Mechanochemical synthesis of magnetically responsive materialsfrom non-magnetic precursors

Ivo Safarik a,b,n, Katerina Horska a, Kristyna Pospiskova b, Jan Filip b, Mirka Safarikova a

a Department of Nanobiotechnology, Institute of Nanobiology and Structural Biology of GCRC, Na Sadkach 7, 370 05 Ceske Budejovice, Czech Republicb Regional Centre of Advanced Technologies and Materials, Palacky University, Slechtitelu 11, 783 71 Olomouc, Czech Republic

a r t i c l e i n f o

Article history:Received 30 October 2013Accepted 7 April 2014Available online 15 April 2014

Keywords:MechanochemistryMagnetic materialsMagnetic adsorbentsMagnetic carriers

a b s t r a c t

Mechanochemical synthesis of various types of magnetically responsive materials from non-magneticpowdered precursors has been developed. The preparation is based on the mechanochemical conversionof ferrous and ferric ions at the presence of alkaline hydroxide into magnetic iron oxides nanoparticles(maghemite identified by XRD measurements). The presence of powdered nonmagnetic materialsduring the mechanochemical process led to the efficient deposition of magnetic nanoparticles on thesurface of the treated materials in the form of individual nanoparticles and their aggregates. Theprepared magnetically responsive materials have been used as adsorbents for xenobiotics removal andas a carrier for enzymes immobilization.

& 2014 Elsevier B.V. All rights reserved.

1. Introduction

Magnetic nano- and microparticles have attracted an increasinginterest in various fields including nanoscience, nanotechnology,biosciences, biotechnology and environmental technology. Manychemical procedures have been used to synthesize magneticparticles, such as classical co-precipitation, reactions in con-strained environments (e.g. microemulsions), sol–gel syntheses,sonochemical reactions, hydrothermal reactions, hydrolysis andthermolysis of precursors, flow injection syntheses, electrospraysynthesis and microwave synthesis [1,2].

Recently mechanochemical procedures have been used tosynthesize magnetic iron oxides and ferrites nanoparticles [3–5].Mechanochemistry represents one of several ways of chemicalactivation. In solid-state mechanochemistry, nonthermal chemicalreactions occur because of the deformation and fracture of solids,which are technically induced by milling or grinding of thematerials. During this process the mechanical energy induceschemical reactions and phase transformations [6].

Up to now, mechanochemical synthesis has been employedmainly for the production of individual iron oxides and ferrites.However, mechanochemistry can be successfully used also for thepreparation of magnetically responsive materials from originally

nonmagnetic powdered precursors. In this paper we present avery simple, generally applicable procedure for the preparation ofmagnetic materials from variety of inorganic, organic and biologi-cal precursors, together with the illustration of their possibleapplications as adsorbents and enzyme carriers. Mechanochemicalpostmagnetization can be very useful for smart magnetic mod-ification of diverse non-magnetic materials.

2. Materials and methods

Materials: FeCl3. 6H2O, FeCl2. 4H2O, montmorillonite, halloysite,Candida rugosa lipase (EC 3.1.1.3), 1,4-butanediol diglycidyl ether(BDDE), sodium (meta)periodate, 1,10-carbonyldiimidazole (CDI),4-nitrophenyl butyrate, dimethyl sulfoxide, Bismarck brown Y andsodium acetate were purchased from Sigma-Aldrich, USA. Micro-crystalline cellulose, safranin O and 4-nitrophenol were fromLachema, Czech Republic, while the common chemicals were fromLach-Ner, Czech Republic. Finally powdered biological materialswith diameters below 1 mm (spruce sawdust, scales from grasscarp (Ctenopharyngodon idella), wheat straw, pistachio nut shells,peanut husks, oak acorns, spent coffee grounds), as well as potatostarch and pine pollen were obtained locally.

The natural ocherous sediment containing biogenic iron oxideswas collected using glass vessels from a water stream in CeskeBudejovice (Czech Republic); it was sieved through a 1 mm sieveto remove larger detrital fraction, then repeatedly washed withdeionized water and air dried at a temperature not exceeding50 1C.

