Synthesis, Characterization And Thermal Properties Of Polymer-magnetite

INSTITUTE OF PHYSICS PUBLISHING

NANOTECHNOLOGY

Nanotechnology 17 (2006) 2046–2053

doi:10.1088/0957-4484/17/8/043

Synthesis, characterization and thermal properties of polymer/magnetite nanocomposites Panagiotis Dallas1 , Vasilios Georgakilas1 , Dimitrios Niarchos1 , Philomela Komninou2, Thomas Kehagias2 and Dimitrios Petridis1 1 2

Institute of Materials Science, NCSR ‘Demokritos’, Agia Paraskevi 15310 Athens, Greece Department of Physics, Aristotle University of Thessaloniki, Thessaloniki, Greece

E-mail: [email protected]

Received 12 December 2005, in final form 26 February 2006 Published 24 March 2006 Online at stacks.iop.org/Nano/17/2046 Abstract A new procedure for the preparation of polymer/magnetic nanoparticle composites is described. Magnetite nanoparticles capped with methacrylate units were dispersed in a toluene solution of the monomer, and polymerization occurred after the addition of the initiator a
-a-azo-iso-butyronitrile at reflux temperature. The structural properties were determined by infrared spectroscopy, x-ray diffractometry and transmission electron microscopy. The thermal properties of the resulting nanocomposite were studied extensively with modulated differential scanning calorimetry. An increase in the glass transition temperature was observed after the incorporation of the nanoparticles.

1. Introduction Nanophase materials, the focus of intense research in the last decade, are associations of atoms or molecules having dimensions in the range of 1–100 nm with size dependent optical, electrical, magnetic and other properties that differ from those in the bulk state [1]. For example, the bands in semiconductor nanoparticles are strongly dependent on the particle size [2], while magnetic nanoparticles exhibit the phenomenon of superparamagnetism that is absent in the bulk material [3]. When such particles are homogeneously incorporated in polymeric matrices, nanocomposite materials result with properties often superior to the constituent members. For example, in clay–polymer nanocomposites the presence of exfoliated clay layers in the polymer enhances the mechanical properties upon small clay loadings [4]. A noteworthy property of nanoparticles is their ability to become dispersible in liquids by appropriate modification of their surface. For example, organic molecules can be chemisorbed by the surfaces of nanoparticles providing protection against agglomeration and air oxidation [5, 6] and simultaneously dispersability in organic solvents (organosols) [7] or aqueous media (hydrosols) [8]. This property makes nanophase materials very attractive since it helps their manipulation, 0957-4484/06/082046+08$30.00

such as in thin film formation or homogeneous dispersion into matrices [9]. As the presence of nanoclusters endows the composite materials with interesting properties, several metallic or metal oxide nanoparticles have been incorporated in various organic or inorganic matrices including polymers [10], resins [11], silica [12], clays [13] and porous materials [14]. In particular, magnetic iron oxide nanoparticles can be well dispersed and protected in a polymer mass which, in turn, acquires the characteristic magnetic properties of the oxides [15, 16]. Novel conductive polymer-magnetic nanocomposites have been designed in a similar manner by combining the magnetic and conducting properties [17]. The incorporation of nanoparticles in a polymer matrix alters, as expected, the mechanical, thermal and other characteristics of the polymer [18]. The usual method for the preparation of such polymer composites is the dispersion of nanoparticles into a melted polymer or dissolving them in a solvent polymer. The nanoparticles are usually capped with organic molecules in order to acquire the necessary solubility in the proper solvent. In this work, we describe the preparation and characterization of magnetic polymer composites with special emphasis on their thermal properties. The composites comprise of magnetite nanoparticles chemically bonded to the polymer chains. Poly (methyl methacrylate) and polystyrene were chosen as

© 2006 IOP Publishing Ltd Printed in the UK

2046

Synthesis, characterization and thermal properties of polymer/magnetite nanocomposites

polymer matrices, since they are well studied and their chains can be generated by polymerizing the respective monomer in the presence of magnetic nanoparticles, the surface of which had already been capped with methacrylate units.

