Article Rheological, Microstructural and Thermal Properties of Magnetic Poly(Ethylene Oxide)/ Oxide Hydrogels Synthesized Using a One-Step Gamma-Irradiation Method

Ivan Mari´c 1, Nataša Šijakovi´cVujiˇci´c 2 , Andela¯ Pustak 1, Marijan Goti´c 3, Goran Štefani´c 3, Jean-Marc Grenèche 4 , Goran Draži´c 5 and Tanja Jurkin 1,* 1 Radiation Chemistry and Dosimetry Laboratory, Division of Materials Chemistry, Ruder¯ Boškovi´cInstitute, BijeniˇckaCesta 54, 10000 Zagreb, Croatia; [email protected] (I.M.); [email protected] (A.P.) 2 Laboratory for Supramolecular Chemistry, Division of Organic Chemistry and Biochemistry, Ruder¯ Boškovi´cInstitute, BijeniˇckaCesta 54, 10000 Zagreb, Croatia; [email protected] 3 Laboratory for Molecular Physics and Synthesis of New Materials, Division of Materials Physics, Ruder¯ Boškovi´cInstitute, BijeniˇckaCesta 54, 10000 Zagreb, Croatia; [email protected] (M.G.); [email protected] (G.Š.) 4 Institut des Molécules et Matériaux du Mans (IMMM UMR CNRS 6283), Le Mans Université, Avenue Olivier Messiaen, F-72085 Le Mans CEDEX 9, France; [email protected] 5 Department of Materials Chemistry, National Institute of Chemistry, Hajdrihova 19, SI-1001 Ljubljana, Slovenia; [email protected] * Correspondence: [email protected]; Tel.: +385-1-4571-255

 Received: 31 July 2020; Accepted: 10 September 2020; Published: 12 September 2020 

Abstract: Magnetic polymer gels are a new promising class of nanocomposite gels. In this work, magnetic PEO/ were synthesized using the one-step γ-irradiation method starting from poly(ethylene oxide) (PEO) and iron(III) precursor alkaline aqueous suspensions followed by simultaneous crosslinking of PEO chains and reduction of Fe(III) precursor. γ-irradiation dose and concentrations of Fe3+, 2-propanol and PEO in the initial suspensions were varied and optimized. With 2-propanol and at high doses magnetic gels with embedded were obtained, as confirmed by XRD, SEM and Mössbauer spectrometry.The quantitative determination of γ-irradiation generated Fe2+ was performed using the 1,10-phenanthroline method. The maximal Fe2+ molar fraction of 0.55 was achieved at 300 kGy, pH = 12 and initial 5% of Fe3+. The DSC and rheological measurements confirmed the formation of a well-structured network. The thermal and rheological properties of gels depended on the dose, PEO concentration and initial Fe3+ content (amount of nanoparticles synthesized inside gels). More amorphous and stronger gels were formed at higher dose and higher content. The properties of synthesized gels were determined by the presence of magnetic iron oxide nanoparticles, which acted as reinforcing agents and additional crosslinkers of PEO chains thus facilitating the one-step gel formation.

Keywords: magnetic hydrogel; gamma-irradiation; poly(ethylene oxide); magnetite; rheological properties; thermal properties; 57Fe Mössbauer spectrometry; XRD; SEM; Fe(II) determination

1. Introduction Magnetic polymer gels (ferrogels) are a new and promising class of nanocomposite hydrogels that have the potential to be used as effective absorbents of toxic in water, protein immobilization, separation, in soft actuators such as artificial muscles, in tissue engineering, drug delivery and hyperthermia applications [1–7]. Ferrogels combine the elastic properties and the defined structure

Nanomaterials 2020, 10, 1823; doi:10.3390/nano10091823 www.mdpi.com/journal/nanomaterials Nanomaterials 2020, 10, 1823 2 of 25 of gels with the magnetic properties of (usually magnetite or ) andNanomaterials respond quickly 2019, 9, x FOR to the PEER external REVIEW magnetic field. Apart from their unique magnetic properties,2 of 29 nanocomposite gel scaffolds with embedded magnetite/maghemite nanoparticles have exhibited defined structure of gels with the magnetic properties of magnetic nanoparticles (usually magnetite superior mechanical, rheological and electrical properties compared to scaffold gels without nanoparticle or maghemite) and respond quickly to the external . Apart from their unique magnetic reinforcement,properties, betternanocomposite biocompatibility, gel scaffolds low with cytotoxicity, embedded and magnetite/maghemite demonstrated antibacterial nanoparticles properties have [1–4]. Furthermore,exhibited ironsuperior oxide mechanical, nanoparticles rheological have been and shown electrical to promoteproperties osteogenic compared di tofferentiation scaffold gels of stem cellswithout [8–10] and nanoparticle can provide reinforcement, the transduction better ofbiocom the dynamicpatibility, mechanical low cytotoxicity, stimulation and demonstrated required for bone formationantibacterial [11]. properties [1–4]. Furthermore, iron oxide nanoparticles have been shown to promote osteogenicNanocomposite differentiation gels can of be st synthesizedem cells [8–10] by aand variety can provide of methods, the transduction including the of radiolyticthe dynamic method. γ-irradiationmechanical (radiolytic stimulation method) required has for the bone advantage formation of [11]. a pure and homogeneous initiation of the polymer crosslinkingNanocomposite reaction and gels reduction can be of synthesized cations, by asa variety well as theof methods, sterility ofincluding the final product.the radiolytic Typically, method. γ-irradiation (radiolytic method) has the advantage of a pure and homogeneous initiation nanocomposite gels are synthesized by two-step methods: (i) γ-irradiation induced crosslinking of of the polymer crosslinking reaction and reduction of metal cations, as well as the sterility of the final the polymerproduct. Typically, in solution nanocomposite in the presence gels ofar pre-preparede synthesized by nanoparticles two-step methods: (NPs) [(i)12 ,γ13-irradiation] or (ii) in situ γ-irradiationinduced crosslinking synthesis of of nanoparticles the polymer in within solution the in already the presence prepared of pre-prepared polymer gel nanoparticles [14–17]. Of (NPs) particular interest[12,13] is the or (ii) one-step in situ γ-irradiation-irradiation synthesis synthesis of ofnanoparticles nanocomposite within gels. the already However, prepared the one-step polymer synthesis gel has the[14–17]. advantage Of particular of simultaneous interest is the crosslinking one-step γ-irradiation of polymer synthesis chains of with nanocomposite the formation gels. of However, network and reductionthe one-step of metal synthesis salts and has thethe formationadvantage of of simult NPs.aneous The one-step crosslinking synthesis of polymer is faster chains and with simpler the and resultsformation in a small of network NPs size and and reduction narrow of size metal distribution salts and the as formation well as homogeneous of NPs. The one-step distribution synthesis of NPs throughoutis faster the and polymer simpler matrix.and results On thein a other small hand, NPs thesize one-step and narrow synthesis size distribution of nanocomposite as well gelsas has beenhomogeneous poorly studied, distribution as it is di offfi NPscult throughout to find favorable the polymer conditions matrix. for On boththe other nanoparticle hand, the synthesisone-step and synthesis of nanocomposite gels has been poorly studied, as it is difficult to find favorable conditions polymer crosslinking. for both nanoparticle synthesis and polymer crosslinking. γ The The-irradiation γ-irradiation method method is is highly highly suitablesuitable for for th thee synthesis synthesis of ofNPs NPs of controlled of controlled size and size shape and shape in ain solution a solution and and in in heterogeneous heterogeneous media media such such as hy asdrogels. hydrogels. It is also It suit is alsoable for suitable forming for the forming three- the three-dimensionaldimensional polymer polymer network. network. However, However, the thestudies studies are mostly are mostly oriented oriented toward toward the synthesis the synthesis of of metalmetal NPs NPs and and gels gels containing containing metal metal NPs. NPs. The radi radiolyticolytic synthesis synthesis of ofiron iron oxide oxide nanoparticles nanoparticles and and nanocompositenanocomposite gels gels is rarelyis rarely studied studied [ 18[18–28].–28]. One of of the the reasons reasons for for this this is a is very a very complex complex iron oxide iron oxide chemistry,chemistry, which which generates generates numerous numerous iron iron oxideoxide and oxyhydroxide oxyhydroxide polymorphs. polymorphs. Another Another difficulty diffi culty is theis highthe high susceptibility susceptibility of of magnetic magnetic iron iron oxideoxide NPs to to (re)oxidation (re)oxidation of offerrous ferrous ions ions (Fe2+ (Fe) to2 +ferric) to ferric 3+ ionsions (Fe3 +(Fe) under) under atmospheric atmospheric conditions, conditions, especially when when they they are are in inthe the nano-size nano-size range. range. The main The main principle of formation of metal and metal oxide nanoparticles by gamma-irradiation of an aqueous principle of formation of metal and metal oxide nanoparticles by gamma-irradiationn+ of an aqueous solution of a metal salt or within a hydrogel is the reduction of metal cations (M ) byn+ hydrated solution of a metal− salt or− within a hydrogel is the reduction of metal cations (M ) by hydrated electrons (e ) (Eo(H2O/ e ) = −2.87 VSHE) and proton radicals (H·) (Eo(H+/H•) = −2.30 VSHE) which are electrons (e )(aqEo(H O/e )aq= 2.87 V ) and proton radicals (H )(Eo(H+/H ) = 2.30 V ) which aq− 2 aq− − SHE · • − SHE are strongstrong reducingreducing agents agents formed formed on on water water radiolysis radiolysis [29], [29],

− • H2O e , H•, •OH, H2O2, HO2•, H2, O2 −, H3O+. (1) aq (1)

TheseThese reducing reducing species species can can easily easily reduce reduce metal metal cationscations with a a more more positive positive standard standard potential potential to a lowerto a oxidationlower oxidation state state or zero-valence or zero-valence state, state, such such as as ferric ferricions ions toto ferrous (E (Eo(Feo(Fe3+/Fe3+/2+Fe) =2 ++0.77) =+ V)0.77 V) 0 o 3+ 0 o 2+ 0 or evenor even Fe0 ((FeE o(((FeE (Fe3+/Fe/Fe0) )= = −0.040.04 V;V; EEo(Fe 2/Fe+/Fe) 0=) =−0.440.44 V) V)[18,26]. [18,26 Due]. Dueto the to high the highenergy energy and and penetration of γ-radiation, −strong reducing species ar−e formed homogeneously throughout the penetration of γ-radiation, strong reducing species are formed homogeneously throughout the system, system, resulting in homogeneous NPs nuclei formation. On the other hand, hydroxyl radicals (•OH), resulting in homogeneous• NPs•− nuclei formation. On the other hand, hydroxyl radicals (•OH), as well as as well as HO2 and O2 radicals formed on water radiolysis, are strong oxidizing species (VNHE, o o • E HO2•E and( OH/H O2•−2O)radicals = +2.34 V formedNHE). In onorder water to ensure radiolysis, strong are reducing strong conditions, oxidizing speciesirradiation (VNHE is carried, (•OH out/ H2O) = +2.34in deoxygenated VNHE). In order solutions to ensure and strong with the reducing addition conditions, of scavengers irradiation of hydroxyl is carried radicals, out in such deoxygenated as 2- solutionspropanol. and The with formed the addition 2-propanol of scavengersradicals can also of hydroxyl reduce metal radicals, cations such (Eo((CH as 2-propanol.3)2CO/(CH3)2C The•OH formed = o 2-propanol−1.8 VNHE radicals). can also reduce metal cations (E ((CH3)2CO/(CH3)2C•OH = 1.8 VNHE). In our previous works, we synthesized different magnetic iron oxide NPs using− γ-irradiation in the presence of various polymers and (polyvinylpyrrolidone (PVP), cetyltrimethylammonium bromide (CTAB), DEAE-dextran, dextran ) [23–25] and within microemulsion droplets [20,21] in order to control the stability of the synthesized magnetic suspensions and the size and morphology of the nanoparticles. The aim of this work was to explore the ability of γ-irradiation technique for the one-step synthesis of magnetic poly(ethylene oxide)/Fe-

Nanomaterials 2020, 10, 1823 3 of 25

In our previous works, we synthesized different magnetic iron oxide NPs using γ-irradiation in the presence of various polymers and surfactants (polyvinylpyrrolidone (PVP), cetyltrimethylammonium bromide (CTAB), DEAE-dextran, dextran sulfate) [23–25] and within microemulsion droplets [20,21] in order to control the stability of the synthesized magnetic suspensions and the size and morphology of the nanoparticles. The aim of this work was to explore the ability of γ-irradiation technique for the one-step synthesis of magnetic poly(ethylene oxide)/Fe-oxide (PEO/Fe-oxide) nanocomposite hydrogels. PEO was selected because it is a semicrystalline, hydrophilic and biocompatible polymer with numerous applications in pharmacy and biomedicine. For instance, it is used as wound dressings and hydrogels for active substance release. Upon γ-irradiation of PEO aqueous solutions, PEO easily crosslinks and forms macroscopic “wall-to-wall” hydrogels [12,30–33]. In this work, we optimized the experimental conditions and synthesized magnetic PEO/iron oxide nanocomposite hydrogels in one step starting from an alkaline aqueous suspension of PEO and Fe3+ precursor. In addition, the influence of γ-irradiation dose and concentrations of Fe3+ precursor and PEO on the microstructural, thermal and rheological properties of such one-step synthesized magnetic nanocomposite gels was studied.

2. Materials and Methods

2.1. Chemicals for the Synthesis All chemicals were of analytical purity and used as received. Iron(III) chloride hexahydrate (FeCl 6H O) (puriss. p.a., Reag. Ph. Eur., 99%) produced by Sigma-Aldrich, St. Louis, MO, 3· 2 ≥ USA; sodium hydroxide (anhydrous, free-flowing, pellets, ACS reagent, 97%) by Sigma-Aldrich, ≥ St. Louis, MO, USA/Honeywell, Muskegon, MI, USA; 2-propanol (CROMASOLV, for HPLC, 99.9%) ≥ by Sigma-Aldrich, St. Louis, MO, USA/Honeywell, Muskegon, MI, USA; poly(ethylene oxide) (PEO) of viscosity average molecular weight (Mv) 400,000 by Sigma-Aldrich, St. Louis, MO, USA and Mili-Q deionized water were used.