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/matlet

Materials Letters

http://dx.doi.org/10.1016/j.matlet.2014.04.0450167-577X/& 2014 Elsevier B.V. All rights reserved.

n Corresponding author at: Department of Nanobiotechnology, Institute of Nano-biology and Structural Biology of GCRC, Na Sadkach 7, 370 05 Ceske Budejovice, CzechRepublic. Tel.: +420 387775608; fax: +420 385310249.

E-mail address: ivosaf@yahoo.com (I. Safarik).URL: http://www.nh.cas.cz/people/safarik (I. Safarik).

Materials Letters 126 (2014) 202–206

Mechanochemical synthesis of magnetically responsive materials:In the standard procedure a mixture of 1.35 g of FeCl3 � 6H2O(0.005 mol), 0.50 g of FeCl2 � 4H2O (0.0025 mol) and 4 g of sodiumchloride (inert material used to avoid particles agglomeration) wasgrounded in a mortar at room temperature for 10 min. Then,appropriate amount of target nonmagnetic powdered material(usually 1–2 g) was added and after thorough mixing the processcontinued for another 10 min. As the last step, powdered potas-sium hydroxide (1.22 g) was added and after mixing the grindingcontinued for 10 min. After KOH addition the mixture becamebrown. During the grinding process the material was scraped fromthe mortar wall occasionally. After finishing the mechanochemicalprocess, the magnetically modified material was thoroughlywashed with water (to remove soluble impurities and free ironoxides particles) and stored in a suspension, or it was air dried.

Structural characterization of iron oxide nanoparticles preparedby mechanochemical procedure: X-ray powder diffraction (XRD)patterns of selected samples were recorded on PANalytical X'PertPRO instrument in Bragg–Brentano geometry with Fe-filteredCuKα radiation (40 kV, 30 mA). The samples were inserted intoconventional front-loading cavity sample holder and scanned inthe 2θ range of 10–901 (step size 0.0171). The commercial stan-dards SRM640 (Si) and SRM660 (LaB6) from NIST were used for theevaluation of the line positions and instrumental line broadening,respectively. The acquired patterns were evaluated using theX'Pert HighScore Plus software (PANalytical) together with PDF-4þdatabase.

Adsorption of organic dyes on magnetically responsive compo-sites: 30 mg of selected magnetically modified adsorbents weremixed with 7.0 mL of water in a test tube. Then, 0.1–3 mL of stockwater solution of a tested dye (1 mg/mL) was added and the totalvolume of the solution was made up to 10.0 mL with water. Thesuspension was mixed on a rotary mixer (Dynal, Norway) for 3 h atroom temperature. The magnetic adsorbent was then separatedfrom the suspension using a magnetic separator (MPC-1 or MPC-6,Dynal, Norway) and the clear supernatant was used for thespectrophotometric measurement. The concentration of free(unbound) dye in the supernatant (Ceq) was determined from thecalibration curve. The amount of dye bound to the unit mass of theadsorbent (qeq) was calculated using the following formula:

qeq ¼ ðCtot�CeqÞ=3 ðmg=gÞ ð1Þ

where Ctot is the total (initial) concentration of dye (μg/mL) used inthe experiment. The value qeq was expressed in mg of adsorbeddye per 1 g of adsorbent. Equilibrium adsorption data were fit toLangmuir adsorption isotherms using SigmaPlot software.

Immobilization of lipase on magnetic cellulose particles: C. rugosalipase was immobilized on magnetically modified cellulose parti-cles. Three various agents were used for activation of hydroxylgroups present in the cellulose structure, namely sodium period-ate, butanediol diglycidyl ether and carbonyldiimidazole. To pre-pare samples, 30 mg of prepared magnetic cellulose particles werewashed with water and magnetically separated by NdFeB magnet.In the next step, particles were treated by the activating agent.Using 1,10-carbonyldiimidazole (CDI), 1.5 mL of 0.5% (w/v) solutionin dimethyl sulfoxide was added to particles. For the periodatemethod, 1.5 mL of 1% (w/v) solution of sodium (meta)periodate(NaIO4) in 0.1 M sodium acetate buffer pH 4 was utilized. Duringactivation by the epoxide method, 1.5 mL of 3% (v/v) solution of1,4-butanediol diglycidyl ether (BDDE) in 0.25 M NaOH was used.Particles were shaken with these activating agents on an auto-matic rotator (20 rpm) for 24 h at room temperature in the dark.After the modification procedure, particles were magneticallyseparated, supernatants were poured off and particles wererepeatedly washed with distilled water. Subsequently, 1.5 mL oflipase solution (1 mg/mL) in 50 mM potassium phosphate buffer,

