Materials Science and Engineering C 29 (2009) 2226–2233

Contents lists available at ScienceDirect

Materials Science and Engineering C

j ourna l homepage: www.e lsev ie r.com/ locate /msec

Preparation and studies on surface modifications of calcium-silico-phosphateferrimagnetic glass-ceramics in simulated body fluid

K. Sharma a, A. Dixit a, Sher Singh b, Jagannath a, S. Bhattacharya a, C.L. Prajapat a, P.K. Sharma c, S.M. Yusuf b,A.K. Tyagi d, G.P. Kothiyal a,⁎a Technical Physics and Prototype Engineering Division, Bhabha Atomic Research Centre, Mumbai 400 085, Indiab Solid State Physics Division, Bhabha Atomic Research Centre, Mumbai 400 085, Indiac Analytical Chemistry Division, Bhabha Atomic Research Centre, Mumbai 400 085, Indiad Chemistry Division, Bhabha Atomic Research Centre, Mumbai 400 085, India

⁎ Corresponding author. Tel.: +91 22 25595652; fax:E-mail address: gpkoth@barc.gov.in (G.P. Kothiyal).

0928-4931/$ – see front matter © 2009 Published by Edoi:10.1016/j.msec.2009.05.009

a b s t r a c t

a r t i c l e i n f o

Article history:Received 18 February 2009Received in revised form 24 April 2009Accepted 14 May 2009Available online 22 May 2009

Keywords:Glass ceramicXPSMagnetiteSimulated body fluid (SBF)Bioactivity

The structure and magnetic behaviour of 34SiO2–(45−x) CaO–16P2O5–4.5 MgO–0.5 CaF2−x Fe2O3 (wherex=5, 10, 15, 20 wt.%) glasses have been investigated. Ferrimagnetic glass-ceramics are prepared by meltquench followed by controlled crystallization. The surface modification and dissolution behaviour of theseglass-ceramics in simulated body fluid (SBF) have also been studied. Phase formation andmagnetic behaviourhave been studied using XRD and SQUID magnetometer. The room temperature Mössbauer study has beendone to monitor the local environment around Fe cations and valence state of Fe ions. X-ray photoelectronspectroscopy (XPS) was used to study the surfacemodification in glass-ceramics when immersed in simulatedbody fluid. Formation of bioactive layer in SBF has been ascertained using X-ray photoelectron spectroscopy(XPS) and scanning electron microscopy (SEM). The SBF solutions were analyzed using an absorptionspectrophotometer. The magnetic measurements indicated that all these glasses possess paramagneticcharacter and the [Fe2+/Fe3+] ions ratio depends on the composition of glass and varied with Fe2O3

concentration in glassmatrix. In glass-ceramics saturationmagnetization increaseswith increase in amount ofFe2O3. The nanostructure of hematite andmagnetite is formed in the glass-ceramicswith 15 and 20wt.% Fe2O3,which is responsible for the magnetic property of these glass-ceramics. Introduction of Fe2O3 induces severalmodifications at the glass-ceramics surface when immersed in SBF solution and thereby affecting the surfacedissolution and the formation of the bioactive layer.

© 2009 Published by Elsevier B.V.

1. Introduction

Calcium-silico-phosphate glasses have potential as implant mate-rials for human body because of their bioactivity and biocompatibility.Hench [1] has reported the first bioactive glass having composition(wt.%) 45% SiO2, 24.5%Na2O, 24.5% CaO and 6% P2O5 commonly knownas 45S5. The bioactivity of these materials is composition dependent.Addition of alumina tends to decrease the bioactivity of these glasses[2]. These glasses and glass-ceramics having Fe2O3 show an importantapplication in cancer treatment by elimination of cancerous cells inbones; by means of hyperthermia [3]. The magnetic properties arisefrom magnetite [Fe3O4] that is produced from the Fe2O3. When thismaterial is placed in the region of the tumor and is subjected to analternating magnetic field, heat is generated by hysteretic losses [4].The tumor is effectively heated and the temperature locally rises to 42–45 °C. As a result, the cancerous cells perish while the healthy ones

+91 22 25505296.

lsevier B.V.

survive [5–7]. Synthesis of glass-ceramics in SiO2–CaO–Fe2O3, SiO2–

CaO–Fe2O3–B2O3–P2O5, SiO2–Al2O3–Fe2O3–P2O5–Li2O and CaO–SiO2–

P2O5–Na2O–Fe2O3 bioglasses, have been reported [8,9]. However,the distribution and the bonding environment of Fe2O3 on theseglasses and glass-ceramics have not been studied in great details.Since the magnetic properties of the material depend on the envi-ronment of Fe, therefore the knowledge of structure and oxidationstates of iron ions is beneficial for synthesis of magnetic glass andglass-ceramic.