2.2. Synthesis of Samples Iron(III) chloride salt was used for the synthesis of iron oxide nanoparticles. PEO/iron(III) precursor suspensions were prepared by firstly dissolving PEO powder to prepare 1.85 wt% aqueous solutions, 3+ followed by the addition of 2 M FeCl3 aqueous solution to the final concentrations of initial Fe ions in solutions being 0.35 10 2 M, 1.75 10 2 M, and 7 10 2 M. These Fe3+ concentrations in solutions × − × − × − correspond to mass percentages of Fe3+ of 1, 5, 20 wt% relative to total PEO and Fe3+ mass, respectively. In addition, one batch of precursor solutions wasprepared from 4 wt% PEO aqueous solutions with the same concentrations of Fe3+ salt. The solutions were irradiated with or without 2-propanol. The final concentration of 2-propanol in solutions was 0.2 M. In addition, few solutions were prepared containing four times more 2-propanol (0.8 M). The pH of suspensions was adjusted to pH = 11.5–12 with 2 M NaOH aqueous solution. The prepared precursor suspensions were bubbled with nitrogen through rubber septa for 30 min in order to remove dissolved before γ-irradiation. γ-irradiation of deoxygenated suspensions in septum-closed glass vials was performed at room temperature in a 60Co γ-irradiation facility located in the Radiation Chemistry and Dosimetry Laboratory at the Ruder¯ Boškovi´cInstitute. The suspensions were irradiated to doses of 50, 130 and 300 kGy and at a dose 1 rate of ~27 kGy h− . Irradiation of PEO/Fe(III) precursor suspensions resulted in the formation of gels or suspensions. The schematic presentation of the synthesis procedure of magnetic PEO/iron oxide nanocomposite hydrogels is given in Figure1. Nanomaterials 2020, 10, 1823 4 of 25 Nanomaterials 2019, 9, x FOR PEER REVIEW 4 of 29

Figure 1. Schematic presentation of the synthesis procedure of magnetic poly(ethylene oxide) (PEO)/iron Figure 1. Schematic presentation of the synthesis procedure of magnetic poly(ethylene oxide) oxide nanocomposite hydrogels. (PEO)/iron oxide nanocomposite hydrogels. 2.3. Characterization of Samples 2.3. Characterization of Samples Synthesized samples were characterized as as-synthesized gels (rheological measurements), as driedSynthesized gels (DSC, samples XRD, Mössbauer, were characterized SEM) or asas suspensionsas-synthesized/isolated gels (rheological precipitates (XRD,measurements), Mössbauer). as Precipitatesdried gels (DSC, were XRD, isolated Mössbauer, from suspensions SEM) or as by suspensions/isolated centrifugation, followed precipitates by washing (XRD, with Mössbauer). ethanol. ScanspeedPrecipitates2236R were high-speed isolated from centrifuge suspensions was used. by centrifugation, The obtained followed gels and isolatedby washing precipitates with ethanol. were driedScanspeed in vacuum 2236R athigh-speed room temperature, centrifuge and was then used. characterized. The obtained gels and isolated precipitates were driedThe in vacuum morphology at room of samplestemperature, was investigatedand then characterized. using a probe Cs-corrected cold field-emission ScanningThe morphology Transmission of Electron samples Microscopewas investigated (STEM, us modeling a probe ARM Cs-corrected 200 CF), and cold the field-emission thermal field emissionScanning scanning Transmission electron Electron microscope Microscope (FE SEM, (STEM, model JSM-7000F)model ARM manufactured 200 CF), and by the JEOL thermal Ltd., Tokyo, field Japan,emission FE SEMscanning was linkedelectron to microscope the EDS/INCA (FE 350 SEM, (energy-dispersive model JSM-7000F) X-ray manufactured analyzer) manufactured by JEOL Ltd., by OxfordTokyo, InstrumentsJapan, FE SEM Ltd., Abingdon,was linked UK. to the EDS/INCA 350 (energy-dispersive X-ray analyzer) manufacturedX-ray powder by Oxford diffraction Instruments (XRD) patterns Ltd., Abingdon, were recorded UK. at room temperature using an APD 2000 X-rayX-ray powder powder diffractometer diffraction (Cu (XRD)Kα radiation, patterns graphite were reco monochromator,rded at room temperature NaI–Tl detector) using manufactured an APD 2000 byX-rayItalStructures powder, Rivadiffractometer Del Garda, Italy.(CuK Theα radiation, XRD patterns graphite were recorded monochromator, over the 15–80 NaI–Tl◦ 2θ range detector) with amanufactured 2θ step of 0.05–0.025 by ItalStructures◦ and a counting, Riva Del time Garda, perstep Italy. of The 15–80 XRD s. patterns were recorded over the 15– 57 80° 2θ Ferange Mössbauer with a 2θ spectra step of were0.05–0.025° recorded and at a counting 20 ◦C in time the transmission per step of 15–80 mode s. using a standard instrumental57Fe Mössbauer configuration spectra by were WissEl recorded GmbH, at Mömbris-Hohl, 20 °C in the transmission Germany. The mode57Co using in the a rhodiumstandard 57 matrixinstrumental was used configuration as a Mössbauer by WissEl source. GmbH, The spectrometer Mömbris-Hohl, was calibrated Germany. at The 20 ◦ CCo using in the standardrhodium αmatrix-Fe foil was spectrum. used as a The Mössbauer velocity source. scale and The all spectrom the dataeter refer was to calibrated the metallic at 20α -Fe°C using absorber the standard at 20 ◦C. Theα-Fe experimentally foil spectrum. observedThe velocity Mössbauer scale and spectra all the were data fitted refer using to the the metallic MossWinn α-Fe program. absorber Additionally,at 20 °C. The 57experimentallyFe Mössbauer observed spectra were Mössbauer recorded spectra at 77 K were using fitted a conventional using the MossWinn constant acceleration program. Additionally, transmission spectrometer57Fe Mössbauer with spectra a 57Co(Rh) were recorded source andat 77 a K bath using cryostat. a conventional The spectra constant obtained acceleration at 77 K transmission were fitted usingspectrometer the MOSFIT with programa 57Co(Rh) (Teillet, source J.; Varret,and a bath F. unpublished cryostat. The MOSFIT spectra program, obtained Universit at 77 Ké weredu Maine) fitted andusing an theα-Fe MOSFIT absorber program was used (Teillet, as a calibration J.; Varret, sample. F. unpublished MOSFIT program, Université du Maine)Diff anderential an α scanning-Fe absorber calorimetry was used (DSC) as a thermogramscalibration sample. were recorded using PerkinElmer, Waltham, MA, USADifferential Diamond scanning DSC calorimeter, calorimetry calibrated (DSC) withthermograms In and Zn standardswere recorded and operating using PerkinElmer, in a dynamic mode.Waltham, Samples MA, USA of dried Diamond gel (5–10 DSC mg) calorimeter, were sealed calibrated into Al pans.with In Two and heating Zn standards and cooling and operating cycles at temperaturesin a dynamic mode. ranging Samples from of40 dried◦C to gel 100 (5–10◦C in mg) an extrawere puresealed nitrogen into Al pans. environment Two heating were and performed cooling − 1 forcycles each at sampletemperatures at a rate ranging of 10 ◦ Cfrom min −−40. The°C to first 100 heating °C in an cycle extra was pure performed nitrogen inenvironment a range 22 ◦wereC to −1 100performed◦C. For eachfor each synthesized sample at gel a rate three of specimens 10 °C min were. The recorded. first heating The cycle temperatures was performed and enthalpies in a range of 22 °C to 100 °C. For each synthesized gel three specimens were recorded. The temperatures and enthalpies of melting and crystallization were determined from the second heating and first cooling cycles, and their averages are presented.

Nanomaterials 2020, 10, 1823 5 of 25 melting and crystallization were determined from the second heating and first cooling cycles, and their averages are presented. The mechanical properties of gels are described using oscillatory rheology. The storage (G0) and loss (G”) moduli of the nanocomposite gels were determined with a mechanical spectrometer (Anton Paar MCR 302, Stuttgart, Germany), using a steel plate plate geometry (PP25, Anton Paar, Graz, − Austria) equipped with a true-gap system and the data were collected using RheoCompass software. The sample temperature was controlled through a Peltier temperature control located on the base of the geometry and with a Peltier-controlled hood (H-PTD 200). A piece of a gel sample (1 mm thick slice) was placed on the base plate of the rheometer, and the plate was set using the true-gap function of the software. Thus, after 5 min at 25 ◦C, the G0 and G” moduli were measured always within the linear viscoelastic region (LVR). After 5 min at 25 ◦C, the yield stress of the gels was determined by applying a strain (γ) sweep between 0.01% and 100%. Rheological properties of the gel material are independent of strain up to yield strain, and beyond yield strain the rheological behavior is nonlinear. Three interval thixotropy test is a standard test which allows tracking of material response resulting from stepwise changes in shear strain making it the most appropriate method for structure recovery tests. In the thixotropic experiments, rheological measurements were conducted on gels at 25 ◦C under initial conditions at which they were in their linear viscoelastic regimes (a strain of 0.1% and angular frequency of 5 rad/s) for 680 s to establish baseline values for G0 and G”. In studies with gels viscoelastic recovery was observed after the cessation of destructive strain. Frequency sweeps (0.05–100 rad/s) were subsequently performed at 25 ◦C at a strain value within LVR to investigate the time-dependent deformation behavior of gels.

2.4. Quantitative Determination of Fe2+ Using 1,10-Phenanthroline UV-Vis Spectrophotometric Method

2.4.1. Chemicals for Spectrophotometric Determination All chemicals were of analytical purity and used as received. Sodium acetate (Merck, Kenilworth, NJ, USA anhydrous for analysis, EMSURE, ACS, Reag. Ph. Eur.), acetic acid (Honeywell, Muskegon, MI, USA puriss.p.a., ACS Reagent, Reag. Ph. Eur., 99.8%), L-ascorbic acid (Sigma-Aldrich, St. Louis, MO, ≥ USA BioXtra, crystalline, 99.0%), 1,10-phenanthroline monohydrate (Sigma-Aldrich, St. Louis, MO, ≥ USA for the spectrophotometric determination, 99.0%), and hydrochloric acid (Fluka (Honeywell), ≥ Muskegon, MI, USA for trace analysis, fuming, 37%) were used. ≥ 2.4.2. Spectrophotometric Determination of Fe2+ The amount of Fe2+ generated upon γ-irradiation was determined using the 1,10-phenanthroline method [34,35]. Immediately after irradiation the samples were acidified (pH 1) by the addition of concentrated ≤ hydrochloric acid (~2.5 vol%) using a syringe through rubber septa. At such low pH, formed iron oxide nanoparticles are dissolved and all Fe2+ formed upon γ-irradiation is preserved from oxidation when the vial is opened. The detailed procedure for the determination of Fe2+ and total iron in such acidified solutions is given in our recently published paper [35].

3. Results and Discussion

3.1. Microstructural Characterization of Gels (and Suspensions) Upon γ-irradiation, the reddish PEO/Fe(III) suspensions turned to reddish or white-green or black gels or black suspensions depending on the dose, 2-propanol concentration and the amount of Fe3+. The photographs of formed nanocomposite gels are shown in Figure2. Nanomaterials 2020, 10, 1823 6 of 25

Nanomaterials 2019, 9, x FOR PEER REVIEW 6 of 29

FigureFigure 2. The2. The photographs photographs of PEOof PEO/Fe-oxide/Fe-oxide nanocomposite nanocompos gelsite gels obtained obtained from from suspension suspension with variouswith Fe3various+ content Fe3+ at content various at doses. various Unless doses. otherwise Unless otherwis indicated,e indicated, the precursor the precursor suspensions suspensions were prepared were fromprepared 1.85 wt% from PEO 1.85 solutions wt% PEO and solutions with the and addition with the of addition 2-propanol of 2-propanol (0.2 M). (0.2 M).

γ-irradiationγ-irradiation of of pure pure PEO PEO solutions solutions (1.85(1.85 wt%) and and PEO/Fe(III) PEO/Fe(III) precursor precursor suspensions suspensions at atpH pH ~ ~ 1212 without without the the addition addition of of 2-propanol 2-propanol resultedresulted in the formation formation of of a apermanent permanent shape shape “wall-to-wall” “wall-to-wall” macroscopicmacroscopic gels. gels. The The resulting resulting nanocomposite nanocomposite gels gels were were reddish-brown reddish-brown and non-magnetic. and non-magnetic. The Thereddish reddish color color and and non-magnetic non-magnetic behavior behavior indicated indicated that thateven evenat the at highest the highest dose of dose 300 of kGy 300 no kGy nosignificant significant reduction reduction of ofFe(III) Fe(III) occurred occurred (Figure (Figure 2). 2The). The reducing reducing conditions conditions were wereimproved improved by the by theaddition addition of of2-propanol 2-propanol (0.2 (0.2 M). M). Irradiation Irradiation of th ofe thepure pure 1.85 1.85 wt% wt% PEO PEO solution solution in the in presence the presence of 2- of 2-propanolpropanol diddid not lead to gel formation. On On the the other other hand, hand, the the irradiation irradiation of ofPEO/Fe(III) PEO/Fe(III) precursor precursor suspensionssuspensions at at pH~12 pH~12 inin thethe presencepresence ofof 2-propanol2-propanol led led to to the the formation formation of of nanocomposite nanocomposite gels gels (Figure 2). The white-green gels were obtained from 1 wt% Fe3+ suspensions, while the black magnetic (Figure2). The white-green gels were obtained from 1 wt% Fe 3+ suspensions, while the black magnetic hydrogels were obtained upon irradiation of 5 and 20 wt% Fe3+ suspensions at higher doses (130 and hydrogels were obtained upon irradiation of 5 and 20 wt% Fe3+ suspensions at higher doses (130 and 300 kGy). The gel obtained from 20% Fe3+ suspensions at 300 kGy was strongly attracted by a 300 kGy). The gel obtained from 20% Fe3+ suspensions at 300 kGy was strongly attracted by a permanent magnet as shown in Figure 3. permanent magnet as shown in Figure3. Figure4 shows FE SEM images of composite gels obtained upon irradiation of suspensions with 2-propanol. The nanoparticles embedded into the polymer matrix are visible. Particles were mainly spherical in shape. The gel at 300 kGy and 20 wt% Fe3+ (Figure4f) consisted of numerous spherical particles and/or particle aggregates of 40 nm in size homogeneously distributed throughout the polymer matrix. The SEM image of gel at 300 kGy and 5 wt% Fe3+ (Figure4e) reveals the presence of larger irregular plate-like particles 70 nm in size, in addition to much smaller spherical ones.