pH 7.5, was added to the activated particles and shaken on auto-matic rotator (20 rpm) for 24 h at 4 1C. After the immobilizationprocedure, particles were magnetically separated, supernatantcontaining unbound enzyme was removed and particles withbound enzyme were repeatedly washed with buffer until noenzyme activity in the supernatant was detected. Magnetic cellu-lose particles with immobilized lipase were stored in buffer at 4 1C.

Lipase assay: Activity of lipase immobilized on magnetic cellu-lose particles was determined spectrophotometrically using0.5 mM 4-nitrophenyl butyrate (dissolved in ethanol) in 50 mMpotassium phosphate buffer, pH 7.5. Particles of magnetic cellulosewith attached lipase were stirred during the reaction in buffercontaining the substrate, then magnetically separated to thebottom of the cuvette to stop the reaction and increasing amountof yellow-colored 4-nitrophenol was measured at 405 nm.

Molar absorption (extinction) coefficient (ε) of the reactionproduct was determined spectrophotometrically for the specificmedium conditions used in this enzyme assay. Coefficient for4-nitrophenol in 50 mM potassium phosphate buffer, pH 7.5 at405 nm was 13815 L/mol cm.

Operational stability of immobilized lipase: Reusability of lipaseimmobilized on magnetic cellulose particles was tested as itsoperational stability; it was repeatedly used for 7 reaction cycles.Particles with attached lipase were washed with buffer betweeneach cycle. Activity of lipase was measured spectrophotometricallyas described previously. Residual activities of lipase after eachcycle were determined and compared taking the initial activity inthe first cycle as 100%.

Time stability of immobilized lipase: Particles of magneticcellulose with immobilized lipase were stored in the reactionbuffer at 4 1C for 30 days and percentage of residual enzymeactivity on the carrier was determined. Possible presence of lipasereleased from the support was tested during this time period bymeasuring the activity of free lipase in the supernatant.

3. Results and discussion

Mechanochemical synthesis of magnetic composite materials: Inmechanochemical synthesis, hydrated solid reactants, namelyferrous and ferric chlorides reacted with KOH during grinding ina mortar to form magnetite nanoparticles. To avoid agglomeration,the excess of sodium chloride was added to the precursors beforegrinding. The following reaction takes place [7]:

2FeCl3 �6H2O(s)þFeCl2 �4H2O(s)þ8KOH(s)-8KCl(s)þFe3O4(s)þ20H2O(g) (2)

Magnetite nanoparticles as the primary product are stable onlyin inert atmosphere [5]. However, in the described process which

Fig. 1. Appearance of original powdered oak acorns suspension (left), suspension ofoak acorns powder after magnetic modification (middle) and demonstration ofmagnetic separation of magnetically modified oak acorns powder (right).

I. Safarik et al. / Materials Letters 126 (2014) 202–206 203

is performed in air, the reaction-generated heat and air oxidationlead to the conversion of magnetite nanoparticles into maghemiteas follows [7]:

4 Fe3O4 (s)þO2 (g)-6 γ-Fe2O3 (s) (3)

The addition of appropriate nonmagnetic powder material tothe reaction mixture and grinding before the KOH addition led to

the homogeneous distribution of ferrous and ferric ions within thematerial matrix. Subsequent addition of KOH followed by grindingled to the formation of magnetite and maghemite nanoparticlesbound to the surface and pores of matrix according to Eqs. (2) and (3).The possible influence of the modified material and its chemicalcomposition may lead to the formation of the non-stoichiometricmaghemite during the grinding process. Anyway magnetic iron

Fig. 2. Scanning electron microscopy of native and magnetically modified materials. 1 – native potato starch (reproduced with permission from [13]); 2 – magnetic potatostarch; 3 – native pine pollen (reproduced with permission from [14]); 4 – magnetic pine pollen; 5 – native pistachio nut shells (reproduced with permission from [15]);6 – magnetic pistachio nut shells; 7 – native halloysite; 8 – magnetic halloysite.