Since these materials are in contact with living tissues whenimplanted in the body, they should not elicit any harmful responsefrom the host tissues. Therefore, the surface chemistry of materialsneeds to meet the requirements of host tissues. In fact, surface of thematerial has a critical influence on the biological response, therefore,most of the applications of these biomaterials are dictated by the wayinwhich a givenmaterial interacts with body fluids. Therefore, a studyof the surface interactions with body fluid is needed to improve theunderstanding of chemistry and physics taking place on surfaces/interfaces of glasses/glass-ceramics. In this regard, XPS spectroscopyis useful in understanding the surface interactions with body fluids as

Table 1Base glass compositions (nominal) in wt.%.

Sample SiO2 P2O5 CaO MgO Fe2O3 CaF2

FB5 34 16 40 4.5 5 0.5FB6 34 16 35 4.5 10 0.5FB7 34 16 30 4.5 15 0.5FB8 34 16 25 4.5 20 0.5

2227K. Sharma et al. / Materials Science and Engineering C 29 (2009) 2226–2233

it provides information on the first 50–100 A° of the sample surface[10].

In the present work, we report preparation and a systematic studyon magnetic behaviour and surface properties of 34SiO2–(45−x) CaO–16P2O5–4.5 MgO–0.5 CaF2−x Fe2O3 (where x=5,10, 15, 20 wt.%) glassand glass-ceramics. This system is of particular interest due to itsmagnetic and bioactive properties. We have studied the effect of Fe2O3

on magnetic and bioactivity related properties in SBF. The surfacemodifications of these samples as a function of exposure time in SBFwere investigated by XPS and SEM/EDX. The dissolution behaviour inthe solution is explained on the basis of surface reactions.

2. Experimental procedure

2.1. Preparation of glass and glass-ceramics

Base glasses of compositions as given in Table 1 were prepared bymelt quench technique. About 100 g batches were prepared bymixingreagent grade SiO2, CaCO3, NH4H2PO4, MgCO3, Fe2O3, and CaF2. Thecharge was calcined at a maximum of 900 °C for 12 h, holding atintermediate temperature for 6–8 h, decided by the decompositiontemperatures of various precursors. Melting was carried out in acovered Pt-10% Rh crucible at 1450–1500 °C in a lowering and raisinghearth furnace. The melt was held for 2 h at this temperature forhomogenization and then poured into a graphite mould followed byannealing at 500 °C. The base glass was powdered in a ball mill andpelletized using a hydraulic press. They were converted into glass-ceramics (hereafter called FBC) through controlled heat treatment.Glass-ceramics FBC5, FBC6, FBC7 and FBC8 with iron concentration 5,10, 15, 20 wt.% respectively, were heat treated at 1000 °C for 6 h.

2.2. Structural and magnetic study of glass and glass-ceramics

X-ray diffraction (XRD) of the powder sample was carried out onPhilips PW 1710 X-ray diffractometer. The magnetic response versusapplied magnetic field H was measured at room temperature, with |H|≤5 kOe using a Superconducting Quantum Interference Device

Fig. 1. XRD patterns of glass-ceramics ha

(SQUID) magnetometer. These data have been analyzed to obtainthe saturation magnetization (Ms), remnant magnetization (Mr), andcoercive field (Hc) for each material. Mössbauer spectra have beenobtained using a spectrometer operated in constant accelerationmode. The source employed is 57Co in Rh matrix of strength 50 mCi.The calibration of the velocity scale is done using iron metal foil. TheMössbauer spectra are fitted with appropriate paramagnetic doubletsand magnetic sextets using least square fit program. The ratio of Fe2+

and Fe3+ ions has been determined from the relative areas obtained bycomputer fitting of experimental spectra.