Nanomaterials 2020, 10, 1823 7 of 25 Nanomaterials 2019, 9, x FOR PEER REVIEW 7 of 29

NanomaterialsFigure 2019 3. The, 9, x photographsFOR PEER REVIEW of magnetic PEOPEO/iron/iron oxideoxide gelgel attractedattracted byby aa permanentpermanent magnet.magnet.8 of 29

Figure 4 shows FE SEM images of composite gels obtained upon irradiation of suspensions with 2-propanol. The nanoparticles embedded into the polymer matrix are visible. Particles were mainly spherical in shape. The gel at 300 kGy and 20 wt% Fe3+ (Figure 4f) consisted of numerous spherical particles and/or particle aggregates of 40 nm in size homogeneously distributed throughout the polymer matrix. The SEM image of gel at 300 kGy and 5 wt% Fe3+ (Figure 4e) reveals the presence of larger irregular plate-like particles 70 nm in size, in addition to much smaller spherical ones.

FigureFigure 4. SEM 4. SEM micrographs micrographs of of PEO PEO/Fe-oxide/Fe-oxide gels ob obtainedtained from from suspensions suspensions with with various various Fe3+ content Fe3+ content at variousat various doses: doses: (a ()a) 5 5 wt%wt% Fe 3+3+ atat 5050 kGy; kGy; (b) (20b )wt% 20 wt%Fe3+ at Fe 503+ kGy;at 50 (c) kGy; 5 wt% ( cFe) 53+ wt%at 130 FekGy;3+ at(d)130 20 kGy; (d) 20wt% wt% Fe3+ Fe at 3130+ at kGy; 130 (e kGy;) 5 wt% (e) Fe 53+ wt% at 300 Fe kGy;3+ at (f 300) 20 wt% kGy; Fe (f3+) at 20 300 wt% kGy. Fe All3+ precursorat 300 kGy. suspensions All precursor suspensionswere prepared were from prepared 1.85 wt% from PEO 1.85 so wt%lutions PEO and solutions with 0.2 M and 2-propanol. with 0.2 M 2-propanol.

Figure 5 shows the Mössbauer spectra recorded at 77 K of nanocomposite gels obtained from suspensions with 2-propanol at 130 kGy and 300 kGy. Generally, the Mössbauer spectrometry is very sensitive to the presence of Fe(II) in inorganic [25,36] and biological samples [37]. The sample 5% Fe3+ at 130 kGy (Figure 5a) consisted of a symmetric doublet which was fitted with quadrupole splitting distribution. Even at this low-temperature the sample was not magnetically ordered, which reveals the ultrasmall particle size. This quadrupole splitting distribution can be ascribed to a poorly crystalline, disordered structure (average quadrupole splitting <Δ> for 5% Fe sample is 0.79 mm s−1 (Figure 5a) and for 20% Fe sample is 0.82 mm s−1 (Figure 5b)). Average isomer shift values of <δ> = 0.45 mm s−1 for both samples are consistent with those of Fe3+ [38]. The spectrum of gel obtained at 300 kGy and 5% Fe3+ consisted of a quadrupole splitting distribution component and the doublet

Nanomaterials 2020, 10, 1823 8 of 25

Figure5 shows the Mössbauer spectra recorded at 77 K of nanocomposite gels obtained from suspensions with 2-propanol at 130 kGy and 300 kGy. Generally, the Mössbauer spectrometry is very sensitive to the presence of Fe(II) in inorganic [25,36] and biological samples [37]. The sample 5% Fe3+ at 130 kGy (Figure5a) consisted of a symmetric doublet which was fitted with quadrupole splitting distribution. Even at this low-temperature the sample was not magnetically ordered, which reveals the ultrasmall particle size. This quadrupole splitting distribution can be ascribed to a poorly crystalline, 1 disordered structure (average quadrupole splitting <∆> for 5% Fe sample is 0.79 mm s− (Figure5a) 1 1 and for 20% Fe sample is 0.82 mm s− (Figure5b)). Average isomer shift values of <δ> = 0.45 mm s− for both samples are consistent with those of Fe3+ [38]. The spectrum of gel obtained at 300 kGy Nanomaterialsand 5% Fe 20193+ consisted, 9, x FOR PEER of a REVIEW quadrupole splitting distribution component and the doublet whose9 of 29 2+ -1 1 parameters correspond to Fe (δ = 1.24 mm s , ∆ = 2.68 mm s− ) (Figure5c). The gel obtained at 2+ -1 −1 whose300 kGy parameters and 20% Fecorrespond3+ exhibited to Fe a collapsing(δ = 1.24 mm sextet s (Figure, Δ = 2.685d). mm Such s ) spectrum(Figure 5c). is The generally gel obtained found 3+ atwith 300 systems kGy and that 20% exhibit Fe exhibited superparamagnetic a collapsing relaxation sextet (Figure phenomena. 5d). Such This spectrum spectrum is generally was best found fitted with systems that exhibit superparamagnetic relaxation phenomena. This spectrum was best fitted with a distribution of the hyperfine magnetic field (average was 23.6 T). The fit resulted in a withbimodal a distribution distribution, of whichthe hyperfine indicates magnetic that the samplefield (average consists < ofBhf two> was types 23.6 of T). particles; The fit resulted a population in a bimodalwith the distribution, bigger particle which size indicates whose magnetic that the sample relaxation consists time of is two longer types than of theparticles; measurement a population time with57 the bigger particle size whose magnetic8 relaxation time is longer than the measurement time of of Fe Mössbauer spectrometry (5 10− s), and a population of smaller particles whose relaxation 57 ×−8 timeFe Mössbauer is somewhat spectrometry shorter than (5 × the 10 measurement s), and a population time of Mössbauerof smaller particles spectrometry. whose relaxation The symmetric time isnature somewhat of the shorter spectrum than suggests the measurement that there istime no of Fe Mössbauer2+ in the sample. spectrometry. Furthermore, The symmetric the fact that nature the of the spectrum suggests that there1 is no Fe2+ in the sample. Furthermore,1 the fact that the isomer shift isomer shift value is 0.45 mm s− and quadrupole shift is 0.00 mm s− , which are the usual values for −1 −1 valuepoorly is crystalline 0.45 mm maghemites and quadrupole sextet at 77shift K [ 39is ],0.00 indicates mm s that, which the sample are the is composedusual values of maghemite.for poorly crystalline maghemite sextet at 77 K [39], indicates that the sample is composed of maghemite. The The spectra measured at 300 K are shown in Figure S1 in the Supplementary Materials (all spectra at spectra measured at 300 K are shown in Figure S1 in the Supplementary Materials (all spectra at room room temperature are consistent with superparamagnetic and/or paramagnetic doublets). temperature are consistent with superparamagnetic and/or paramagnetic doublets).

3.00x105 3.20x105

5 3.15x10 2.90x105 δ -1 -1 < > = 0.45 mm s 5 <δ > = 0.45 mm s 3.10x10 Δ -1 -1 < > = 0.82 mm s <Δ > = 0.79 mm s 2.80x105 3.05x105

5 3.00x10 5 10 12 2.70x10 5 10 8

2.95x10 8 6 5 5 6

Counts Counts 2.60x10 2.90x10 4 4

5 2 2

2.85x10 / % Probability Probability / % / Probability 5 0 2.50x10 0 5 -2 2.80x10 0.00.51.01.52.02.5 0.0 0.5 1.0 1.5 2.0 2.5 Quadrupolar splitting / mm s-1 (a) Quadrupolar splitting / mm s-1 (b) 2.75x105 2.40x105 -4 -2 0 2 4 -4 -2 0 2 4 Doppler velocity / mm s-1 Doppler velocity / mm s-1 5.00x105 1.28x106 4.90x105 A = 92% 6 5 A = 8% 1.26x10 4.80x10 -1 <δ > = 0.45 mm s δ = 1.24 mm s-1 -1 -1 5 <Δ > = 0.74 mm s Δ 6 -1 4.70x10 = 2.68 mm s 1.24x10 <δ > = 0.45 mm s -1 Γ = 0.39 mm s ε -1 10 10 <2 > = 0.00 mm s 5 4.60x10 1.22x106 = 23.6 T 8 8 hf

5 6 6

Counts 6 4.50x10 Counts 1.20x10 4 4 5

2 Probability / % 4.40x10 2 Probability/ % 1.18x106 0 0 5 0 102030405060 4.30x10 0.0 0.5 1.0 1.5 2.0 (d) -1 (c) Quadrupolar splitting / mm s Hyperfine magnetic field / T 1.16x106 -4 -2 0 2 4 -10 -5 0 5 10 -1 -1 Doppler velocity / mm s Doppler velocity / mm s

FigureFigure 5. TheThe 77 K Mössbauer spectra of PEO/Fe-oxide PEO/Fe-oxide gels obtained from suspensions with: 5 5 wt% wt% FeFe3+3+ atat 130 130 kGy kGy ( (aa);); 20 20 wt% wt% Fe Fe3+3+ atat 130 130 kGy kGy ( (bb);); 5 5 wt% wt% Fe Fe3+3+ atat 300 300 kGy kGy (c (c);); 20 20 wt% wt% Fe Fe3+3+ atat 300 300 kGy kGy ( (dd).). All precursor suspensions were were prepared prepared from from 1. 1.8585 wt% wt% PEO PEO solutions solutions and and with with 0.2 0.2 M M 2-propanol. 2-propanol. MössbauerMössbauer parameters parameters are are given: δδ = isomerisomer shift shift relative relative to to αα-Fe-Fe at at 20 20 °C;◦C; Δ∆ = quadrupolequadrupole splitting; splitting; ΓΓ = lineline width; width; BBhfhf == hyperfinehyperfine field, field, A A == relativerelative area. area. Δ∆ valuesvalues are are given given as as an an average average value value of quadrupole splitting distributions ( aa––cc).). BBhfhf valuevalue is is given given as as an an aver averageage value value of of hyperfine hyperfine field field distributiondistribution ( (dd).). Line Line width width values values were were fixed fixed during during the the fitting fitting of of quadrupole splitting distribution 1 1 ((a–c)) or or hyperfine hyperfine field field distribution (d). Error: Error: δδ == ± 0.010.01 mm mm s− s1; Δ; ∆= ±= 0.010.01 mm mm s−1; s Bhf; =B ± 0.3= T.0.3 T. ± − ± − hf ±

Figure 6 presents the XRD patterns of unirradiated precursor (Figure 6a) and PEO/Fe-oxide gels obtained upon irradiation of suspensions with 2-propanol (0.2 M) and 5 wt% Fe3+ at 50 kGy (Figure 6b), 130 kGy (Figure 6c) and 300 kGy (Figure 6d), as well as 20 wt% Fe3+ at 130 kGy (Figure 6e) and 300 kGy (Figure 6f). The XRD patterns (Figure 6a–e) show two maxima which can be attributed to iron oxide phases. The XRD patterns of the precursor and the gel obtained at 50 kGy (Figure 6a,b) were ascribed to ferrihydrite and NaCl as an impurity. The XRD patterns of gels obtained at 130 and 300 kGy (Figure 6a–e) were attributed to magnetite NPs. On the other hand, the gel obtained from 20 wt% Fe3+ suspension at 300 kGy (Figure 6f) had sharper and more distinct maxima, indicating improved crystallinity of the formed magnetite NPs, which is in line with the results of Mössbauer spectrometry. The distinct maxima at ~19 and ~23° on the XRD patterns of composite gels obtained at 50 and 130 kGy are the result of partial crystallization of PEO gels upon drying. These maxima completely disappeared on the XRD pattern of nanocomposite gel obtained at 300 kGy and only wide

Nanomaterials 2020, 10, 1823 9 of 25

Figure6 presents the XRD patterns of unirradiated precursor (Figure6a) and PEO /Fe-oxide gels obtained upon irradiation of suspensions with 2-propanol (0.2 M) and 5 wt% Fe3+ at 50 kGy (Figure6b), 130 kGy (Figure6c) and 300 kGy (Figure6d), as well as 20 wt% Fe 3+ at 130 kGy (Figure6e) and 300 kGy (Figure6f). The XRD patterns (Figure6a–e) show two maxima which can be attributed to iron oxide phases. The XRD patterns of the precursor and the gel obtained at 50 kGy (Figure6a,b) were ascribed to ferrihydrite and NaCl as an impurity. The XRD patterns of gels obtained at 130 and 300 kGy (Figure6a–e) were attributed to magnetite NPs. On the other hand, the gel obtained from 20 wt% Fe3+ suspension at 300 kGy (Figure6f) had sharper and more distinct maxima, indicating improved crystallinity of the formed magnetite NPs, which is in line with the results of Mössbauer spectrometry. The distinct maxima at ~19 and ~23◦ on the XRD patterns of composite gels obtained at 50 and 130 kGy are the result of partial crystallization of PEO gels upon drying. These maxima completely disappeared on the XRD pattern of nanocomposite gel obtained at 300 kGy and only wide amorphous halo is visible, due to the very dense crosslinking density of gel obtained at 300 kGy. The results of line broadening analyses are given in Table1. The volume-averaged domain sizes ( Dv) of the dominant crystalline phase in the synthesized samples were estimated using the Scherrer equation:

D = 0.9λ/(β cosθ) (2) hkl hkl × where Dhkl is the volume average domain size in the direction normal to the reflecting planes (hkl), λ is the X-ray wavelength (CuKα), θ is the Bragg angle, and βhkl is the pure full width of the diffraction line (hkl) at half the maximum intensity. The volume-average domain size (Dv) of the 110 lines (D110) of ferrihydrite in the unirradiated precursor was estimated to 1.7, whereas upon γ-irradiation to 50 kGy the average domain size increased to 4.5 nm (Table1). At 130 kGy and 300 kGy (5 wt% Fe 3+) ferrihydrite transformed to magnetite of about 2.3 nm in size (D311  2.3 nm). The sample irradiated 3+ with the dose of 130 kGy with 20 wt% Fe had a somewhat larger crystallite size (D311  3.3 nm) than the sample prepared from suspension with lower Fe3+ concentration, which is expected. However, at 300 kGy and 20 wt% Fe3+ the crystallite size of magnetite nanoparticles increased significantly when compared to the 130 kGy sample. This significant increase in the crystallinity of nanoparticles was not observed for gels prepared from suspensions lower initial Fe3+ concentration where the magnetite crystallite size did not change noticeably. TEM analysis of gel obtained at 130 kGy from suspension with 20 wt% Fe3+ is shown in Figure7. Figure7a shows the STEM bright-field micrograph of the gel with slightly high-frequency FFT filtered BF-STEM image of very thin area (where individual Fe atom columns can bee seen in some nanoparticles). The very small (~3 nm) particles can be seen confirming the results obtained by XRD line-broadening analysis. The much larger particle size observed using SEM in the scattering mode (Figure4) compared to the TEM determination in the transmission mode (Figure7a) is because the SEM sees the iron oxide nanoparticles “disguised” by polymer whereas the TEM can see the “pure” individual iron oxide nanoparticles. EDXS (energy-dispersive X-ray spectroscopy) elemental mapping (Figure7b) shows that three major elements are iron, oxygen, and carbon, and that they are homogeneously distributed throughout the sample indicating good of nanoparticles within the gel matrix. SAED (selected area electron diffraction) patterns (Figure7c,d) match magnetite and NaCl impurity thus confirming the results of XRD, i.e., the formation of magnetite nanoparticles inside the PEO gel. The Mössbauer results suggested that magnetite nanoparticles were completely oxidized to maghemite. This discrepancy arises because the SAED measures the sample in high vacuum under reducing conditions. On the contrary, Mössbauer spectrometry is very sensitive to Fe(II) at the ambient conditions, and as a rule the Mössbauer spectrometry for small nanoparticles below 5 nm shows no Fe(II). This is because very small magnetite nanoparticles can easily be oxidized in air. The amount of Fe2+ generated upon γ-irradiation was quantitatively determined using 1,10-phenanthroline UV-Vis spectrophotometric method. Figure8 shows the Fe 2+ fraction ([Fe2+/(Fe2++Fe3+)]) in the γ-irradiated samples prior to their isolation and coming in contact with air. This was obtained by the addition of concentrated HCl immediately after irradiation through Nanomaterials 2019, 9, x FOR PEER REVIEW 10 of 29

amorphous halo is visible, due to the very dense crosslinking density of gel obtained at 300 kGy. The results of line broadening analyses are given in Table 1. The volume-averaged domain sizes (Dv) of the dominant crystalline phase in the synthesized samples were estimated using the Scherrer Nanomaterials 2020, 10, 1823 10 of 25 equation:

Dhkl = 0.9λ/(βhkl × cosθ) (2) rubber septa (described in experimental). The Fe2+ molar fraction depended on dose, pH and initial concentrationwhere Dhkl is the of Fe volume3+ in precursor average domain suspensions. size in The the Fedirection2+ fraction normal increased to the reflecting with the increase planes (hkl of dose), λ andis the pH X-ray and with wavelength the lower (Cu initialKα), Feθ is3+ theconcentration. Bragg angle,At and 130 βhkl kGy is the and pure initial full 5width wt% Feof 3the+ the diffraction reduction yieldline was(hkl) 34%at half in the comparison maximum to intensity. 22% ingels The obtainedvolume-average from 20 domain wt% Fe size3+ suspension. (Dv) of the 110 The lines reduction (D110) yieldof ferrihydrite of 31% obtained in the unirradiated for the sample precur at 300sor kGywas fromestimated 20 wt% to 1.7, Fe3 whereas+ suspension upon resultedγ-irradiation in a highlyto 50 3+ magnetickGy the gel.average The highestdomain reduction size increased yield ofto Fe4.52+ nm(54.7%) (Table was 1). achievedAt 130 kGy at the and dose 300 of kGy 300 (5 kGy, wt%pH Fe= 12) ≅ forferrihydrite 5% Fe3+ precursor transformed suspension. to magnetite It shouldof about be 2.3 noted nm in that size at (D similar311 2.3 experimentalnm). The sample conditions irradiated the 3+ ≅ γ-irradiationwith the dose in of the 130 presence kGy with of 20 dextran wt% Fe sulfate had a or somewhat DEAE-dextran larger polymercrystallite generated size (D311 almost 3.3 nm) 100% than of 3+ Fethe2+ atsample 130 kGy prepared [25,35]. from More suspension reducing conditionswith lower were Fe concentration, obtained at higher which pH. is Thisexpected. can be However, explained at by 300 kGy and 20 wt% Fe3+ the crystallite size of magnetite nanoparticles increased significantly when the higher yield of hydrated electrons in a highly alkaline medium [19,40–43]. In an alkaline medium, compared to the 130 kGy sample. This significant increase in the crystallinity of nanoparticles was the hydrogen atoms are converted to hydrated electrons. Because of the fact that the higher reducing not observed for gels prepared from suspensions lower initial Fe3+ concentration where the magnetite conditions were obtained at higher pH, all investigated gels were synthesized at pH ~ 12. crystallite size did not change noticeably.

1200 5000 Ferrihydrite Card No. 29-0712 Ferrihydrite Card No. 29-0712 (a) NaCl Card No. 5-0628 (b) NaCl Card No. 5-0628 1000 4000

800 3000 • 600 2000 • Counts 400 • • Counts

1000 200

0 0 20 30 40 50 60 70 80 20 30 40 50 60 70 80 2θ / ° (CuKα) 2θ / ° (CuKα)

Magnetite Card No. 19-0629 Magnetite Card No. 19-0629 3000 (c) NaCl Card No. 5-0628 (d) NaCl Card No. 5-0628 2000 2500

2000

• 1500 • • 1000 • Counts Counts 1000

500

0 0 20 30 40 50 60 70 80 20 30 40 50 60 70 80 2θ / ° (CuKα) 2θ / ° (CuKα) 5000 3500 (e) Magnetite Card No. 19-0629 (f) Magnetite Card No. 19-0629 4000 NaCl Card No. 5-0628 3000 NaCl Card No. 5-0628 2500 3000 2000

2000 1500 Counts Counts 1000 1000 • • 500 • •

0 0 20 30 40 50 60 70 80 20 30 40 50 60 70 80 θ α 2 / ° (CuK ) 2θ / ° (CuKα)

Figure 6. XRD patterns of unirradiated precursor (a) and PEO/Fe-oxide gels obtained from suspensions with: 5 wt% Fe3+ at 50 kGy (b), 5 wt% Fe3+ at 130 kGy (c), 5 wt% Fe3+ at 300 kGy (d) 20 wt% Fe3+ at 130 kGy (e) and 20 wt% Fe3+ at 300 kGy (f), at pH ~ 12. The two maxima at ~19 and ~23 degrees marked with red dots correspond to crystalline PEO. Nanomaterials 2020, 10, 1823 11 of 25

Table 1. The volume-averaged domain size (Dv) of the dominant crystalline phase of selected samples.

Sample Phase hkl 2θ/◦ FWHM/◦ Dhkl/nm unirradiated precursor (5 wt% ferrihydrite 110 ~35 5.1 1.7 Fe3+, 0.2 M 2-propanol) PEO + halite (impurity) gel - 50 kGy (5 wt% Fe3+, 0.2 M ferrihydrite 110 35.8 1.9 4.5 2-propanol) PEO + halite (impurity) gel - 130 kGy (5 wt% Fe3+, 0.2 M magnetite 311 35.5 3.6 2.3 2-propanol) PEO + halite (impurity) gel - 300 kGy (5 wt% Fe3+, 0.2 M magnetite - 35.3 3.5 2.4 2-propanol) halite + unidentified impurity gel - 130 kGy (20 wt% Fe3+, 0.2 M magnetite 311 35.5 2.6 3.3 2-propanol) PEO + halite (impurity) gel - 300 kGy (20 wt% Fe3+, 0.2 M magnetite 311 35.4 0.6 13.9 2-propanol) PEO + halite (impurity) powder - 50 kGy (20 wt% Fe3+, magnetite 311 35.5 3.6 2.3 0.8 M 2-propanol) powder - 130 kGy (20 wt% Fe3+, magnetite 311 35.5 2.6 3.3 0.8 M 2-propanol) Nanomaterials 2019, 9, x FOR PEER REVIEW 12 of 29

Figure 7. STEMFigure bright-field7. STEM bright-field image image of theof the gel gel obtained obtained at at 130 130 kGy kGy from from 20 wt% 20 Fe wt%3+ suspension Fe3+ suspension (a); (a); EDXS elemental mapping of the gel where yellow, blue, and red colors represent iron (Fe), oxygen EDXS elemental mapping of the gel where yellow, blue, and red colors represent iron (Fe), oxygen (O), (O), and carbon (C), respectively (b); SAED (selected area electron diffraction) of spherical and carbonnanoparticles (C), respectively (NPs) with (b); marked SAED interplanar (selected distances area electron (c); SAED diff ofraction) spherical of NPs spherical indexednanoparticles as (NPs) withmagnetite marked interplanar(Fe3O4) (d). distances (c); SAED of spherical NPs indexed as magnetite (Fe3O4)(d).

The amount of Fe2+ generated upon γ-irradiation was quantitatively determined using 1,10- phenanthroline UV-Vis spectrophotometric method. Figure 8 shows the Fe2+ fraction ([Fe2+/(Fe2++Fe3+)]) in the γ-irradiated samples prior to their isolation and coming in contact with air. This was obtained by the addition of concentrated HCl immediately after irradiation through rubber septa (described in experimental). The Fe2+ molar fraction depended on dose, pH and initial concentration of Fe3+ in precursor suspensions. The Fe2+ fraction increased with the increase of dose and pH and with the lower initial Fe3+ concentration. At 130 kGy and initial 5 wt% Fe3+ the reduction yield was 34% in comparison to 22% in gels obtained from 20 wt% Fe3+ suspension. The reduction yield of 31% obtained for the sample at 300 kGy from 20 wt% Fe3+ suspension resulted in a highly magnetic gel. The highest reduction yield of Fe2+ (54.7%) was achieved at the dose of 300 kGy, pH = 12 for 5% Fe3+ precursor suspension. It should be noted that at similar experimental conditions the γ- irradiation in the presence of dextran sulfate or DEAE-dextran polymer generated almost 100% of Fe2+ at 130 kGy [25,35]. More reducing conditions were obtained at higher pH. This can be explained by the higher yield of hydrated electrons in a highly alkaline medium [19,40–43]. In an alkaline

Nanomaterials 2019, 9, x FOR PEER REVIEW 13 of 29

medium, the hydrogen atoms are converted to hydrated electrons. Because of the fact that the higher Nanomaterialsreducing 2020conditions, 10, 1823 were obtained at higher pH, all investigated gels were synthesized at pH12 ~ of 12. 25

0.6 5 % Fe3+, pH = 9 5 % Fe3+, pH = 12 3+ 0.5 20 % Fe , pH = 12 20 % Fe3+, pH = 9 ] 3+ 0.4 ]+[Fe 2+ 0.3 ] / ([Fe

2+ 0.2 [Fe

0.1

0.0 0 50 100 150 200 250 300 D / kGy

Figure 8. The Fe2+ fraction ([Fe2+/(Fe2++Fe3+)]) in relation to dose, pH and initial amount of Fe3+ inFigure precursor 8. The suspensions Fe2+ fraction as determined ([Fe2+/(Fe2++Fe using3+)]) thein relation 1,10-phenanthroline to dose, pH UV-Visand initial spectrophotometric amount of Fe3+ in method.precursor Samples suspensions were obtained as determined from 1.85 using wt% PEOthe precursor1,10-phenanthroline suspensions UV-Vis with 0.2 spectrophotometric M 2-propanol. method. Samples were obtained from 1.85 wt% PEO precursor suspensions with 0.2 M 2-propanol. The increase of 2-propanol concentration in initial suspensions to 0.8 M did not result in the formationThe of increase gels, but of in 2-propanol the formation concentration of black magnetic in init suspensions.ial suspensions Figure to 90.8 shows M did the not XRD result patterns in the andformation Mössbauer of gels, spectrum but in at roomthe formation temperature of blac of blackk magnetic magnetic suspensions. powders isolated Figure from 9 shows suspensions the XRD 3+ obtainedpatterns upon and irradiationMössbauer of spectrum 20 wt% Fe at roomsuspensions temperat withure 0.8 of M black 2-propanol magnetic at 50powders kGy (Figure isolated9a) and from atsuspensions 130 kGy (Figure obtained9b,c). upon These irradiation samples consist of 20 ofwt% magnetite Fe3+ suspensions with the volume-average with 0.8 M 2-propanol domain at size 50 ofkGy approximately(Figure 9a) and 3 nm at 130 insize kGy (Table (Figure1). 9b,c). The doublet These samp in theles Mössbauer consist of magnetite spectrum, with which the can volume-average be assigned todomain small superparamagnetic size of approximately maghemite 3 nm in (oxidized size (Table magnetite 1). The doublet nanoparticles), in the Mössbauer is in accordance spectrum, with XRDwhich Nanomaterials 2019, 9, x FOR PEER REVIEW 14 of 29 linecan broadening be assigned analysis to small (D311 superparamagnetic= 3.2 nm). maghemite (oxidized magnetite nanoparticles), is in accordance with XRD line broadening analysis (D311 = 3.2 nm). 125 1000 PEOFe-11 Magnetite Card No. 19-0629 PEOFe-16 Magnetite Card No. 19-0629 (a) 900 (b) 100 800 700 75 600 500

50 Counts 400 Counts 300 25 200 100 0 0 10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 2θ / ° (CuKα) 2θ / ° (CuKα)

3.52x105

3.48x105

3.44x105 δ = 0.35 mm s-1 Δ = 0.75 mm s-1 5 Counts 3.40x10 Γ = 0.66 mm s-1

5 3.36x10 (c) 3.32x105 -10 -5 0 5 10 Doppler velocity / mm s-1

Figure 9.FigureXRD patterns9. XRD patterns and Mössbauer and Mössbauer spectrum spectrum at room at r temperatureoom temperature of PEOof PEO/Fe-oxide/Fe-oxide black black magnetic powdersmagnetic obtained powders from 1.85obtained wt% from PEO 1.85 suspensions wt% PEO suspensions with 20 wt% with Fe 203 +wt%and Fe four3+ and times four times more more 2-propanol (0.8 M) at2-propanol 50 kGy ((0.8a) and M) at 130 50 kGy (a ()b and,c). 130 kGy (b,c).