2. Experimental details 2.1. Preparation of capped magnetite nanoparticles To a solution of 2.5 g of FeSO4 ·7H2 O in 30 ml H2 O were added 1.2 g KOH in 20 ml H2 O, followed by the dropwise addition of 0.5 ml of diluted H2 O2 (2%) until the green colour (from Fe(OH)2 ) of the solution turned black. Following this, 30 ml of toluene and 2.5 ml of oleic acid were added. The mixture was heated to reflux temperature for 1 h and was kept at room temperature until the two phases were clearly separated. After separation of the organic phase, the product was isolated by adding 30 ml acetone, washed several times with the same solvent and dried in air. 2.2. Exchange of the oleate units by methacrylate units 70 mg of oleic acid capped magnetite nanoparticles were dispersed in 30 ml toluene solution of methacrylic acid MA (2 ml), and the resulting mixture was refluxed for 3 h. The methacrylate capped magnetite was precipitated by ethanol, separated by centrifugion, washed with ethanol and air dried. 2.3. Preparation of polymethylmethacrylate–iron oxide composites An amount of iron oxide nanoparticles functionalized with methacrylic acid (0, 15, 60, 300 mg respectively) were mixed with 3 ml of methyl methacrylate monomer in 30 ml of toluene, and 30 mg of the initiator AIBN (a-a
-azo-iso-butyronitrile) was added to the resulting mixture. The mixture was heated to reflux for 3 h under argon flow. After completion of the reaction, the polymer/iron oxide composite was precipitated by adding 70 ml of hexane to the reaction mixture. The product was isolated by centrifugion, washed with hexane and dried at 60 ◦ C. The iron loading in the prepared samples was determined as 0, 2, 6 and 32% wt by thermogravimetric analysis, the first sample serving as a reference. Furthermore, in another control experiment, iron oxide nanoparticles capped with oleic acid were dispersed into pre-prepared PMMA. The polymer was first dissolved in toluene and, to the solution, the oleic acid capped iron oxide was added (6% by wt loading in pure magnetite) and the mixture was stirred. The composite was precipitated in hexane, washed with hexane and dried in 60 ◦ C. 2.4. Preparation of polystyrene–iron oxide composites An amount of methacrylate functionalized nanoparticles (0, 55, 120 mg respectively) was mixed with 5 ml of styrene monomer, and then 30 mg of the initiator AIBN in 30 ml of toluene was added. Three samples were prepared with iron loading determined with thermogravimetric analysis 0, 3 and 6% by wt. The mixture was heated to reflux for 20 h under argon flow. After the reaction was completed, the mixture was added to 70 ml of methanol whereupon the polymer/iron oxide

composite was precipitated, washed with methanol and dried at 60 ◦ C. 2.5. Chemicals Methacrylic acid and methylmethacrylate were purchased from Merck (99%), while oleic acid and FeSO4 ·7H2 O were purchased from Riedel Haans (99%) and Fluka (99%) respectively. The initiator AIBN was purchased from Merck (99%) and H2 O2 from Panreal.

3. Characterization of the products XRD patterns were recorded on powder samples using a ˚ Siemens 500 Diffractometer. Cu Kα radiation (λ = 1.5418 A) ◦ −1 Specimens for was used with a scan rate 0.03 s . transmission electron microscopy (TEM) were prepared as follows: a small amount of the material was immersed into an ultrasonic bath together with the appropriate solvent, in order to remove the polystyrene matrix and expose the crystal grains. Then the liquid containing the grains was injected on carbon-coated copper grids and dried under a lamp. TEM observations were carried out in a Jeol 100 CX electron microscope, operated at 100 kV. FT-IR spectra were collected on a Nicolet 20 SXC spectrometer. The specimen was made by mixing the corresponding composites with KBr (Aldrich, 99%, FT-IR grade). Thermal analysis was performed with a modulated differential scanning calorimetry instrument (M DSC Perkin-Elmer DSC-7 system) under a continuous N2 stream. The temperature modulation was 0.8 ◦ C min−1 , and the heating rate 5 ◦ C min−1 . The glass transition temperature (Tg ) was calculated from the reversible processes. The separation of the total heat flow into its reversing and non reversing components was based on the changes occurring in the measured heat capacity instead of ‘thermodynamic reversibility’. Thermogravimetric measurements (TGA) were performed on a Perkin-Elmer Pyris TGA/DTA while the magnetization curves were recorded on a vibrating sample magnetometer (VSM) with the external magnetic field varying from −20 to 20 kOe. M¨ossbauer spectroscopy was performed in a constant acceleration apparatus with a Co (Rh) source. Isomer shift values were relative to iron.