2.3. In-vitro bioactivity analysis in SBF

The interactions of the samples with simulated body fluid werestudied. The pellets were immersed in SBF solution for 1–4 weeks,incubated at 37.4 °C. Samples were removed periodically and theirsurfaces were analyzed using XPS technique. For XPS measurement,samples were mounted on the specimen holder using silver paste. Theconducting path was provided from bottom to the top surface of thesample by silver paste, to avoid the surface charging effect. The samplechamber was then evacuated to a vacuum better than 1×10−9 Torr.The sample was excited by Mg-Kα radiations (hυ=1254.6 eV), photo-electron spectra were analyzed using a VG make CLAIM 2 analyzersystem. The core level peaks for the constituent elements are identifiedand marked on the spectra.

Quantitative evaluation of chemical composition was made for theconstituent elements of the immersed glass-ceramics by estimatingthe peak area and using the calculated cross sections (p) as sensitivityfactors [11]. After the sampleswere removed, changes in concentrationof Ca, P, Si and Fe in SBF solutions were measured using an absorptionspectrophotometer (Atomic Absorption Spectrophotometer ChemitoAA 203). The pH of the SBF solutions was monitored periodicallyduring the experiment. The morphological analyses were carried outbymeans of SEM (Model: Tescan VegaMV 2300T/40) technique. Priorto mounting samples in analysis chamber of microscope, thin Auconducting coating was deposited on the sample surface to preventcharging effects.

3. Results and discussion

3.1. Structural and magnetic studies

The XRD patterns of heat treated samples are shown in Fig.1.Whenglasses FB6, FB7 and FB8 were heat treated at 1000 °C for 6 h, calciumphosphate (Ca2P2O7), wollastonite and magnetite were developed asmajor crystalline phases.

ving different Fe2O3 concentrations.

Fig. 2. M-H plots of the different glass-ceramics samples having different Fe2O3(x)concentrations.

Fig. 3. Mössbauer spectra for glass samples having different amounts of Fe2O3(x)recorded at room temperature.

2228 K. Sharma et al. / Materials Science and Engineering C 29 (2009) 2226–2233

The CaO has more affinity towards phosphate and Ca2P2O7 phaseformed on heat treatment. Magnetite has been formed in the glassmatrix by the reaction:

6Fe2O3→4Fe3O4 þ O2:

The extent of conversion of hematite (Fe2O3) into Fe3O4 depends onthe reducing environment produced by the presence of MgO, CaO inthe melt [12]. These oxides have the high affinity for oxygen andtherefore reduction of the Fe2O3 will depend on their relativequantities (composition) in the batch. Magnetite phase is responsiblefor the magnetic properties in the glass-ceramic samples.

M-H plots of the different glass-ceramic samples are shown inFig. 2. Magnetic parameters estimated from the M-H plot are given inTable 2. Saturation magnetization increased with increase in Fe2O3

concentration from 10 to 20 wt.%. The quantity of magnetic phasepresent in the glass-ceramic samples was determined from the ratioof saturation magnetization of sample with that of magnetite (Ms=92 emu/g) [13].

There is an increase in saturation magnetization with an increase inFe2O3 concentration, which is attributed to the development of mag-netite phase in the samples. The volume fraction of themagnetite phaseincreased, with the increase in iron concentration. This implies thatmore conversion of Fe2O3 (nonmagnetic phase) into Fe3O4 (magneticphase) has resulted in an increased value of saturation magnetization.

Table 3

3.1.1. 57Fe Mössbauer spectroscopyThe Mössbauer spectra for different glass samples are shown in

Fig. 3. It is found that these spectra were composed of two para-magnetic doublets. Related Mössbauer parameters like isomer shift(IS), quadrupole splitting (QS) determined from these spectra aregiven in Table 3. Typical values of Fe3+ tetrahedrally (Th) coordinatedto oxygen in silicate glass are in the range of 0.20–0.32 mm s−1

while for octahedrally (Oh) coordinated these values are in the rangeof 0.35–0.55 mm s−1. For Fe2+ these values are in the range of 0.90–0.95mms−1(tetrahedral) and 1.05–1.10mms−1(octahedral) [14]. The

Table 2Magnetic parameters estimated from M-H plots (Fig. 2).

