymers) coordinated on surfaces by HA molecules than by molecular HA complexes with ionic iron. The formation of these ferric polymers was reported during FeIII salts hydrolysis,2,42 and it could easily take place during preparation of the Fe-HA sample. We did not use inert atmosphere for this synthesis, and it is known that complexation with leonardite HA does not prevent Fe(II) from oxidation.22,25,28 The ferric polymers are composed of chaotically ordered {FeIII(O, OH)6} octahedra resulting in a lack of long-range order. Carboxylic and phenolic groups of the humic macroligands can partially substitute oxygen or hydroxyl groups in the {FeIII(O, OH)6} octahedra resulting in entrapment of iron polynuclear complexes into the branched humic molecules. These iron-HS interactions will further inhibit particle growth and facilitate formation of small partially amorphous (20 nm, but they were partially permeable for the NPs with sizes