4. Results and discussion Magnetite nanoparticles capped with oleic acid were prepared by a one step method involving the partial oxidation of Fe(II) in alkaline solutions by dilute H2 O2 . The reaction was conducted in the presence of oleic acid, and under biphasic conditions [19]. The amount of oleic acid chemisorbed by the surface of the nanoparticles was estimated from the mass loss in the TGA analysis to be 24 wt% of the total weight. The surface bound oleate groups can be exchanged with methacrylate units by refluxing an excess of the monomer in toluene to yield the corresponding methacrylate capped magnetite. The exchange reaction ensures the chemical bonding of the methacrylate units to the surface iron atoms of the Fe3 O4 nanoparticles which, in turn, undergo the wanted polymerization with the vinyl groups of the methyl methacrylate or styrene monomers in toluene. This procedure 2047

P Dallas et al

Methacrylic acid Methyl methacrylate absorbed on the surface of iron oxide nanoparticle CH3 n

C

Poly(methyl methacrylate)

CH3 CH2 + x

C

COOMe

CH2

AIBN Toluene

C

C O

O

O-

COOMe

COOMe

O-

Figure 1. The incorporation of Fe3 O4 nanoparticles to the polymer matrix through chemical bonding.

2048

1,0

1517 1420

2843

0,6

2921

transmittance

0,8

0,4

0,2 (a) 0,0 3500

3000

2500 2000 1500 wavenumber (cm-1)

1000

500

1,0

0,8 transmittance

is expected to lead to grafting of the nanoparticles to the polymer chains, as presented schematically in figure 1. The capping agents chemisorbed by the surfaces of the magnetic iron oxide nanoparticles can be observed by the FT IR spectra presented in figure 2. The CH2 symmetric (2853 cm−1 ) and asymmetric (2916 cm−1 ) vibrations of the aliphatic alkyl chains are observed. The bands at 1517 and 1420 cm−1 arise from the antisymmetric and symmetric stretching vibration of the carboxylic group. The observed difference between these vibrations, 97 cm−1 for oleic acid indicates bidentate chelation of the COO− to the surface iron atoms. The C=C vibration was absent from the oleate capped nanoparticles, whereas it was clearly seen in the methacrylic acid spectrum. The vinyl stretching vibration of the methacrylic acid is observed at 1645 cm−1 (figure 2(b)). The difference between the two spectra and the presence of the methacrylic acid vibration modes after the exchange process indicated replacement of oleic by methacrylic acid. Figure 3 shows the XRD pattern of magnetite capped with oleic acid with the crystal planes corresponding to each diffraction peak [20]. The pattern is typical of an inverse spinel structure, meaning the formation of either magnetite (Fe3 O4 ) or maghemite (γ -Fe2 O3 ). The average size of the nanoparticles was calculated to be about 17 nm, according to the Scherrer equation [21]. D = 0.9λ/(2θ ) cos θ , where D is the crystalline domain size, (2θ ) is the full width at half maximum of the strongest peak and λ is the x-ray wavelength ˚ (λ = 1.5418 A). As the two oxides are isostructural, the XRD technique cannot clearly differentiate between them, especially in the nanophase state because the characteristic reflections are broad and appear almost at the same 2θ positions. M¨ossbauer spectroscopy can clearly identify the two oxides, owing to the presence of divalent iron in magnetite. The spectrum recorded at 5 K, bottom in figure 4(c), was fitted with two sextets (A and B) with the following parameters: site A IS (isomershift) = 0.32 mm s−1 , QS (quadropole splitting) = 0, Hh (hyperfinefield) = 489 Oe. Site B IS = 0.58 mm s−1 , QS = 0, Hh = 450 Oe. The high IS value for site B, which lies between typical trivalent and divalent iron values, indicates the presence of the Fe(II) state and accordingly the formation of Fe3 O4 . Similarly, the values of Hh are consistent with the magnetite phase [22]. Moreover,

0,6 1645

0,4

1555 1426

0,2 (b) 3500

3000

2500 2000 1500 wavenumber (cm-1)

1000

500

Figure 2. (a) IR spectrum of the oleic acid modified magnetite. (b) IR spectrum after the exchange process with methacrylic acid.

the M¨ossbauer spectra recorded at different temperatures, figures 4(a)–(c), are indicative of small particles exhibiting superparamagnetism, meaning that a fine particle is able to reverse its magnetic moment by thermal activation in the timescale of the M¨ossbauer effect. In superparamagnetism, all spins of iron atoms within a particle point in the same

Synthesis, characterization and thermal properties of polymer/magnetite nanocomposites

100

150

intensity (a.u)

311

98

100

96

220

94

440 222

50

511 400 422

92

(a)R.T.