As can be seen from the above results, a critical step in the one-step synthesis of magnetic iron oxide nanocomposite gels was to find a balance between the good reducing conditions required for the formation of magnetic particles and the conditions suitable for the formation of the polymer network. The permanent shape of wall-to-wall gels obtained on irradiation without 2-propanol is strong evidence of a three-dimensional network and PEO intermolecular crosslinking. It is known that on irradiation of dilute PEO aqueous solutions, the main mechanism for crosslinking of PEO chains is a reaction with hydroxyl radicals formed on water radiolysis (Equation (1)) [31,32,44]. If conditions are favorable (a relatively low dose rate and a diluted PEO concentration), owing to good mobility in dilute solutions, such formed PEO macroradicals preferably crosslink resulting in macrogel formation.

CH2-CH2-O-CH2-CH2-O- + •OH (H•) → -CH2-CH2-O-CH2-CH•-O- + H2O (H2) (3)

• -CH2-CH-O- 2 -CH2-CH-O- (4)

-CH2-CH-O-

The precursor particles did not significantly disturb the intermolecular crosslinking of PEO chains. Although during the process of formation of PEO network, •OH radicals, which are oxidizing agents, are partially removed from the system, the reducing conditions when irradiated without 2- propanol were still not strong enough to reduce the Fe(III) precursor nanoparticles. When 2-propanol in a concentration of 0.2 M was added in a system to enhance reducing conditions, magnetite nanoparticles were formed (Figures 5–8). 2-propanol scavenges hydroxyl radicals which are oxidizing agents, and prevents back oxidation of formed ferrous ions

Nanomaterials 2019, 9, x FOR PEER REVIEW 14 of 29

125 1000 PEOFe-11 Magnetite Card No. 19-0629 PEOFe-16 Magnetite Card No. 19-0629 (a) 900 (b) 100 800 700 75 600 500

50 Counts 400 Counts 300 25 200 100 0 0 10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 2θ / ° (CuKα) 2θ / ° (CuKα)

3.52x105

3.48x105

3.44x105 δ = 0.35 mm s-1 Δ = 0.75 mm s-1 5 Counts 3.40x10 Γ = 0.66 mm s-1

5 3.36x10 (c) 3.32x105 -10 -5 0 5 10 Doppler velocity / mm s-1

Figure 9. XRD patterns and Mössbauer spectrum at room temperature of PEO/Fe-oxide black magnetic powders obtained from 1.85 wt% PEO suspensions with 20 wt% Fe3+ and four times more 2-propanol (0.8 M) at 50 kGy (a) and 130 kGy (b,c). Nanomaterials 2020, 10, 1823 13 of 25 As can be seen from the above results, a critical step in the one-step synthesis of magnetic iron oxide nanocomposite gels was to find a balance between the good reducing conditions required for As can be seen from the above results, a critical step in the one-step synthesis of magnetic iron the formation of magnetic particles and the conditions suitable for the formation of the polymer oxide nanocomposite gels was to find a balance between the good reducing conditions required for network. The permanent shape of wall-to-wall gels obtained on irradiation without 2-propanol is the formation of magnetic particles and the conditions suitable for the formation of the polymer strong evidence of a three-dimensional network and PEO intermolecular crosslinking. It is known network. The permanent shape of wall-to-wall gels obtained on irradiation without 2-propanol is that on irradiation of dilute PEO aqueous solutions, the main mechanism for crosslinking of PEO strong evidence of a three-dimensional network and PEO intermolecular crosslinking. It is known that chains is a reaction with hydroxyl radicals formed on water radiolysis (Equation (1)) [31,32,44]. If on irradiation of dilute PEO aqueous solutions, the main mechanism for crosslinking of PEO chains is conditions are favorable (a relatively low dose rate and a diluted PEO concentration), owing to good a reaction with hydroxyl radicals formed on water radiolysis (Equation (1)) [31,32,44]. If conditions mobility in dilute solutions, such formed PEO macroradicals preferably crosslink resulting in are favorable (a relatively low dose rate and a diluted PEO concentration), owing to good mobility in macrogel formation. dilute solutions, such formed PEO macroradicals preferably crosslink resulting in macrogel formation. CH2-CH2-O-CH2-CH2-O- + •OH (H•) → -CH2-CH2-O-CH2-CH•-O- + H2O (H2) (3) CH -CH -O-CH -CH -O- + •OH (H•) -CH -CH -O-CH -CH•-O- + H O (H ) (3) 2 2 2 2 → 2 2 2 2 2

• -CH2-CH-O- 2 -CH2-CH-O- (4)

-CH2-CH-O- (4) The precursor particles did not significantly disturb the intermolecular crosslinking of PEO chains. The precursor particles did not significantly disturb the intermolecular crosslinking of PEO Although during the process of formation of PEO network, OH radicals, which are oxidizing agents, chains. Although during the process of formation of PEO network,• •OH radicals, which are oxidizing are partially removed from the system, the reducing conditions when irradiated without 2-propanol agents, are partially removed from the system, the reducing conditions when irradiated without 2- were still not strong enough to reduce the Fe(III) precursor nanoparticles. When 2-propanol in a propanol were still not strong enough to reduce the Fe(III) precursor nanoparticles. When 2-propanol concentration of 0.2 M was added in a system to enhance reducing conditions, magnetite nanoparticles in a concentration of 0.2 M was added in a system to enhance reducing conditions, magnetite were formed (Figures5–8). 2-propanol scavenges hydroxyl radicals which are oxidizing agents, nanoparticles were formed (Figures 5–8). 2-propanol scavenges hydroxyl radicals which are and prevents back oxidation of formed ferrous ions oxidizing agents, and prevents back oxidation of formed ferrous ions

•OH (H•) + (CH ) CHOH H O (H ) + (CH ) C•OH (5) 3 2 → 2 2 3 2 Thus formed 2-propanol radicals can act as additional reducing agent [45]:

3+ 2+ + Fe + (CH ) C•OH Fe + (CH ) CO + H (6) 3 2 → 3 2 The addition of 2-propanol had a pronounced effect on the formation of PEO gels. Irradiation of pure PEO solutions with 2-propanol (0.2 M) did not result in the formation of any gel content, even at 300 kGy. By scavenging the •OH radicals, which are also initiators of PEO crosslinking, 2-propanol reduced the yield of PEO macroradicals, and hence the crosslinking degree. On the other hand, on irradiation of PEO/Fe(III) precursor suspensions, magnetic gels were formed, and the amount of gel depended on the dose and initial amount of Fe3+ (Figure2). This suggests that PEO chains are additionally crosslinked through formed iron oxide nanoparticles. An additional reason may lie in the possible contribution of iron oxide NPs to the formation of higher yield of hydroxyl radicals, like the one observed for nanosilica by Le Caër et al. [46]. They showed that on irradiation of the silica/water system, an exciton was formed that may be scavenged by water molecules and further react to produce additional hydroxyl radicals, protons, and hydrogen. Besides, the additional hydroxyl radicals may be formed in the Fenton reaction, because the Fe3+ that is reduced to Fe2+ by hydrated electrons may be reoxidized by radiolytically formed H2O2 thus producing additional •OH radicals. At higher 2-propanol concentrations (0.8 M) there were not enough hydroxyl radicals left for the polymer to crosslink, resulting in the formation of highly magnetic magnetite NPs suspensions (Figure8). When more concentrated PEO suspensions (4 wt%) pure or with various amount of Fe3+ were irradiated, black wall-to-wall hydrogels were obtained, similar to those obtained by irradiation without 2-propanol, probably due to the higher yield of PEO intermolecular crosslinking (photos in Figure2). Therefore, by finding the optimal conditions we were able to synthesize magnetic PEO/iron oxide nanocomposite gels in a single step. Nanomaterials 2020, 10, 1823 14 of 25

3.2. Thermal Characterization of Gels The thermal properties of obtained nanocomposite gels in dependence on the γ-irradiation dose and the initial Fe3+ concentration were studied by differential scanning calorimetry. The DSC thermograms are given in Figure S2 in Supplementary Materials, while the phase transformation enthalpies and temperatures of the second heating and first cooling cycles are given in Figure 10, Figure 11, Figures S3 and S4. The results are presented in dependence on both the irradiation dose and the mass percentage of Fe3+ in precursor suspensions, as some effects are easier to observe. It must be emphasized again that when comparing the results of pure PEO gels with those of , one has to take into account that pure PEO gels were formed on irradiation without 2-propanol, because with 2-propanol in pure PEO solutions gels were not formed. Therefore, the difference between pure PEO gels and nanocomposite gels is not solely due to the amount of NPs within the gel, but also due to the influence of 2-propanol on network formation. Enthalpies and temperatures of melting and crystallization decreased with the dose over the entire dose range for both pure PEO gels and PEO/iron oxide nanocomposite gels (Figure 10, Figure 11, Figures S3 and S4). At 300 kGy almost completely amorphous gels with a high degree of crosslinking were obtained, especially in the case of pure PEO gel and gel obtained from 20 wt% Fe3+ suspension (Figure 10). At higher doses there is a higher yield of PEO intermolecular crosslinking; crosslinks are dense enough to significantly restrain mobility and impede crystallization of PEO chains on drying, which resulted in more amorphous gels. The smallest changes of enthalpies in dependence on dose were for gels with the highest NPs content (highest initial Fe3+ content) (Figure 10). The decrease of melting (Tm) and crystallization (Tc) temperatures with the dose was the most abrupt for pure PEO gels (Figure 11). A high density of crosslinks is the main reason for such low Tm and Tc. Crosslinks increase the number of defects in the crystalline phase resulting in less “perfect“ crystallites and consequently decrement in the melting temperature. Crosslinks also impose restrictions on molecular motions of PEO chains and at high doses the high density of PEO crosslinks and a small segment of PEO chains between two crosslink junctions seriously impede crystallizability, resulting in a significant lowering of Tc. Such a decrease of Tm and enthalpies of melting with the dose for PEO gels was also reported by other authors [12,30,47]. The increased amount of initial Fe3+ salt at the same dose led to the increased amount of formed gel with reduced enthalpies, suggesting the enhanced PEO crosslinking through formed Fe-oxide NPs (Figure 10 and Figure S4). In general, all nanocomposite gels at a certain dose had higher melting and crystallization temperatures than pure PEO gels (Figure 11). The amount of nanoparticles (initial Fe3+ content) at a certain dose had a significant effect on enthalpies decrease, but very little impact on temperatures (Figure 10). While there was approximately a 20 ◦C jump in Tm and Tc for gel obtained from 1% Fe3+ suspension compared to pure PEO gels at the certain dose, the further increase of initial Fe3+ to 20% resulted in a higher amount of formed gels with lower enthalpies but only slightly lower Tc and Tm. A similar trend in temperature changes with the increase of silica NPs we observed in our previous work [12]. The observed change in the melting enthalpies and temperatures with the NPs content can be explained by the impact of NPs. NPs can act as nucleating agents—they induce heterogeneous crystallization centers and thus facilitate crystallization of uncrosslinked PEO segments and increase Tc [48]. On the other hand, at high NPs concentration, due to worsened dispersion and partial NPs agglomeration, NPs can restrain the mobility of partially crosslinked PEO chains and its crystallizability. At the same time, agglomerates can be a serious obstacle for PEO crosslinking. Since in composites obtained with 2-propanol, more gel was formed with initial Fe3+ increase, the main reason for the decrease in enthalpies is obviously the additional crosslinking of PEO chains by NPs. So, while pure PEO gels are formed by intermolecular crosslinking of PEO chains, in nanocomposites some type of interactions/bonding between Fe-oxide NPs and PEO chains such as H-bonds or coordination bonding [49–51] are likely to contribute to the radiation-induced crosslinking of PEO chains, resulting in further amorphization. The highest dose dependence of enthalpies and temperatures for pure PEO Nanomaterials 2020, 10, 1823 15 of 25

gels support the above conclusion. The interactions of Fe2O3 and TiO2 particles and polymer have been observed observed by Popescu et al. [52], whereas Davenas et al. [53] noted that silica filler can act as an additional crosslinker that upon irradiation formed covalent bonds to the polymer matrix. Similarly, Agrawal et al. [54] observed laponite nanoparticles acting as additional junction points in physically associative PLA-PEO-PLA gel. Criado-Gonzalez et al. [55] and Peng at al. [56] observed the formation of a weak network due to crosslinking through coordination bonding of Fe(III) cations with alginate chains and PAA chains, respectively. The role of NPs as nucleating agents can also be inferred from the behavior of gels obtained without 2-propanol. For nanocomposite gels prepared without 2-propanol the sole effect of nanoparticle content can be observed compared to the pure PEO gels (Figure 10, Figure 11and Figure S4). The pure PEO gels and nanocomposite gels prepared from 4 wt% PEO suspensions had higher melting and crystallization enthalpies and temperatures than those obtained from 1.85 wt% PEO solutions at the same dose (Figure 10 and Figure S3). For pure PEO gels exactly the opposite would be expected; that irradiation of a more concentrated polymer solution would result in a higher crosslinks density and a more amorphous gel. The phase transformation temperatures of nanocomposite gels containing the same NPs content did not depend on PEO concentration, they were almost the same for Nanomaterials1.85 and 4 2019 wt%, 9, suspensions. x FOR PEER REVIEW 16 of 29