I. Safarik et al. / Materials Letters 126 (2014) 202–206204

oxides nanoparticles bound to the treated material lead to theformation of stable magnetically responsive composite materials,which can be easily separated using a simple magnet or magneticseparator (Fig. 1). The magnetic iron oxide particles deposition isusually very stable, and during the subsequent washing steps onlya very low amount of particles were released and washed out.

The presence of magnetic iron oxides nanoparticles and theiraggregates on the surface of selected magnetically modifiedmaterials (magnetic potato starch, pine pollen, pistachio nut shellsand halloysite) can be clearly seen using scanning electron micro-scopy (Fig. 2). It can be expected that the strong binding ofmagnetic iron oxides particles to the surface of non-magneticmaterials has been achieved by a subtle balance of van der Waals,electrostatic and hydrophobic interactions both between themagnetic particles and the treated material surface and betweenthe adsorbed magnetic particles [8].

The structure of iron oxide particles in selected magneticallymodified materials (namely spruce sawdust) was identified usingXRD as maghemite γ-Fe2O3 (Fig. 3) with observed remnants ofsoluble salts. The very broad diffraction lines imply nanocrystallinecharacter of the maghemite particles – the size of X-ray coherentdomains (i.e., corresponding approximately to mean particle size)calculated according to the Scherrer formula equals to 4 nm [9].

Adsorption of organic dyes on magnetic composites: Recently, largenumber of low cost, easily available materials have been tested andused as adsorbents for xenobiotics removal [10]. Magnetic modifica-tion of such adsorbents simplifies their separation from the treatedsolutions and suspensions. That is why adsorption properties ofseveral materials magnetized with described mechanochemicalprocedure (namely powdered magnetically responsive montmorillo-nite, grass carp scales, oak acorns and biogenic iron oxides) weretested with two water-soluble organic dyes, belonging to differentdye classes, namely Bismarck brown Y (diazo dye) and safranin O(safranin dye). It was shown in preliminary experiments that theadsorption of the tested dyes reached equilibrium in approximately60–120 min. Incubation time of 3 h was used for adsorption experi-ments. The equilibrium adsorption isotherms for both tested dyesand four magnetic adsorbents (prepared by addition of 1 g ofmaterial during the standard mechanochemical procedure) areshown in Fig. 4. The experimental data were analyzed by means ofnon-linear regression calculation using SigmaPlot software and itwas shown that they follow the Langmuir isotherm equation. TheLangmuir model is valid for monolayer adsorption onto a surfacewith a finite number of identical sites. The well-known expression

for the Langmuir model is given by

qeq ¼QmaxbCeq

1þbCeqð4Þ

where qeq (expressed in mg/g) is the amount of the adsorbed dye perunit mass of magnetic adsorbent and Ceq (expressed in mg/L) is theunadsorbed dye concentration in solution at equilibrium. Qmax

(expressed in mg/g) is the maximum amount of the dye per unitmass of adsorbent to form a complete monolayer on the surfacebound at high dye concentration and b is a constant related to theaffinity of the binding sites (expressed in L/mg) [11].

Because the adsorption of the tested dyes can be successfullydescribed by the Langmuir isotherm, it is possible to calculate themaximum adsorption capacity which is a very important parameterdescribing the adsorption process (Table 1). In the case of two dyesand four adsorbents tested, the highest value was found for Bismarckbrown Y and grass carp scales (43.3 mg/g), while the lowest valuewas obtained for safranin O and magnetic oak acorns (12.9 mg/g).The maximum adsorption capacities of the developed materials arenot so high as observed in the case of some other alternativemagnetic biosorbents [12]. The possible reason for lower adsorptionis the high coverage of the magnetically modified material surfacewith magnetic iron oxides (see Fig. 2). In order to test this possibility,two batches of magnetic montmorillonite were prepared, differing bythe amount of native montmorillonite added during the standardmechanochemical procedure (one or two grams); higher amount ofnative material during the mechanochemical process results in lowersurface coverage with magnetic iron oxides. The maximum adsorp-tion capacities were 38.6 and 52.4 mg/g for one gram and two grambatches, respectively.