0 20

30

40

50

60

70

80

90

o

2θ

Figure 3. XRD pattern of the Fe3 O4 nanoparticles capped with oleic acid.

intensity (a.u)

100.0 99.6 99.2 98.6 (b)L.N.

98.4 100.0 99.5 99.0 98.5 98.0 97.5 97.0

(c)L.He

96.5 -10

-5

0 5 velocity (mm/s)

10

100,0 99,8 intensity (a.u)

direction, but thermal fluctuations cause this direction to vary with a frequency that depends upon the particle size, anisotropy energy and temperature. If this frequency is greater than the Larmor procession frequency of the 57 Fe nucleus (10−8 s−1 ), the magnetic hyperfine splitting collapses to a singlet or doublet line, the latter if a quadrapole interaction is present. In the opposite case of slow relaxation of the iron spins, a complete magnetic hyperfine splitting is observed. However, because of the size distribution in a sample, the spectra consist of a doublet from small particles with short relaxation times, together with a sextet due to larger particles with longer relaxation times [23]. Furthermore, the relative area of the doublet increases with increasing temperature as a result of increasing the relaxation frequency, while the magnetic sextet increases with lowering the temperature as the relaxation frequency slows down. According to this description, the variable temperature M¨ossbauer spectra demonstrate that the present magnetite particles capped with oleic acid are nanoscale objects with a distribution of sizes. TEM microscopy from a sample, after dissolution of the poly(methyl)methacrylate matrix, revealed that the nanoparticles are cubic shaped with a mean diameter of 20 nm (figure 5(a)). Electron diffraction analysis showed that the iron oxide crystallized in the form of magnetite (Fe3 O4 ). A bright-field micrograph, depicting the crystal grains of the material is shown in figure 5(a), whereas the corresponding polycrystalline electron diffraction (ED) pattern is given as an inset. Measurements performed on the 1–7 rings of the ED pattern (figure 5(b)) resulted in the following interatomic distances: d1 = 4.83 nm, d2 = 2.95 nm, d3 = 2.52 nm, d4 = 2.08 nm, d5 = 1.70 nm, d6 = 1.61 nm and d7 = 1.48 nm. These values match, almost perfectly, the interplanar spacings of the 111, 220, 311, 400, 333 and 440 crystal planes of magnetite (Fe3 O4 ). These data establish that no change in the magnetic iron state occurred during polymerization and isolation. Moreover, as no other spots or rings are seen in the ED, all crystal grains are pure magnetite. TEM micrographs of the nanomagnetic poly(methyl)methacrylate composite with 6% wt iron oxide loading demonstrate that the nanoparticles are not uniformly

99,6 99,4 99,2 (d) 99,0 -10

-5

0

5

10

velocity (mm/sec)

Figure 4. M¨ossbauer spectra recorded at various temperatures for the oleic acid capped magnetite (a) 300 K, (b) 77 K, (c) 5 K, (d) PMMA composite with 6% wt magnetite recorded at 300 K.

distributed in the polymer, but exist as agglomerates of different sizes and shapes as shown in figure 6. The M¨ossbauer spectrum of this composite at room temperature is shown in figure 4(d). The spectrum compared with that of the pure magnetite partner, figure 4(a), clearly shows enhanced magnetic sextets reflecting the agglomeration of magnetite particulates during the polymerization process. Hysteresis loops at room temperature for two Fe3 O4 – poly(methyl)methacrylate nanocomposites with 6 and 32% wt iron oxide both show field dependent magnetization curves typical of ultrafine particle magnetic composites. The 2049

P Dallas et al Fe3O4 loading 32% wt

10

0 Magnetization (emu/gr)

Magnetization (emu/gr)

Fe3O4 loading 6% wt

-10

0

-20000

-20000

0 20000 Magnetic Field (Oe)

-10000 0 10000 Magnetic Field (Oe)

20000

Figure 7. Magnetization curves recorded at room temperature for PMMA nanocomposites with 6% and 32% wt iron oxide loading respectively. Inset: magnification of magnetization curves near zero magnetic field. Table 1. Glass transition temperature (Tg ) and enthalpy (H ) values of the polymer composites with various iron oxide content. Figure 5. (a) TEM micrographs of oleic acid capped Fe3 O4 nanoparticles. The electron diffraction pattern is given as an inset. (b) The electron diffraction pattern with the corresponding rings indicated.