Magnetic parameter FBC6 FBC7 FBC8

Saturation magnetization (Ms ((emu/g)) 0.0304 0.129 0.64Remenant magnetization (Mr (emu/g)) 0.0037 0.020 0.164Coercive field (Hc (Oe)) 211 167 103Magnetic phase (wt.%) .032 0.14 0.69Hysteresis area (erg/g) 121.8 513 2795

IS values indicate that in glass structure, Fe3+ and Fe2+ are present inoctahedral and tetrahedral coordination, respectively.

The presence of some iron ions in the Fe2+ state can be attributed tothe reduction of Fe2O3 by the presence of MgO, CaO in the melt [12].From Table 3, it is seen that in the glass containing 5 wt.% Fe2O3, nearly68% of the total iron entered as Fe3+ ions in octahedral coordinationi.e. entered in the glass as a modifier. The rest entered the glassnetwork as former. The iron ions in the octahedral coordinationincrease to 78% of the total iron ions, as Fe2O3 content increases to15 wt.%. These Fe3+ (Oh) ions increased at the expense of thetetrahedral coordinated ions (Fe2+ (Th)). When content of Fe2O3 wasraised to 20wt.%, the decrease in the Fe3+ (Oh) ions has beenobserved.This means that addition of Fe2O3 (xN15 wt.%) favours more of Fe2+

ions to substitute glass network as glass forming units.Table 3 shows the dependence of Fe2+/Fe3+ ions ratio along with

various parameters obtained fromMössbauer spectra on the composi-tion for different glass samples. From this, it is seen that the Fe2+/Fe3+

ions ratio depends on composition. It decreases with increasing Fe2O3

up to 15wt.% and then increases with increase in Fe2O3 to 20 wt.%. Theisomer shift value for Fe2+ ions of the different components firstdecreaseswith the increase in the Fe2O3 content at the expense of CaO,then increases with increase in Fe2O3 to 20 wt.%. This result can beattributed to the replacement of the iron (Fe2O3with xN15wt.%) in thenetwork as former thus causing a decrease in the s electron density atthe iron nuclei and hence an increase in the IS values. The increase of

Various parameters obtained from room temperature Mössbauer spectra of glasssamples.

Sample (Paramagnetic) Fe3+ (Paramagnetic) Fe2+ Fe2+/Fe3+QS IS RI Γ QS IS RI Γ

FB5 0.935 0.2916 68.16 0.7436 2.36 0.7274 31.84 0.6983 0.47FB6 0.99 0.2724 77.30 0.6543 2.492 0.7015 22.69 0.5296 0.29FB7 1.03 0.2717 78.89 0.6491 2.532 0.6888 21.11 0.5416 0.26FB8 0.89 0.344 57.09 0.6505 2.37 0.7864 42.90 0.6787 0.73

IS: Isomer shift (mm/s), QS: Quadrupole Splitting (mm/s), RI: relative intensity.Hint: Internal Field (kG), Γ: Line width (mm/s).

Fig. 4. Mössbauer spectra for glass-ceramics samples having different amounts ofFe2O3(x) recorded at room temperature.

Table 5Concentration (ppm) of different constituents (Si, Ca and P) of SBF solutions afterimmersing glass-ceramic samples for different durations.

Sample Si Ca P

FBC6 One week 14 1.5 83Two week 53 26 19

FBC7 One week 14 1.1 77Two week 17 13 10

FBC8 One week 18 0.8 79Two week 27 15 21

SBF 12 0.9 77

Fig. 5. pH of SBF solutions at different time periods during immersion of samples.