Figure 10. Melting (∆Hm) and crystallization (∆Hc) enthalpies of the 2nd heating cycles and 1st cooling Figure 10. Melting (ΔHm) and crystallization (ΔHc) enthalpies of the 2nd heating cycles and 1st cooling cycles, respectively, of the obtained gels in dependence on the irradiation dose and the mass percentage cycles,of Fe 3+respectively,in precursor suspensions.of the obtained Unless gels otherwise in dependence indicated, theon precursorthe irradiation suspensions dose were and prepared the mass 3+ percentagefrom 1.85 of wt% Fe PEOin precursor solutions andsuspensions. with 0.2 M Unless 2-propanol. otherw Massise indicated, percentage the of precursor Fe3+ of 1, suspensions 5, 20 wt% were(relative prepared to total from PEO 1.85 and wt% Fe3+ PEOmass) solutions in the precursor and with suspensions 0.2 M 2-propanol. correspond Mass to concentrations percentage of FeFe33++ of 1, ions5, 20 of wt% 0.35 (relative10 2 M, 1.75to total10 PEO2 M, andand 7 Fe103+ mass)2 M in in the the case precursor of 1.85 wt% suspensions PEO suspensions. correspond In the to × − × − × − concentrationscase of 4 wt% PEOof Fe suspensions,3+ ions of 0.35 Fe ×3+ 10concentrations−2 M, 1.75 × 10 of−2 1.75M, and10 72 ×M 10 and−2 M 7 in 10the2 caseM correspond of 1.85 wt% to 2.3 PEO × − × − suspensions.and 9.3 wt% In Fe the3+. case of 4 wt% PEO suspensions, Fe3+ concentrations of 1.75 × 10−2 M and 7 × 10−2 M correspond to 2.3 and 9.3 wt% Fe3+.

NanomaterialsNanomaterials 20192020, 9, ,x10 FOR, 1823 PEER REVIEW 16 of 2517 of 29

Figure 11. Melting (Tm) and crystallization (Tc) temperatures of the 2nd heating cycles and 1st cooling Figure 11. Melting (Tm) and crystallization (Tc) temperatures of the 2nd heating cycles and 1st cooling cycles, respectively, of the obtained gels in dependence on the irradiation dose and the mass percentage cycles,of Fe respectively,3+ in precursor of suspensions. the obtained Unless gels otherwise in dependence indicated, the on precursor the irradiation suspensions dose were and prepared the mass percentagefrom 1.85 of wt% Fe3+ PEO in precursor solutions and suspensions. with 0.2 M 2-propanol.Unless otherwise indicated, the precursor suspensions were prepared from 1.85 wt% PEO solutions and with 0.2 M 2-propanol. 3.3. Rheological Properties of Gels EnthalpiesThe rheological and temperatures properties of of the melting gels were and investigated crystallization at different decreased conditions with within the dose the linear over the entireviscoelastic dose range region for (LVR).both pure PEO gels and PEO/iron oxide nanocomposite gels (Figures 10, 11, S3, S4). At 300The kGy amplitude almost sweepcompletely test at amorphous room temperature gels with for gels a high obtained degree at variousof crosslinking doses starting were fromobtained, especially1.85 wt% in PEOthe case precursor of pure suspensions PEO gel and is presented gel obtained in Figure from 12 a,b,c20 wt% and Fe Figure3+ suspension S5 in Supplementary (Figure 10). At higherMaterials doses andthere the is values a higher are givenyield inof TablePEO2 .interm All investigatedolecular crosslinking; gels behave like crosslinks viscoelastic are solids dense with enough G values (storage modulus) higher than G” values (loss modulus), confirming a well-ordered gel to significantly0 restrain mobility and impede crystallization of PEO chains on drying, which resulted network [57]. Storage moduli (as well as yield and flow point) increased with the irradiation dose in more amorphous gels. The smallest changes of enthalpies in dependence on dose were for gels for both pure PEO gels and all nanocomposite gels indicating higher network density. Slightly stiffer with the highest NPs content (highest initial Fe3+ content) (Figure 10). The decrease of melting (Tm) gels were formed at higher dose. The increase of the storage modulus G0 and crosslink density with and thecrystallization irradiation dose (Tc) hastemperatures been observed with for the di ffdoseerent was radiation the most crosslinked abrupt gels,for pure like PEOPEO[ 58gels] and (Figure 11). PVPA high hydrogels density [59 of]. Allcrosslinks nanocomposite is the main gels hadreason higher for storage such low moduli, Tm and yield T pointsc. Crosslinks and flow increase points the numbercompared of defects to pure in PEO the gels crystalline at a certain phase dose. resulting Generally, atin aless particular “perfect“ dose, crystallites storage moduli, and yield consequently point decrementand flow in point the increasedmelting temperature. with the increase Crosslinks of Fe3+ content also inimpose precursor restrictions suspensions. on molecular At 50 and 130 motions kGy of PEOstorage chains moduli and at increasedhigh doses with the initial high Fedensity3+ content, of PEO while crosslinks at 300 kGy and there a small was ansegment extreme of increase PEO chains 3+ betweenfor gel two prepared crosslink from junctions 5 wt% Fe seriouslysuspension impede compared crystallizability, to pure PEO resulting gel but no in further a significant increase lowering for gel prepared from 20 wt% Fe3+ suspension. The quantity loss factor, tan(δ) = G”/G , determines the of Tc. Such a decrease of Tm and enthalpies of melting with the dose for PEO gels0 was also reported relative elasticity of viscoelastic materials. The gels with a value of tan( ) = 0.1 and lower belong to by other authors [12,30,47]. δ stiff gels and are indicative of well-ordered systems. For 50 kGy loss factor value was low for pure The increased amount of initial Fe3+ salt at the same dose led to the increased amount of formed PEO (0.01) compared to gels prepared from 5 % (0.11) and 20% (0.08) Fe3+ suspensions, meaning that gel withpure PEOreduced has a betterenthalpies, ordered suggesting microstructure the ofenhanced the gel compared PEO crosslinking to nanocomposites. through At formed 130 kGy Fe-oxide the NPs (Figures 10 and S4). In general, all nanocomposite gels at a certain dose had higher melting and crystallization temperatures than pure PEO gels (Figure 11). The amount of nanoparticles (initial Fe3+ content) at a certain dose had a significant effect on enthalpies decrease, but very little impact on temperatures (Figure 10). While there was approximately a 20 °C jump in Tm and Tc for gel obtained from 1% Fe3+ suspension compared to pure PEO gels at the certain dose, the further increase of initial Fe3+ to 20% resulted in a higher amount of formed gels with lower enthalpies but only slightly lower

Nanomaterials 2020, 10, 1823 17 of 25 values of storage modulus, the yield, and flow points of gel obtained at 20% initial Fe3+ content were 2 to 4 times higher than for gel from initial 5 wt% Fe3+ and pure PEO, respectively, but the loss factor values of these gels were similar. The gel obtained from suspensions with the highest Fe3+ content (20%) at 300 kGy showed the longest yielding zone (the zone between yield point and flow point) Nanomaterialsthrough a range 2019, 9, of x FOR 700 PEER Pa indicating REVIEW very stiff structure. 20 of 29

10,000 50 kGy 1.85% PEO / Pa / Pa G' 1000 G''

100 20 % 5 %

Storage modulus modulus Storage Loss modulus 10 0 % (a) 0.1 1 10 100 1000 Shear stress τ / Pa 10,000 / Pa / Pa G' 1000 20 % G'' 5 % 0 %

100 Storage modulus modulus Storage Loss modulus 130 kGy 1.85% PEO (b) 10 0.1 1 10 100 1000 Shear stress τ / Pa 10,000 / Pa / Pa G' 1000 20 % G'' 5 % 0 % 100

300 kGy Storage modulus modulus Storage modulus Loss 1.85% PEO 10 (c) 0.1 1 10 100 1000 Shear stress τ / Pa 10,000 / Pa / Pa G' 1000 G''

100 9.3 % 2.3 % Storage modulus modulus Storage Loss modulus 130 kGy 0 % 4% PEO 10 (d) 0.1 1 10 100 1000 Shear stress τ / Pa

Figure 12. AmplitudeAmplitude sweep sweep test test ( (GG′0 ((■) )and and GG″”( (▲N)) values) values) of of pure pure PEO PEO gels gels and and nanocomposite nanocomposite gels gels obtained at ((aa)) 5050 kGy,kGy, ( b(b)) 130 130 kGy, kGy, and and (c )(c 300) 300 kGy kGy from from 1.85 1.85 wt% wt% PEO PE suspensions,O suspensions, and and at (d at) 130(d) kGy130 3+ kGyfrom from 4.0 wt% 4.0 wt% PEO PEO suspensions, suspensions, at 25 at◦C. 25 The °C. initialThe initial mass mass percentage percentage of Fe of (relativeFe3+ (relative to the to total the total PEO 3+ PEOand Feand Femass)3+ mass) in precursor in precursor suspensions suspensions is indicated. is indicated.

Nanomaterials 2020, 10, 1823 18 of 25

Table 2. Results of amplitude sweep test of pure PEO gels and PEO/Fe-oxide nanocomposite gels.

Initial Mass Mass Percentage Loss Factor G Yield Flow Percentage of PEO Aqueous D/kGy 0 (tan δ = 3+ (Max)/Pa Point/Pa Point/Pa of Fe */% Solution/% G”/G0) 0 1.85 50 816 56.0 166.3 0.01 5 1.85 50 1115 117.4 192 0.11 20 1.85 50 1907 192.8 375 0.08 0 1.85 130 2336 98.5 377.3 0.04 5 1.85 130 5519 201.9 439.2 0.05 20 1.85 130 7300 455.3 788.3 0.04 0 1.85 300 2784 124.2 559.7 0.04 5 1.85 300 10,325 386.3 736.3 0.03 20 1.85 300 7220 355.0 1086 0.03 0 4 130 6507 298.6 1261 0.03 2.3 ** 4 130 1800 82.4 517.4 0.05 9.3 ** 4 130 1200 72.8 450.2 0.09 Pure PEO gels were prepared by irradiation without the addition of 2-propanol. * Initial wt% of Fe3+ relative to the total mass of PEO and Fe3+ in precursor suspensions. ** 4 wt% PEO suspensions had the same initial concentration of Fe3+ salt as in 1.85 wt% PEO, i.e., suspensions contained ~2x more PEO but the same amount of Fe3+, resulting in 2.3 and 9.3 wt% of Fe3+ relative to the total PEO and Fe3+ mass.

The results show that a higher degree of intermolecular crosslinking with increased dose increases the storage modulus and the gel becomes stiffer. But pure PEO with a high degree of exclusively intermolecular PEO crosslinking was still softer than nanocomposite gels. The increase in elastic moduli, flow, and yield points of nanocomposite gels and the formation of stronger gels confirms a well-ordered gel network, well-ordered microstructure and indicates good NPs dispersion. This increase is not only due to the reinforcing effect of inorganic NPs, but the formed magnetic NPs facilitate the formation of gels by acting as additional crosslinkers. Similarly, Agrawal et al. [54] found that G0 of PLA-PEO-PLA gel increased dramatically when the amount of laponite particles was increased, indicating the formation of new junctions by the nanoparticles. Blyakhman et al. [1] reported the increase in Young’s modulus of PAAm gel resulting from the addition of a low concentration of magnetic NPs, reflecting the direct effect of magnetic NPs on gel elasticity, and reported the possibility that MNPs act as crosslinking agents. All this is consistent with the decrease in enthalpies and temperatures of these gels with increasing 3+ dose and Fe content, as observed from DSC measurements. In addition, the similar values of G0 of 20% compared to 5% gels at 300 kGy (in line with the similar melting enthalpies) (Figure 12 and Figure S5) may be due to the likely agglomeration of NPs at high NPs content. The amplitude sweep test of gels obtained at 130 kGy from 4 wt% PEO precursor suspensions showed the opposite behavior depending on the initial Fe3+ content (Figure 12d) to gels obtained from 1.85 wt% PEO suspensions. The gels obtained from 4 wt% PEO suspensions had the same initial concentration of Fe3+ salt as in 1.85 wt% PEO (resulting in 2.3 and 9.3 wt% Fe3+ relative to the total initial mass of PEO and Fe3+). The pure PEO gel obtained from 4 wt% PEO was by far the strongest by all parameters, while the gel with the highest NPs content was the weakest. G0 values, yield point and flow point were in descending order with increasing initial Fe3+ amount, while the loss factor (tan δ) increased. As expected, stronger gels can be obtained by increasing the polymer concentration in the starting suspension, resulting in a higher density of the network formed by intermolecular crosslinking. But the effect of NPs did not show an improved behavior as in the case of lower polymer concentration. Obviously, at higher polymer concentration the effect of intermolecular crosslinking of PEO chains is dominant over crosslinking through Fe-oxide NPs. Figure 13 presents the frequency sweep measurements. Frequency sweeps describe the time-dependent behavior of a sample in the non-destructive deformation range. The investigated samples are true chemically crosslinked gels. The frequency sweep of pure PEO and magnetic nanocomposite gels showed constant storage modulus (G0) values within the entire frequency range Nanomaterials 2020, 10, 1823 19 of 25

(100 rad/s to 0.05 rad/s). Pure PEO and PEO/Fe-oxide gels showed similar behavior at lower frequencies, but at higher frequencies, pure PEO gel obtained more ordered structure, better homogeneity (lower loss factor) compared to nanocomposite gels with higher loss factor values (because of increase of loss modulus at higher frequences). The same remarks related to internal gel microstructure have been seen forNanomaterials pure PEO 2019 gels, 9, x andFOR PEER nanocomposites REVIEW at different doses except for gels from 4 wt%22 ofof 29 PEO.