Iron oxide content (%) in PMMA composites

Tg ( ◦ C)

H (J g−1 ◦ C−1 )

0 2 6 32 6 (simple mixing)

99.59 112.85 114.13 118.98 109.51

0.3488 0.2420 0.2304 0.0866 0.0265

Table 2. Glass transition temperature (Tg ) and enthalpy (H ) values of the polymer composites with various iron oxide content.

Figure 6. TEM micrographs of the polymer nanocomposite with 6% wt magnetite, showing the formation of nanoparticles agglomerates in the polymer matrix.

hysteresis loops of two samples with 6 and 32% wt loadings in magnetite at RT are presented in figure 7. The saturation magnetization values at RT are 13 emu g−1 for the sample with 32% wt iron oxide and 3 emu g−1 for the sample with 6% wt iron oxide. Very low coercivity values were observed in the two samples (240 and 220 Oe for the samples with 32% wt and 6% wt, respectively) and probably arise from the large agglomerates, the magnetic moment of which is blocked at room temperature, as also suggested by the RT M¨ossbauer spectra of the 6% wt sample (figure 4(d)). Similar low coercivities for nanoscale magnetite have been reported [24, 25]. The values expressed in pure magnetite, 44 emu g−1 (sample 32% wt) and 50 emu g−1 (sample with 6% wt) are far from the value of bulk Fe3 O4 , 92 emu g−1 , but 2050

Iron oxide content(%) in PS composites

Tg (◦ C)

H (J g−1 ◦ C−1 )

0 3 6

72.12 72.44 74.30

0.2310 0.2600 0.2535

they are in good agreement with values reported for Fe3 O4 particles of similar sizes [24, 25]. Figure 8 presents the DSC thermographs for the polymethylmethacrylate nanocomposites. The glass transition temperature (Tg ) of pure PMMA is estimated at 99 ◦ C, while for the polymer/magnetite composites it increased to 110 ◦ C. The composite formed by simple mixing of nanoparticles with the methyl methacrylate polymer had a Tg value near 109 ◦ C (see table 1). The Tg ( ◦ C) and H (J g−1 ◦ C−1 ) values of polystyrene and polystyrene nanocomposites are reported in table 2, and the diagrams are presented in figure 9. All samples were first dried at 60 ◦ C under vacuum, in order to evaporate the solvent. Considering the thermal properties of composites, we must point out the important property of the glass transition This is defined as the temperature (Tg ) of polymers. temperature where the plastic becomes hard and brittle when cooled rapidly after heating. At the glass transition temperature, the weak secondary bonds that stick the polymer chains together are broken, and the macromolecule starts to

Synthesis, characterization and thermal properties of polymer/magnetite nanocomposites

0,0

0 % wt

2 % wt

heat flow reversible heat flow

0,0

-0,5

Heat flow (W/g)

Heat flow (W/g)

non reversible

-1,0 -1,5

heat flow reversible heat flow non reversible heat flow

-0,5

-1,0

-2,0 (a)

-1,5

50

0,00

75 100 125 Temperature (°C)

0,0

6 % wt

75 100 125 Temperature (°C)

32 % wt

150

heat flow (W/g) reversible heat flow non reversible heat flow

-0,5

-0,25

Heat flow (W/g)

(b)

50

150

heat flow reversible heat flow non reversible heat flow

-0,50

Heat flow (W/g)

-2,5

-0,75 -1,00

-1,5 -2,0

(d)

-1,25 (c) 50

-1,0

75 100 125 Temperature (°C)

0,0

75

150

6 % simple mixing

100 125 Temperature (°C)

150

heat flow reversible heat flow non reversible heat flow

Heat flow (W/g)

-0,5 -1,0 -1,5 -2,0 (e)