2229K. Sharma et al. / Materials Science and Engineering C 29 (2009) 2226–2233

QS up to 15 wt.% Fe2O3 attributed to an increase in the asymmetryaround the iron ions. Further, QS decreases with increase in Fe2O3

content to 20 wt.% as iron enters in network as former.The Mössbauer spectra for different glass-ceramics samples are

shown in Fig. 4 and relatedMössbauer parameters are given in Table 4.On heat treatment the samples (FBC) show magnetic and paramag-netic components due to Fe3+ and Fe2+ ions depending upon theircompositions. The glass-ceramics samples with 5 and 10 wt.% Fe2O3

show paramagnetic doublet arising from Fe3+ and Fe2+ ions. When15 wt.% Fe2O3 is added, two magnetic (Fe3+ in octahedral coordina-tion) and two paramagnetic components (Fe3+ in tetrahedral co-ordination) are observed. Further increasing Fe2O3 to 20 wt.%, similarspectra (two magnetic and two paramagnetic contributions) havebeen observed, though the contribution of magnetic component hasincreased to 48% in FBC8 from 35% in FBC7. Isomer shift values of theseglasses suggest that Fe3+ ions are in octahedral coordination inmagnetic component and in tetrahedral coordination in paramagneticcomponent.

Along with the doublet, the ratio of Fe2+ and Fe3+ varied with theincrease in iron content. This paramagnetic doublet can be assignedwithout ambiguity to hematite Fe2O3 [15]; this correlateswell with thestarting composition and XRD analysis of the sample showing thepresence of such a phase. The spectra of FBC7 and FBC8 glass-ceramicsshow the two paramagnetic doublets and two sextets. There seems to

Table 4Various parameters obtained from room temperature Mössbauer spectra of glass-ceramics samples.

Sample Com Magnetic (sextets) Paramagnetic (doublets)

Hint QS IS RI QS IS RI Γ

FBC5 Com1 0.9480 0.2412 58.60 0.5688Com2 1.5147 0.0545 41.39 0.6305

FBC6 Com1 0.9281 0.2633 51.74 0.4915Com2 1.5592 0.1143 48.25 0.4966

FBC7 Com1 440.37 0.07 0.2636 14.92 1.5413 0.1316 32.33 0.5230Com2 515.90 0.10 0.3727 21.03 0.8813 0.268 31.70 0.4470

FBC8 Com1 448.07 0.09 0.3248 20.44 1.5570 0.1307 14.94 0.5493Com2 515.12 0.11 0.3844 28.28 0.7660 0.2761 36.32 0.3683

Com: component.

be two kinds of Fe3+ ions in these glass-ceramics samples. The twosextets (see Table 4), can be associated to the nanostructure hematite[16], FeO Fe2O3 (which presents a sextet spectrum at room tempera-ture for single crystalline samples). The nanostructure hematite, andmagnetite formed are responsible formagnetic propertyof these glass-ceramics.

3.2. In-vitro bioactivity analysis in SBF

3.2.1. Chemical analysis of SBF solutionsChemical analyses of SBF solutions after immersion period of

1week and 2weeks are given in Table 5. It is observed that silica, PO4−3

Fig. 6. XPS wide scans of glasses FB6, FB7 and FB8 samples.

2230 K. Sharma et al. / Materials Science and Engineering C 29 (2009) 2226–2233

and Ca leach out from the surface. After 1 week immersion, theconcentrations of Si, Ca and P have increased slightly when comparedto as prepared SBF concentration. Amount of Si leaching out from thesurface increased marginally, as the Fe2O3 concentration increasedfrom 10 to 20 wt.%. The pH of the SBF solutions was also monitoredduring the experiment. Fig. 5 shows the change inpHduring the courseof experiment. The pH was found to increase from 7.4 to 8.26. Thischange in basicity is sharp from 7.4 to 8.1 (up to 1 week) and thengradual from 8 to 8.3 for 1–4 weeks.

3.2.2. XPS analysisX-ray photoelectron scans were recorded in the binding energy

(BE) region 0 to 1000 eV for as prepared glass samples. XPS spectra ofglass samples are shown in Fig. 6. The peak at the binding energyaround 284.6 eV is for the C1s core level, which has been used asreference energy. The doublet around 710 eV arises from Fe2pphotoelectrons. The peak at around 346 eV assigned to Ca2p, decreases

Fig. 7. XPS wide scans of glass-ceramics FBC6, FBC7, and FBC8 samples: (A) before immersi(D) after immersion for 4 weeks.

in intensity as the Fe2O3 content increases. The peak at around 530 eVis assigned to O1s.