Figure 13.FigureFrequency 13. Frequency sweep sweep tests tests (G 0(G(′ ()■) and and GG”(″ (▲N) values) values) of of pure pure PEO PEO gels gelsand nanocomposite and nanocomposite gels gels obtainedobtained at (a) 50 at kGy,(a) 50 ( kGy,b) 130 (b) kGy, 130 kGy, and and (c) 300 (c) 300 kGy kGy from from 1.85 1.85 wt% wt% PEO PEO suspensions,suspensions, and and at at(d) ( d130) 130 kGy kGy from 4.0 wt% PEO suspensions, at 25 °C. Initial mass percentage of Fe3+ in3+ precursor suspensions from 4.0 wt% PEO suspensions, at 25 ◦C. Initial mass percentage of Fe in precursor suspensions is indicated. is indicated. Results of 3ITT thixotropy test are given in Figures 14 and S6. Thixotropy is a time-dependent Resultsphenomenon. of 3ITT thixotropyIt can be defined test areas the given shear in thinning Figure 14 behavior and Figure of a viscoelastic S6. Thixotropy gel sample is a time-dependentupon the phenomenon.application It can of a be destructive defined strain as the and shear the thinningsubsequent behavior recovery of the a viscoelastic viscoelastic properties gel sample after upon the applicationcessation of a of destructive the strain. Figure strain 14 and shows the subsequentthe evaluation recovery of complex of theviscosity viscoelastic and Figure properties S6 the after cessationevolution of the strain. of G′ and Figure G″ of 14 gels shows after theapplying evaluation the destructive of complex strain. viscosity The sample and behavior Figure S6switched the evolution from gel-like to sol-like, with G″ values higher than G′. After that, the original conditions reapplied of G0 and G” of gels after applying the destructive strain. The sample behavior switched from gel-like and the recovery of viscoelastic properties of gels was observed. Recovery ratios of gels within 60 s to sol-like, with G” values higher than G0. After that, the original conditions reapplied and the recovery

Nanomaterials 2020, 10, 1823 20 of 25 of viscoelastic properties of gels was observed. Recovery ratios of gels within 60 s are given in Table3. The initial Fe3+ amount, that is the amount of NPs formed, and the density of the PEO network formed at a specific γ-irradiation dose influenced the self-recoverable properties of the new nanocomposite materials. The most prominent recovery of the viscoelastic properties was observed for gels from 1.85 wt% PEO suspensions at 50 kGy and gels from 4 wt% PEO suspensions at 130 kGy (Table3). The recovery of pure PEO gel at the specific irradiation dose was always better compared to the nanocomposite gels (indicating very good microstructure integrity of pure PEO gel), except for gels obtained from 4 wt% PEO suspensions. The lowest recovery was observed for gel obtained at 300 kGy from 20% Fe3+ suspension (57.1% recovery in 60 s). Better recovery was observed for all gel samples with lowerNanomaterialsG0 values 2019, 9 compared, x FOR PEER REVIEW to the recoverable properties of the most potent gels. 24 of 29

Figure 14.Figure3-interval 14. 3-interval thixotropy thixotropy test test (3ITT) of ofpure pure PEO PEO gel and gel nanocomposite and nanocomposite gels obtained gels at ( obtaineda) 50 at (a) 50 kGy,kGy, ( b)) 130 130 kGy, kGy, (c) 300 (c) kGy 300 from kGy 1.85 from wt% 1.85 PEO wt%suspensions PEO suspensionsand (d) from 4 andwt% PEO (d) fromsuspensions 4 wt% PEO suspensionsat 130 at kGy 130 showed kGy showed complex complex viscosity viscosity(ղ*) as a function ( ղ *) as of a functiontime and ofapplication time and of application different strains of di fferent (LVR-DR-LVR) at 25 °C. Linear viscoelastic region (LVR): strain = 0.1%, frequency = 5 Hz; destructive strains (LVR-DR-LVR) at 25 ◦C. Linear viscoelastic region (LVR): strain = 0.1%, frequency = 5 Hz; region (DR): strain =300%, frequency = 5 Hz. Initial mass percentage of Fe3+ in precursor suspensions destructive region (DR): strain =300%, frequency = 5 Hz. Initial mass percentage of Fe3+ in precursor is indicated. suspensions is indicated. The obtained magnetic PEO/iron oxide nanocomposite gels showed promising rheological properties for potential application in tissue engineering and as wound dressings. By further optimization of the system and irradiation conditions, magnetic gels with tailored properties for a particular application could be synthesized.

1

Nanomaterials 2020, 10, 1823 21 of 25

Table 3. Self-recoverable properties of pure PEO gels and nanocomposite gels obtained at various doses and compositions of precursor suspensions determined in 3-interval thixotropy test.

wt% Fe3+ 0 5 20 0 5 20 0 5 20 0 2.3 9.3 wt% PEO 1.85 1.85 1.85 1.85 1.85 1.85 1.85 1.85 1.85 4.0 4.0 4.0 Dose (kGy) 50 50 50 130 130 130 300 300 300 130 130 130 Recovery 3ITT test (%) 96.1 90.7 84.7 70.8 59.7 60.5 87.0 57.6 57.1 74.6 90.4 92.6 t = 60 s

The obtained magnetic PEO/iron oxide nanocomposite gels showed promising rheological properties for potential application in tissue engineering and as wound dressings. By further optimization of the system and irradiation conditions, magnetic gels with tailored properties for a particular application could be synthesized.

4. Conclusions γ-irradiation proved to be a suitable method for the one-step synthesis of magnetic PEO/iron oxide nanocomposite hydrogels. For the one-step irradiation synthesis of magnetic PEO/iron oxide hydrogels the appropriate balance between conditions suitable for polymer crosslinking and conditions suitable for the reduction of Fe(III)-precursor was crucial. γ-irradiation generated Fe2+ was quantitatively determined using the 1,10-phenanthroline UV-Vis spectrophotometric method. A maximum Fe2+ mole fraction of 55% was achieved at a dose of 300 kGy and with 5 wt% of the initially added Fe3+ at pH ~ 12. The thermal, viscoelastic, and magnetic properties of the gels depended on the irradiation dose, and the PEO and initial Fe3+ concentration, i.e., amount of magnetic iron oxide NPs inside the gels. Stronger gels were formed at the higher dose and higher magnetite NPs content (in the case of 1.85 wt% PEO). Both rheological measurements and DSC results suggested that the pronounced increase in strength and stiffness of nanocomposite gels was not only due to the reinforcing effect caused by the presence of iron oxide NPs, but that the formed magnetic iron oxide nanoparticles acted as additional crosslinkers of the PEO chains, thus facilitating the formation of potent gels. At higher PEO concentrations (4 wt%), the effect of intermolecular crosslinking of PEO chains was dominant over the effect of Fe-oxide NPs. γ-irradiation of aqueous suspensions containing PEO, Fe3+, and 2-propanol in alkali offered a possibility to obtain magnetic PEO/iron oxide nanocomposite gels with low crystallinity and improved strength. By further optimization of the system and irradiation conditions (pH, polymer molecular mass, polymer and precursor concentration, as well as dose and dose rate) magnetic gels with tailored properties for a specific application could be synthesized.

Supplementary Materials: The following are available online at http://www.mdpi.com/2079-4991/10/9/1823/s1, Figure S1: The room temperature Mössbauer spectra of PEO/Fe-oxide gels obtained from suspensions with: (a) 5 wt% Fe3+ at 130 kGy; (b) 20 wt% Fe3+ at 130 kGy; (c) 5 wt% Fe3+ at 300 kGy; (d) 20 wt% Fe3+ at 300 kGy. All precursor suspensions were prepared from 1.85 wt% PEO solutions and with 0.2 M 2-propanol. Mössbauer parameters are given: δ = isomer shift relative to α-Fe at 20 ◦C; ∆ = quadrupole splitting; Γ = line width. Error: δ = 0.01 mm s-1; ∆ = 0.01 mm s-1, Figure S2: DSC thermographs of the 2nd heating (a,b) and the 1st cooling (c,d)± cycles of pure PEO± gel and nanocomposite gels obtained at 50, 130 and 300 kGy from 1.85 wt% PEO precursor suspensions with various Fe3+ content. Unless otherwise indicated, suspensions contained 0.2 M 2-propanol, Figure S3: Melting enthalpies and temperatures of the 1st heating cycles of the obtained gels in dependence on the irradiation dose and the mass percentage of Fe3+ in precursor suspensions. Unless otherwise indicated, the precursor suspensions were prepared from 1.85 wt% PEO solutions and with addition of 0.2 M 2-propanol. All gels obtained at 300 kGy with 2-propanol, and pure PEO gel at 130 kGy, were totally amorphous in the first heating cycle (no melting enthalpies). Gels obtained by irradiation from 5 wt% Fe3+ suspensions without 2-propanol had two melting maxima (both are given on graph), Figure S4: Melting (∆Hm) and crystallization Nanomaterials 2020, 10, 1823 22 of 25

(∆Hm) enthalpies and temperatures (Tm and Tc) of the 2nd heating cycles and the 1st cooling cycles, respectively, of gels obtained at various doses in dependence on the mass percentage of Fe3+ in 1.85 wt% PEO precursor suspensions. Unless otherwise indicated suspensions contained 0.2 M 2-propanol, Figure S5: Comparison of amplitude sweep test (G0 () and G”(N) values) of nanocomposite gels obtained at 130 kGy and 300 kGy (1.85 wt% 3+ PEO solution), at 25 ◦C. Initial mass percentage of Fe in precursor suspensions is indicated, Figure S6: 3-interval thixotropy test (3ITT) (storage G0 () and loss G”(N) modulus) of pure PEO gel and nanocomposite gels obtained at (a) 50 kGy, (b) 130 kGy, (c) 300 kGy from 1.85 wt% PEO suspensions and (d) from 4 wt% PEO suspensions at 130 kGy as a function of time and application of different strains (LVR-DR-LVR) at 25 ◦C. Linear viscoelastic region (LVR): strain = 0.1%, frequency = 5 Hz; destructive region (DR): strain = 300%, frequency = 5 Hz. Initial mass percentage of Fe3+ in precursor suspensions is indicated. Author Contributions: Conceptualization, T.J.; investigation, T.J., I.M.; formal analysis, I.M., N.Š.V., A.P., M.G., G.Š., J.-M.G., G.D.; methodology: T.J.; resources, T.J., M.G., N.Š.V., J.-M.G., G.D.; writing—original draft preparation, T.J., I.M., N.Š.V., M.G.; writing—review and editing, T.J.; supervision of experiments, T.J.; project administration/coordination, T.J.; funding acquisition, T.J. All authors have read and agreed to the published version of the manuscript. Funding: This work was financially supported by the Croatian Science Foundation under the project UIP-2017-05-7337 “The impact of polymers on the radiolytic synthesis of magnetic nanoparticles” (POLRADNANOP). Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

References

1. Blyakhman, F.A.; Makarova, B.E.; Fadeyev, F.A.; Lugovets, D.V.; Safronov, A.P.; Shabadrov, P.A.; Shklyar, T.F.; Melnikov, G.Y.; Orue, I.; Kurlyandskaya, G.V. The contribution of magnetic nanoparticles to ferrogel biophysical properties. Nanomaterials 2019, 9, 232. [CrossRef][PubMed] 2. Blyakhman, F.A.; Safronov, A.P.; Zubarev, A.Y.; Shklyar, T.F.; Makeyev, O.G.; Makarova, B.E.; Melekhin, V.V.; Larrañaga, A.; Kurlyandskaya, G.V. Polyacrylamide ferrogels with embedded maghemite nanoparticles for biomedical engineering. Results Phys. 2017, 7, 3624–3633. [CrossRef] 3. Hasan, A.; Morshed, M.; Memic, A.; Hassan, S.; Webster, T.J.; Marei, H.E.-S. Nanoparticles in tissue engineering: Applications, challenges and prospects. Int. J. Nanomed. 2018, 13, 5637–5655. [CrossRef] [PubMed] 4. Cvek, M.; Zahoranova, A.; Mrlik, M.; Sramkova, P.; Minarik, A.; Sedlacik, M. Poly(2-oxazoline)-based magnetic hydrogels: Synthesis, performance and cytotoxicity. Surf. B Biointerfaces 2020, 190, 110912. [CrossRef] 5. Li, Y.; Huang, G.; Zhang, X.; Li, B.; Chen, Y.; Lu, T.; Lu, T.J.; Xu, F. Magnetic hydrogels and their potential biomedical applications. Adv. Funct. Mater. 2012, 23, 1–13. [CrossRef] 6. Ramanujan, R.V.; Lao, L.L. The mechanical behavior of smart magnet-hydrogel composites. Smart Mater. Struct. 2006, 15, 952–956. [CrossRef] 7. Thoniyot, P.; Tan, M.J.; Karim, A.A.; Young, D.J.; Loh, X.J. Nanoparticle-Hydrogel composites: Concept, design, and applications of these promising, multifunctional materials. Adv. Sci. 2015, 2, 1400010. [CrossRef] 8. Hu, S.; Zhou, Y.; Zhao, Y.; Xu, Y.; Zhang, F.; Gu, N.; Ma, J.; Reynolds, M.A.; Xia, Y.; Xu, H.H.K. Enhanced bone regeneration and visual monitoring via superparamagnetic scaffold in rats. J. Tissue Eng. Regen. Med. 2018, 12, e2085–e2098. [CrossRef][PubMed] 9. Wang, Q.; Chen, B.; Ma, F.; Lin, S.; Cao, M.; Li, Y.; Gu, N. Magnetic iron oxide nanoparticles accelerate osteogenic differentiation of mesenchymal stem cells via modulation of long noncoding RNA INZEB2. Nano Res. 2017, 10, 626–642. [CrossRef] 10. Wang, Q.; Chen, B.; Cao, M.; Sun, J.; Wu, H.; Zhao, P.; Xing, J.; Yang, Y.; Zhang, X.; Ji, M.; et al. Response of MAPK pathway to iron oxide nanoparticles in vitro treatment promotes osteogenic differentiation of hBMSCs. Biomaterials 2016, 86, 11–20. [CrossRef] 11. Henstock, J.R.; Rotherham, M.; Rashidi, H.; Shakesheff, K.M.; El Haj, A.J. Remotely activated mechanotransduction via magnetic nanoparticles promotes mineralization synergistically with bone morphogenetic protein 2: Applications for injectable cell therapy. Stem Cells Transl. Med. 2014, 3, 1363–1374. [CrossRef][PubMed] Nanomaterials 2020, 10, 1823 23 of 25