50

75

100 125 Temperature (°C)

150

Figure 8. DSC curves of the nanocomposites with PMMA and with the following iron oxide loading per cent: (a) 0%, (b) 2%, (c) 6%, (d) 32%, (e) simple mixture, 6%.

move. Tg can be measured from the curve of the reversible processes on the modulated DSC analysis. The analysis showed an increase in Tg as the concentration of nanoparticles increased. This behaviour probably arises from branching formed when islands of nanoparticles are bonded to different polymeric chains. This lowers the mobility of the chains, and as a result the glass transition temperature increases in the nanocomposites. In addition, increasing the concentration of nanoparticles makes the nanocomposite more brittle, leaving even less ‘free’ space for the polymer macromolecules to move. This also includes an increase in the glass transition temperature [26]. Tg of the sample prepared by simple mixing also increased to 109.5 ◦ C in comparison with 99.5 ◦ C of the pure polymer. This indicates that simple mixing of the nanoparticles in the polymer matrix results in an increase of Tg . On the other hand, this value is lower than that, 114.1 ◦ C, observed for nanocomposite B having the same loading of 6% in iron oxide. Accordingly, it can be concluded that the

increase in Tg depends not only on the presence of the magnetic nanoparticles, but also on the bonding of the nanoparticles to the polymeric chains. Figure 10, illustrates the two different ways in which the nanocomposites can be dispersed in the polymer matrix. A similar increase in Tg for the polystyrene/iron oxide composites was not observed. This probably arises from the more ‘plastic’ character of the polystyrene since the Tg of the pristine polymer is lower.

5. Conclusions In this paper, poly(methyl methacrylate) or polystyrene/magnetite nanocomposites were prepared and fully characterized. A constant increase in the Tg of the nanocomposites was observed. This increase was higher when the nanoparticles were chemically bonded to the polymeric chains. This 2051

P Dallas et al 0 % wt

0,0

heat flow reversible heat flow non reversible heat flow

Heatflow (W/g)

-0,5 -1,0 -1,5 -2,0

(a)

25

50

0,0

75 100 Temperature (°C)

125

3 % wt

6 % wt

0,0

Heat flow (W/g)

-0,5

Heat flow (W/g)

heat flow reversible heat flow non reversible heat flow

-1,0 -1,5

heat flow reversible heat flow non reversible heat flow

-0,5

-1,0

-1,5 -2,0 (b)

(c)

25

50

75 100 Temperature (°C)

25

125

50

75 100 Temperature (°C)

125

Figure 9. DSC thermographs of the polystyrene nanocomposites with the following iron oxide loading per cent: (a) 0% (simple polystyrene) (b) 3%, (c) 6%.

MeOOC C MeOOC

MeOOC

C MeOOC

C

COOMe COOMe

O-O COOMe

O O

O O

C COOMe

OO

COOMe

COOMe

COOH

MeOOC C

C COOMe

MeOOC

COOMe

MeOOC

MeOOC

COOMe C CH H2 COOMe

COOMe COOMe

MeOOC

MeOOC

COOMe COOMe COOMe H H2 C C MeOOC

COOMe

COOMe

COOMe

Figure 10. The two different ways for the incorporation of the nanoparticles in the polymer matrix, either by chemical bonding or simple mixing.

incorporation and the chemical bonding may positively influence the mechanical properties of the magnetic polymer nanocomposites.

References [1] Weller H 1993 Angew. Chem. Int. Edn 32 41 Schmid G (ed) 1994 Clusters and Colloids (New York: VCH) Klabunde K J (ed) 2001 Nanoscale Materials in Chemistry (New York: Wiley–Interscience) Fendler J H (ed) 1998 Nanoparticles and Nanostructured Films (Weinheim: Wiley–VCH)

2052

[2] [3] [4] [5]

Alivisatos A P 1996 Science 271 933 Kodama R H 1999 J. Magn. Magn. Mater. 200 359 Lan T and Pinnavaia T J 1994 Chem. Mater. 6 2217 Bourlinos A B, Simopoulos A and Petridis D 2002 Chem. Mater. 14 899 Hyeon T 2003 Chem. Commun. 927 [6] Ghosh M, Lawes G, Gayen A, Subbanna G N, Reiff W M, Subramanian M A, Ramirez A P, Zhang J P and Seshadri R 2004 Chem. Mater. 16 118 [7] Sun S and Zeng H 2002 J. Am. Chem. Soc. 123 8204 Rockenberger J, Scher E C and Alivisatos A P 1999 J. Am. Chem. Soc. 121 11595