XPS analysiswas used to study the surfacemodifications of samplesin SBF. Fig. 7A depicts the XPS spectra of glass-ceramics samples beforeimmersion in SBF and Fig. 7(B–D) shows the wide scan XPS spectra ofglass-ceramics samples immersed in the SBF for 1, 2 and 4 weeks,respectively. We observe very low intensity peak correspondingto iron in all the samples; while the intensities of peaks corre-sponding to Si and P change with the variation in the immersion timeof samples.

XPS scans of as prepared FBC6, FBC7and FBC8 glass-ceramic arecompared with XPS spectra of these samples immersed in SBF for1 week (Fig. 8(A–C)). The sample surface got depleted of Si and Caduring its interaction with SBF, which is evident from these spectra asthe intensities of peaks corresponding to Si and Ca decrease. The XPSspectra also show the appearance of the N1s core level signal which isabsorbed from the solution.

on in SBF; (B) after immersion in SBF for 1 week, (C) after immersion for 2 weeks and

Fig. 8. (A) XPS scans of as prepared glass-ceramic FBC6 before immersion and after immersion immersed in SBF for 1 week (FBC6a), (B) XPS scans of as prepared glass-ceramic FBC7before immersion and after immersion immersed in SBF for 1 week (FBC7a), (C) XPS scans of as prepared glass-ceramic FBC8 before immersion and after immersion immersed in SBFfor 1 week (FBC8a), (D) XPS spectra of FBC7 samples immersed in SBF for 2 weeks (FBC7b) with that of immersed in SBF for 4 weeks (FBC7c). (E) XPS wide scans of glass-ceramics(FBC6, FBC7, and FBC8) samples immersed in SBF for one month. (F) Surface composition (Si and P) for glass-ceramics samples after different immersion times in SBF.

2231K. Sharma et al. / Materials Science and Engineering C 29 (2009) 2226–2233

XPS scans of FBC7 samples immersed in SBF for 2weeks and 4weeksare compared in Fig. 8D. It clearly shows the depletion of Siwith increasein immersion time. There is also an increase in signal intensitycorresponding to P. XPS scans of FBC6, FBC7 and FBC8 samples, afterimmersion in SBF for 4 weeks are compared in Fig. 8E. We observed a

decrease in signal intensity for Ca2p and Ca2s in FBC8, compared to thatof FBC7 and FBC6. Fig. 8F shows the surface composition for glass-ceramics samples after different immersion times in SBF.With exposureto SBF solution (1 week to 2 week) percentages of Si remained very lowwhile percentages of P increased implying thegrowthof Ca–P rich phase.

2232 K. Sharma et al. / Materials Science and Engineering C 29 (2009) 2226–2233

3.2.3. SEM analysisFig. 9(A–B) illustrates the evolution of microstructure of the FBC 7

(15 wt.% iron oxide) sample upon immersion in SBF, for differentperiods of time. Surfacemorphology exhibits growth of the particles ofdifferent sizes ranging from 0.05 to 5 µm. The samples exhibited thegrowth of fine layer of particles on the surface, with some localizedcoagulation of the particles. These particles grow and increase thesurface coverage with increase in immersion time (from 2 weeks to4 weeks).

Surface morphologies for the FBC6 and FBC8 samples immersed inSBF for 4 weeks are shown in Fig. 10(A–B).These surfaces also exhibitthe fine spread of particles but the size of particles is smaller ascompared with FBC7 sample. Typical size is observed in the range of50–100 nm, but less area is covered by the additional layer in FBC6 andFBC8 samples. A number of clusters are visible at the surface. There aresome voids visible on the surface. These voids are possibly formed inthe initial stage of reaction of SBF fluid with the surface.

In the glass-ceramics studied, the amount of silica is kept constantand the role of increasing Fe2O3 on surface chemistry was elucidatedthrough in-vitro experiment. The pH of the SBF solutions initiallyincreased rather fast and then gradually reached to 8.2 (Fig. 5). The fastchange in sample FBC7, implies the prompt surface interaction withbody fluid. The change in pH indicates that the SBF solution ions attackthe surface and dissolve ions from the surface into the solution. This

Fig. 9. SEM micrograph of FBC7 glass-ceramics samples after immersion in SBF for;(A) 2 weeks, and (B) 4 weeks.