12. Jurkin, T.; Puci´c, I. Irradiation effects in poly(ethylene oxide)/silica nanocomposite films and gels. Polym. Eng. Sci. 2013, 53, 2318–2327. [CrossRef] 13. Marinovi´c-Cincovi´c,M.T.; Radosavljevi´c,A.N.; Krsti´c,J.I.; Spasojevi´c,J.P.; Bibi´c,N.M.; Mitri´c,N.M.; Kaˇcarevi´c-Popovi´c,Z.M. Physicochemical characteristics of gamma irradiation crosslinked poly(vinyl alcohol)/magnetite ferrogel composite. Hem. Ind. 2014, 68, 743–753. [CrossRef] 14. Eid, M. Preparation and characterization of natural polymers as stabilizer for magnetic nanoparticles by gamma irradiation. J. Polym. Res. 2013, 20, 112. [CrossRef] 15. Jovanovi´c,Ž.; Krklješ, A.; Stojkovska, J.; Tomi´c,S.; Obradovi´c,B.; Miškovi´c-Stankovi´c,V.;Kaˇcarevi´c-Popovi´c,Z. Synthesis and characterization of silver/poly(N-vinyl-2-pyrrolidone) hydrogel nanocomposite obtained by in situ radiolytic method. Radiat. Phys. Chem. 2011, 80, 1208–1215. [CrossRef] 16. Krklješ, A.; Nedeljkovi´c,J.M.; Kaˇcarevi´c-Popovi´c,Z.M. Fabrication of Ag-PVA hydrogel nanocomposite by γ-irradiation. Polym. Bull. 2007, 58, 271–279. [CrossRef] 17. Spasojevi´c, J.; Radosavljevi´c, A.; Krsti´c, J.; Jovanovi´c, D.; Spasojevi´c, V.; Kalagasidis-Kruši´c, M.; Kaˇcarevi´c-Popovi´c,Z. Dual responsive antibacterial Ag-poly(N-isopropylacrylamide/itaconic acid) hydrogel nanocomposites synthesized by gamma irradiation. Eur. Polym. J. 2015, 69, 168–185. [CrossRef]

18. Abedini, A.; Daud, A.R.; Hamid, M.A.A.; Othman, N.K. Radiolytic formation of Fe3O4 nanoparticles: Influence of radiation dose on structure and magnetic properties. PLoS ONE 2014, 9, e90055. [CrossRef] 19. Ekoko, G.B.; Lobo, J.K.-K.; Mvele, O.M.; Muswema, J.L.; Yamambe, J.-F.S.; Mangwala, P.K. Gamma irradiation

inducing the synthesis of magnetic Fe3O4 nanorod particles in alkaline medium. Int. J. Mater. Sci. Appl. 2014, 3, 339–343. [CrossRef] 20. Goti´c,M.; Jurkin, T.; Musi´c,S. Factors that may influence the micro-emulsion synthesis of nanosize magnetite particles. Colloid Polym. Sci. 2007, 285, 793–800. [CrossRef] 21. Goti´c,M.; Jurkin, T.; Musi´c,S. From iron(III) precursor to magnetite and vice versa. Mater. Res. Bull. 2009, 44, 2014–2021. [CrossRef] 22. Jurkin, T.; Zadro, K.; Goti´c,M.; Musi´c,S. Investigation of solid phase upon γ-irradiation of ferrihydrite-ethanol suspension. Radiat. Phys. Chem. 2011, 80, 792–798. [CrossRef] 23. Jurkin, T.; Štefani´c,G.; Draži´c,G.; Goti´c, M. Synthesis route to δ-FeOOH nanodiscs. Mater. Lett. 2016, 173, 55–59. [CrossRef] 24. Jurkin, T.; Goti´c,M.; Štefani´c,G.; Puci´c,I. Gamma-irradiation synthesis of iron oxide nanoparticles in the presence of PEO, PVP or CTAB. Radiat. Phys. Chem. 2016, 124, 75–83. [CrossRef] 25. Mari´c,I.; Štefani´c,G.; Goti´c,M.; Jurkin, T. The impact of dextran sulfate on the radiolytic synthesis of magnetic iron oxide nanoparticles. J. Mol. Struct. 2019, 1183, 126–136. [CrossRef] 26. Sutherland, T.I.; Sparks, C.J.; Joseph, J.M.; Wang, Z.; Whitaker, G.; Sham, T.K.; Wren, J.C. Effect of ferrous concentration on the kinetics of radiation-induced iron-oxide nanoparticle formation and growth. Phys. Chem. Chem. Phys. 2017, 19, 695–708. [CrossRef]

27. Wang, S.; Xin, H.; Qian, Y. Preparation of nanocrystalline Fe3O4 by γ-ray radiation. Mater. Lett. 1997, 33, 113–116. [CrossRef] 28. Yakabuskie, P.A.; Joseph, J.M.; Keech, P.; Botton, G.A.; Guzonas, D.; Wren, J.C. Iron oxyhydroxide colloid formation by gamma-radiolysis. Phys. Chem. Chem. Phys. 2011, 13, 7198–7206. [CrossRef] 29. Belloni, J.; Mostafavi, M.; Remita, H.; Marignier, J.-L.; Delcourt, M.-O. Radiation-induced synthesis of mono- and multi-metallic clusters and nanocolloids. New. J. Chem. 1998, 22, 1239–1255. [CrossRef] 30. Jurkin, T.; Puci´c,I. Poly(ethylene oxide) irradiated in the solid state, melt and aqueous solution—A DSC and WAXD study. Radiat. Phys. Chem. 2012, 81, 1303–1308. [CrossRef] 31. Sava¸s,H.; Güven, O. Gelation, swelling and water vapor permeability behavior of radiation synthesized poly(ethylene oxide) hydrogels. Radiat. Phys. Chem. 2002, 64, 35–40. [CrossRef] 32. Ulanski, P.; Rosiak, J.M.; Zainuddin, A. Pulse radiolysis of poly(ethylene oxide) in aqueous solutions. I. Formation of macroradicals. Radiat. Phys. Chem. 1995, 46, 913–916. [CrossRef] 33. Ulanski, P.; Rosiak, J.M.; Zainuddin, A. Pulse radiolysis of poly(ethylene oxide) in aqueous solutions. II. Decay of macroradicals. Radiat. Phys. Chem. 1995, 46, 917–920. [CrossRef] 34. Jiang, C.Z.; Yang, S.Y.; Gan, N.; Pan, H.C.; Liu, H. A method for determination of [Fe3+]/[Fe2+] ratio in superparamagnetic iron oxide. J. Magn. Magn. Mater. 2017, 439, 126–134. [CrossRef] 35. Mari´c,I.; Goti´c,M.; Štefani´c,G.; Pustak, A.; Jurkin, T. γ-irradiation generated ferrous ions affect the formation of magnetite and feroxyhyte. Radiat. Phys. Chem. 2020, 170, 108648. [CrossRef] Nanomaterials 2020, 10, 1823 24 of 25

36. Mari´c,I.; Draži´c,G.; Štefani´c,G.; Zadro, K.; Goti´c,M.; Jurkin, T. Characterization of radiolytically synthesized feroxyhyte and magnetite nanoparticles. Mater. Charact. 2020, 159, 110038. [CrossRef] 37. Kamnev, A.A.; Tugarova, A.V. Sample treatment in Mössbauer spectroscopy for protein-related analyses: Nondestructive possibilities to look inside metal-containing biosystems. Talanta 2017, 174, 819–837. [CrossRef] 38. Lopez-Herrera, M.E.; Greneche, J.M.; Varret, F. Analysis of the Mössbauer quadrupole spectra of some amorphous fluorides. Phys. Rev. B 1983, 28, 4944. [CrossRef] 39. Vandenberghe, R.E.; De Grave, E. Application of Mössbauer spectroscopy in Earth Sciences. In Mössbauer Spectroscopy Tutorial Book, 1st ed.; Yoshida, Y., Langouche, G., Eds.; Springer: Berlin, Germany, 2013; pp. 91–186. 40. Gracien, E.B.; Ruimin, Z.; Kanza, L.K.; Lopaka, I. Effects of pH on the morphology of iron oxides synthesized under gamma-irradiation. J. Radioanal. Nucl. Chem. 2006, 270, 473–478. [CrossRef] 41. Spinks, J.W.T.; Woods, R.J. Introduction to Radiation Chemistry, 3rd ed.; John Wiley & Sons: New York, NY, USA, 1990; p. 262. 42. Spotheim-Maurizot, M.; Mostafavi, M.; Douki, T.; Belloni, J. Radiation Chemistry. From Basics to Application in Material and Life Sciences; EDP Sciences: Les Ulis, France, 2008; pp. 120–121. 43. Swiatla-Wojcik, D. Computation of the effect of pH on spur chemistry in water radiolysis at elevated temperatures. Nukleonika 2008, 53, S31–S37. 44. Matheson, M.S.; Mamou, A.; Silverman, J.; Rabani, J. Reaction of hydroxyl radicals with polyethylene oxide in aqueous solution. J. Phys. Chem. 1973, 77, 2420–2424. [CrossRef] 45. Dey, G.R.; Remita, H.; Mostafavi, M. Radiolytic reduction of Fe(II) in 2-propanol. Chem. Phys. Lett. 2006, 431, 83–87. [CrossRef] 46. Le Caër, S.; Rotureau, P.; Brunet, F.; Charpentier, T.; Blain, G.; Renault, J.P.; Mialocq, J.-C. Radiolysis of confined water: Hydrogen production at a high dose. Chem. Phys. Chem. 2005, 6, 2585–2596. [CrossRef] [PubMed] 47. Minkova, L.; Stamenova, R.; Tsvetanov, C.; Nedkov, E. Structural studies of radiation-crosslinked poly(ethylene oxide). J. Polym. Sci. Part. B 1989, 27, 621–642. [CrossRef] 48. Wypych, G. Handbook of Fillers, 2nd ed.; ChemTec Publishing: Toronto, ON, Canada, 2000; pp. 485–495, 501–505. 49. Karao˘glu,E.; Kavas, H.; Baykal, A.; Toprak, M.S.; Sözeri, H. Effect of hydrolyzing agents on the properties of

poly(ethylene glycol)-Fe3O4 nanocomposite. Nano Micro Lett. 2011, 3, 79–85. [CrossRef] 50. Mathur, S.; Moudgil, B.M. Adsorption Mechanism of PEO on oxide surfaces. J. Colloid Interface Sci. 1997, 196, 92–98. [CrossRef] 51. Weeber, R.; Kantorovich, S.; Holm, C. Ferrogels crosslinked by magnetic particles: Field-driven deformation and elasticity studied using computer simulations. J. Chem. Phys. 2015, 143, 154901. [CrossRef] 52. Popescu, E.D.; Zaharescu, T.; Gavrila,ˇ G. Insulation properties of degrading LDPE modified with metallic oxides. Optoelectr. Adv. Mater. Rapid Commun. 2011, 5, 1056–1061. 53. Davenas, J.; Stevenson, I.; Celette, N.; Cambon, S.; Gardette, J.L.; Rivaton, A.; Vignoud, L. Stability of polymers under ionizing radiation. The many faces of radiation interaction with polymers. Nucl. Instrum. Methods Phys. Res. Sect. B 2002, 191, 653–661. [CrossRef] 54. Agrawal, S.K.; Sanabria-DeLong, N.; Tew, G.N.; Bhatia, S.R. Nanoparticle-reinforced associative network hydrogels. Langmuir 2008, 24, 13148–13154. [CrossRef] 55. Criado-Gonzalez, M.; Corbella, L.; Senger, B.; Boulmedais, F.; Hernández, R. Photoresponsive nanometer-scale iron alginated hydrogels: A study of gel-sol transition using quartz crystal microbalance. Langmuir 2019, 35, 11397–11405. [CrossRef][PubMed] 56. Peng, F.; Li, G.; Liu, X.; Wu, S.; Tong, Z. Redox-responsive gel-sol/sol-gel transition in poly(acrylic acid) aqueous solution containing Fe(III) ions swiched by light. J. Am. Chem. Soc. 2008, 130, 16166–16167. [CrossRef][PubMed] 57. Yang, X.; Zhu, Z.; Chen, X. Thermal and rheological properties of poly(vinyl alcohol) and water-soluble chitosan hydrogels prepared by combination of γ-ray irradiation and freeze thawing. J. Appl. Polym. Sci. 2008, 109, 3825–3830. [CrossRef] Nanomaterials 2020, 10, 1823 25 of 25

58. Van Brederode, R.A.; Rodriguez, F.; Cocks, G.C. Crosslinking poly(ethylene oxide) in dilute solutions by gamma rays. J. Appl. Polym. Sci. 1968, 12, 2097–2104. [CrossRef] 59. Fechine, G.J.M.; Barros, J.A.G.; Alcântara, M.R.; Catalani, L.H. Fluorescence polarization and rheological studies of the poly(N-vinyl-2-pyrrolidone) hydrogels produced by UV radiation. Polymer 2006, 47, 2629–2633. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).