Synthesis, characterization and thermal properties of polymer/magnetite nanocomposites

[8]

[9] [10]

[11]

[12] [13] [14]

[15]

Bourlinos A, Simopoulos A and Petridis D 2002 Chem. Mater. 14 899 Li Z, Chen H, Bao H and Gao M 2004 Chem. Mater. 16 1391 Fu L, Dravid V P and Johnson D L 2001 Appl. Surf. Sci. 181 173 Kang Y S, Risbud S, Rabolt J F and Stroeve P 1996 Chem. Mater. 8 2209 Yokoi H and Kantoh T 1993 Bull. Chem. Soc. Japan 66 1536 Tian Y C and Fendler J H 1996 Chem. Mater. 8 969 Fried T, Shemer G and Markovich G 2001 Adv. Mater. 13 1158 Chatterjee J, Haik Y and Chen C J 2001 J. Magn. Magn. Mater. 225 21 Novakova A, Lanchinskaya V Y, Volkov A V, Gendler T S, Kiseleva T, Moskvina M A and Zezin S B 2003 J. Magn. Magn. Mater. 35 258 Gich M et al 2003 Appl. Phys. Lett. 82 4307 Ziolo R F, Giannelis E P, Weinstein B A, O’Horo M P, Gianguly B N, Mehrotra V, Russell M W and Huffman D R 1992 Science 257 219 Abe T, Tachibana Y, Uematsu T and Iwamoto M 1995 Chem. Commun. 1617 Froba M, Kohn R and Bouffaud G 1999 Chem. Mater. 11 2858 Bourlinos A, Karakassides M, Simopoulos A and Petridis D 2000 Chem. Mater. 12 2640 Bourlinos A B, Simopoulos A, Boukos N and Petridis D 2001 J. Phys. Chem. B 105 7434 Shi J, Hua Z and Zhang L 2004 J. Mater. Chem. 14 795 Bourlinos A B, Radek Z and Petridis D 2003 Micropor. Mesopor. Mater. 58 155 Novakova A A, Lanchinskaya V Y, Volkov A V, Gendler T S, Kiseleva T Y, Moskvina M A and Zezin S B 2003 J. Magn. Magn. Mater. 258 354

[16]

[17]

[18] [19] [20] [21]

[22] [23]

[24] [25] [26]

Hitoshi I, Hajime F, Motoo A and Masayuki K 1993 J. Magn. Magn. Mater. 118 263 Ninjbadgar T, Yamamoto S and Fukuda T 2004 Solid State Sci. 6 879 Hitoshi I, Hajime F and Masayuki K 1992 J. Colloid Interface Sci. 148 534 Rupali G and Amitabha D 2002 Chem. Mater. 12 608 Tada M, Hatanaka S, Sanbonsugi H, Matsushita N and Abe M 2003 J. Appl. Phys. 93 756 Lopez D, Cendoya I, Tores F, Tejada J and Mijangos C 2001 Polym. Eng. Sci. 41 1845 Bourlinos A B, Bakandritsos A, Georgakilas V, Tzitzios V and Petridis D 2005 J. Mater. Sci. at press Hyeon T, Lee S S, Park J, Chung Y and Na H B 2001 J. Am. Chem. Soc. 123 12798 Klug H P and Alexander L E 1962 X Ray Diffraction Procedures for Polycrystalline and Amorphous Materials (New York: Wiley) pp 491–538 Ji T, Fang J, Golob V, Tang J and O’Connor Ch 2002 J. Appl. Phys. 92 6833 Gee S H, Hong Y K, Erickson D W, Park M H and Sur J C 2003 J. Appl. Phys. 93 7560 Morup S, Damestic J A and Topsol H 1980 Applications of Moessbauer Spectroscopy vol II, ed R L Cohen (New York: Academic) Butterworth M D, Bell S A, Armes S P and Simpson A W 1996 J. Colloid Interface Sci. 183 91 Chang H S W, Chiou C C and Chen Y W 1997 J. Solid State Chem. 128 87 Golding B 1959 Polymers and Resins (New York: Van Nostrand)

2053