Fig. 10. SEM micrograph of glass-ceramics samples after immersion in SBF for 4 weeks;(A) FBC6, and (B) FBC8.

slowchange in the pH is due to the fact that these interactions betweenbody fluid and sample surface are time dependant and kinetic innature. This interaction is controlled by the diffusion [17–19]. This lossof Si from the surface to the solution results in Si–OH bond at thesurface. Thus interactions on the sample surface, resulted in marginalincrease in Si, Ca and P concentrations in the solutions (Table 5).The sharp decrease in P concentrations in the solution is observed,migration of Ca2+ and PO4

3− from the solution to the surface forming aCaO–P2O5 rich layer. This inference is supported by the fact that there isan increase in the basic nature of the solutions (Table 5) and is attri-buted to the increase in absorption of phosphate from the solution tothe surface.

XPS study of glass-ceramics showed that the interactions of SBFstartedwith the surfacemodifications of the samples upon exposure toSBF solution. As described a marked increase in P content and a netdecrease in Si on the sample surfacewas found for samples. This can beinterpreted as the formation of a calcium phosphate layer on thesample surface, caused by the adsorption of phosphate ions from theSBF solution. The formation of a CaO–P2O5-rich layer on the surface ofbioactive glasses upon interactionwith simulated body fluids has beenreported bymany authors [20–23]. This behaviourwas observed for allsamples; however the area of surface covered with Ca–P layerremained very low on the sample surface with 20 wt.% iron oxidecontent. The different interactions of glass-ceramics with SBF solution

2233K. Sharma et al. / Materials Science and Engineering C 29 (2009) 2226–2233

are evident from the wide scan XPS spectra of these samples. In orderto verify the hypothesis of simultaneous growth of a mixed calciumphosphate layer, scanning electron microscopic (SEM) investigationswere conducted after immersion of samples in SBF as mentionedearlier.

The Si content remained lowon the sample surface, as confirmedbythe low signal of the Si2p peak. As far as the FBC6 and FBC8 sampleswere concerned, very thin calcium phosphate layer was formed uponexposure to the SBF solution for 4 weeks (Fig. 10A and B). Theattenuation of the Si2p signal, observed in the wide scan spectra forFBC7, provides the evidence of complete coverage of the glass-ceramicafter 4 weeks of exposure (Fig. 9A and B).

The area of additional covering of layer depends on the surfacecomposition and structure of glass-ceramics. As the concentration ofFe2O3 has been increased at the expense of CaO in the matrix, it willtake a longer time for super-saturation to occur in the solution, whichresults in delay in migration of Ca2+ on the surface through the SiO2

rich layer. Further the rate of apatite formation in SBF decreases withincreasing Fe2O3 content in glass as it suppresses the dissolution ofcalcium, thus inhibiting the formation of the silica gel layer. Thereforeit takes a longer time to precipitate CaO–P2O5 rich film from thesupersaturated solution. Hence rate of apatite formation (bioactivity)is reduced as observed in our samples. These results suggest that thesurface morphology of the FBC7 glass-ceramic is favourable for thefaster growth of bioactive layer.

4. Conclusion

The magnetic bioactive glass-ceramics have been obtained fromglasses with compositions 34SiO2–(45−x) CaO–16P2O5–4.5 MgO–0.5CaF2−x Fe2O3 (where x=5,10,15, 20wt.%). The glasses prepared werefound to be paramagnetic in nature. Fe3+ and Fe2+ ions in these glassesare present in octahedral and tetrahedral coordination, respectively. TheFe2+ and Fe3+ ratio calculated by Mössbauer spectroscopy in glass, isfound to be a function of composition. In glass-ceramics, Fe3+ ions arefound in octahedral coordination and Fe2+ in tetrahedral coordination.The glass containing15wt.% Fe2O3, nearly 16%of the total ironenteredasFe3+ ions in tetrahedral coordination, i.e. entered in the glass network asa glass former. There was an increase in the saturation magnetizationwith increase in Fe2O3 content. The samples with iron concentrationabove 10 wt.% were ferrimagnetic. The biodegradation and the bio-activity of glass-ceramics are found to depend on the composition. The

interaction with SBF solution yielded the calcium phosphate layerformation on the surface of glass-ceramics samples. The surfaces ofthese materials were modified differently with increase in the amountof Fe2O3 in the glass matrix. According to the measured elementalcomposition and the surface study, the glass-ceramic having 15 wt.%Fe2O3 is biosorbable and bioactive, thus exhibiting the possibility ofbeing used as implant for hyperthermia application.