Immobilization of lipase on magnetic cellulose particles: Selectedmagnetically modified materials can be used as carriers forimmobilization of enzymes and other important biologically activecompounds. C. rugosa lipase was immobilized on magnetic cellu-lose particles prepared by mechanochemical method. Threereagents, namely sodium periodate, butanediol diglycidyl etherand carbonyldiimidazole were used for activation of hydroxylgroups present in cellulose structure. Table 2 shows the activityof immobilized lipase on one milligram of magnetic cellulose.

During the study of operational stability, lipase immobilized onmagnetic cellulose using carbonyldiimidazole activation retained98% of initial activity after 7 cycles. In the case of sodium periodate

Fig. 3. Typical X-ray powder diffraction pattern of magnetically modified material -spruce sawdust (top) and separate maghemite nanoparticles prepared by the samemechanochemical procedure (bottom). Vertical lines indicate theoretical positionsof most intense maghemite diffractions (taken from PDF card No.: 00-039-1346),asterisks indicate diffractions of remnants of soluble salts.

Table 1Maximum adsorption capacities (Qmax; mg/g) of four magnetically modifiedadsorbents (one gram of material used during the mechanochemical process) forthe tested dyes (Bismarck brown Y and safranin O). Langmuir adsorption isothermwas used for calculation.

Magnetic adsorbents Bismarck brown Y Safranin O

Grass carp scales Qmax¼43.3 Qmax¼30.3Oak acorns Qmax¼20.9 Qmax¼12.9Montmorillonite Qmax¼38.6 Qmax¼41.4Biogenic iron oxides Qmax¼36.4 Qmax¼15.9

Table 2Activity of lipase immobilized on magnetic cellulose particles using three immo-bilization methods. CDI, 1,10-carbonyldiimidazole; BDDE, 1,4-butanediol diglycidyl ether.

Immobilization method Activity of enzyme onmagnetic cellulose particles (nkat/mg)

CDI 1.48NaIO4 4.09BDDE 4.55

I. Safarik et al. / Materials Letters 126 (2014) 202–206 205

and butanediol diglycidyl ether activation procedures immobilizedlipase retained 90 and 83% of initial activity after 7 cycles, respec-tively (Fig. 5). After storage at 4 1C for the time period of 1 month,lipase immobilized using carbonyldiimidazole retained approxi-mately 97%, using sodium periodate 80% and using butanedioldiglycidyl ether 70% of its initial activity, respectively.

4. Conclusions

A fast and simple mechanochemical conversion of powderednonmagnetic materials into their magnetic derivatives has beendeveloped. Magnetically responsive composites have been used as

adsorbents for organic xenobiotics removal and as a biocompatiblecarrier for enzymes immobilization. Low-cost, efficient and bio-compatible magnetic composites can substantially improve andsimplify wide range of biotechnology and environmental technol-ogy processes.

Acknowledgments

The authors thank Dr. Dalibor Jancik (RCPTM, Palacky Univer-sity, Olomouc, Czech Republic) and Laboratory of Electron Micro-scopy (Biological Centre, Ceske Budejovice, Czech Republic) formaking SEM images. This research was supported by the GrantAgency of the Czech Republic (Project no. 13-13709S), by theResearch Project LH12190 (Ministry of Education, Youth and Sportsof the Czech Republic), by Technology Agency of the CzechRepublic (Competence Centres, Project no. TE01020218) and bythe Operational Program Research and Development for Innova-tions – European Development Fund (CZ.1.05/2.1.00/03.0058) ofthe Ministry of Education, Youth and Sports of the Czech Republic.

References

[1] Safarik I, Safarikova M. Magnetic nanoparticles for in vitro biological andmedical applications: an overview. In: Thanh NTK, editor. Magnetic Nanopar-ticles: From Fabrication to Biomedical and Clinical Applications. Boca Raton,FL 33487-2742, CRC Press/Taylor and Francis; 2012. p. 215–42.

[2] Laurent S, Forge D, Port M, Roch A, Robic C, Elst LV, et al. Magnetic iron oxidenanoparticles: synthesis, stabilization, vectorization, physicochemical charac-terizations, and biological applications. Chem Rev 2008;108:2064–110.