Acknowledgements

The authors wish to thank Drs. V.C. Sahni and J.V. Yakhmi forencouragement and support to this work.

References

[1] L.L. Hench, J. Am. Ceram. Soc. 81 (1998) 1705.[2] L.L. Hench, J. Am. Ceram. Soc. 74 (1991) 1487.[3] T. Kokubo, M. Shigematsu, Y. Nagashima, M. Tashiro, T. Nakamura, T. Yamamuro,

S. Higashi, Bull. Inst. Chem. Res., Kyoto Univ. 60 (1982) 260.[4] O. Bretcanu, E. Verné, M. Coisson, P. Tiberto, P. Allia, J. Magn. Magn. Mater.

300 (2006) 412.[5] Y. Ebisawa, T. Kokubo, K. Ohura, T. Yamamuro, J. Mater. Sci: Mater. Med. 4 (1993)

225.[6] M. Ikenaga, K. Ohura, T. Yamamuro, Y. Kotoura, M. Oka, T. Kokubo, J. Orthop. Res.

11 (1993) 849.[7] K. Takegami, T. Sano, H. Wakabayashi, J. Sodona, T. Yamazaki, S. Morita, T. Shibuya,

A. Uchida, J. Biomed. Mater. Res. 43 (1998) 210.[8] T. Leventouria, A.C. Kis, J.R. Thompson, I.M. Anderson, Biomaterials 26 (2005)

4924.[9] R.K. Singh, G.P. Kothiyal, A. Srinivasan, J. Mag. Magn. Mater. 320 (2008) 1352.[10] L.M. Bjursten, L. Emanuelsson, L.E. Ericson, P. Thomsen, J. Lausmaa, L. Mattson,

U. Rolander, B. Kaserno, Biomaterials 11 (1990) 596.[11] J. F Molder, W. F Stickle, P.E. Sobol, K.D. Bomen, J. Chastain, Handbook of X-ray

Photoelectron Spectroscopy, Perkin-Elmer Corporation, Physical Electronics Divi-sion, Minnesota, 1992, p. 25.

[12] X. Fang, C.S. Ray, A.M. Milanokovic, D.E. Day, J. Non-Cryst. Solids 283 (2001) 162.[13] D. Cullity, Introduction to Magnetic Materials, Addison-Wesley, Reading, 1972.[14] J.A. Johnson, C.E. Johnson, J. Phys., Condens. Matter 17 (2005) 381.[15] C. Ederer, N.A. Spaldin, Nat. Mater. 3 (2004) 849.[16] J. De Sitter, C. Dauwe, E. De Grave, A. Govaert, Solid State Commun. 18 (1976) 645.[17] L.L. Hench, D.E. Clark, J. Non-Cryst. Solids 28 (1978) 83.[18] R.W. Douglas, T.M. El-Shamy, J. Am. Ceram. Soc. 50 (1967) 1.[19] C.M. Jantzen, Mater. Res. Soc. 125 (1988) 143.[20] M.R. Filgueiras, G. La Torre, L.L. Hench, J. Biomed. Mater. Res. 27 (1993) 445.[21] S.B. Cho, K. Nakanishi, T. Kokubo, N. Soga, C. Ohtsuki, T. Nakamura, Biomed. Mater.

Res. 33 (1996) 145.[22] A. Balamurugan, G. Balossier, S. Kannan, J. Michel, A.H.S. Rebelo, J.M.F. Ferreira,

Acta Biomateriala 3 (2007) 255.[23] G. Lusvardi, D. Zaffe, L. Menabue, C. Bertoldi, G. Malavasi, U. Consolo, Acta

Biomater. 5 (2009) 419.