[3] Bellusci M, Aliotta C, Fiorani D, La Barbera A, Padella F, Peddis D, et al.Manganese iron oxide superparamagnetic powder by mechanochemicalprocessing, nanoparticles functionalization and dispersion in a nanofluid.J Nanoparticle Res 2012:14 (Article No. 904).

[4] Das D, Sureshkumar MK, Koley S, Mithal N, Pillai CGS. Sorption of uranium onmagnetite nanoparticles. J Radioanal Nucl Chem 2010;285:447–54.

[5] Lin CR, Chu YM, Wang SC. Magnetic properties of magnetite nanoparticlesprepared by mechanochemical reaction. Mater Lett 2006;60:447–50.

[6] Beyer MK, Clausen-Schaumann H. Mechanochemistry: the mechanical activa-tion of covalent bonds. Chem Rev 2005;105:2921–48.

[7] Lu J, Yang SH, Ng KM, Su CH, Yeh CS, Wu YN, et al. Solid-state synthesis ofmonocrystalline iron oxide nanoparticle based ferrofluid suitable for magneticresonance imaging contrast application. Nanotechnology 2006;17:5812–20.

[8] Saito T, Koopal LK, Nagasaki S, Tanaka S. Adsorption of heterogeneouslycharged nanoparticles on a variably charged surface by the extended surfacecomplexation approach: charge regulation, chemical heterogeneity, and sur-face complexation. J Phys Chem B 2008;112:1339–49.

[9] Patterson AL. The Scherrer formula for X-Ray particle size determination. PhysRev 1939;56:978–82.

[10] Srinivasan A, Viraraghavan T. Decolorization of dye wastewaters by biosor-bents: a review. J Environ Manag 2010;91:1915–29.

[11] Safarik I, Rego LFT, Borovska M, Mosiniewicz-Szablewska E, Weyda F, Safar-ikova M. New magnetically responsive yeast-based biosorbent for the efficientremoval of water-soluble dyes. Enzyme Microb Technol 2007;40:1551–6.

[12] Safarik I, Horska K, Safarikova M. Magnetically responsive biocomposites forinorganic and organic xenobiotics removal. In: Kotrba P, Mackova M, Macek T,editors. Microbial Biosorption of Metals. Dordrecht, Heidelberg, London,New York: Springer; 2011. p. 301–20.

[13] Varatharajan V, Hoover R, Li JH, Vasanthan T, Nantanga KKM, Seetharaman K,et al. Impact of structural changes due to heat-moisture treatment at differenttemperatures on the susceptibility of normal and waxy potato starchestowards hydrolysis by porcine pancreatic alpha amylase. Food Res Int2011;44:2594–606.

[14] Nakagawa T, Edouard JL, de Beaulieu JL. A scanning electron microscopy (SEM)study of sediments from Lake Cristol, southern French Alps, with specialreference to the identification of Pinus cembra and other Alpine Pinus speciesbased on SEM pollen morphology. Rev Palaeobot Palynol 2000;108:1–15.

[15] Yang T, Lua AC. Characteristics of activated carbons prepared from pistachio-nut shells by physical activation. J Colloid Interface Sci 2003;267:408–17.

Fig. 4. Equilibrium adsorption isotherms of Bismark brown Y (top) and safranin O(bottom) on the tested magnetically modified adsorbents (one gram of materialused during the mechanochemical process). Ceq: equilibrium liquid-phase concen-tration of the unadsorbed (free) dye (mg/L); qeq: equilibrium solid-phase concen-tration of the adsorbed dye (mg/g). ( ♦ ) biogenic iron oxides; ( � ) oak acorns;( ▲ ) montmorillonite; ( ■ ) grass carp scales.

Fig. 5. Operational stability of lipase immobilized on magnetic cellulose particlesusing three immobilization methods. Chart demonstrates relative residual activityof immobilized enzyme (%) after each cycle. ♦ - CDI activation; ■ - NaIO4 activation;▲ - BDDE activation. CDI - 1,10-carbonyldiimidazole; BDDE - 1,4-butanedioldiglycidyl ether.

I. Safarik et al. / Materials Letters 126 (2014) 202–206206