High-coercivity FePt nanoparticle assemblies embedded in silica thin films

This article has been downloaded from IOPscience. Please scroll down to see the full text article.

2009 Nanotechnology 20 025609

(http://iopscience.iop.org/0957-4484/20/2/025609)

Download details:

IP Address: 155.69.4.4

The article was downloaded on 08/06/2010 at 09:06

Please note that terms and conditions apply.

View the table of contents for this issue, or go to the journal homepage for more

Home Search Collections Journals About Contact us My IOPscience

IOP PUBLISHING NANOTECHNOLOGY

Nanotechnology 20 (2009) 025609 (5pp) doi:10.1088/0957-4484/20/2/025609

High-coercivity FePt nanoparticleassemblies embedded in silica thin filmsQ Yan1,2,3, A Purkayastha1, A P Singh1, H Li1, A Li2,R V Ramanujan3 and G Ramanath1

1 Department of Materials Science and Engineering, Rensselaer Polytechnic Institute, Troy,NY 12180, USA2 Department of Materials Science and Engineering, Nanjing University, Nanjing 210000,People’s Republic of China3 School of Materials Science and Engineering, Nanyang Technological University, 639798,Singapore

E-mail: ramanath@rpi.edu and alexyan@ntu.edu.sg

Received 2 September 2008, in final form 27 October 2008Published 11 December 2008Online at stacks.iop.org/Nano/20/025609

AbstractThe ability to process assemblies using thin film techniques in a scalable fashion would be a keyto transmuting the assemblies into manufacturable devices. Here, we embed FePt nanoparticleassemblies into a silica thin film by sol–gel processing. Annealing the thin film composite at650 ◦C transforms the chemically disordered fcc FePt phase into the fct phase, yieldingmagnetic coercivity values Hc > 630 mT. The positional order of the particles is retained due tothe protection offered by the silica host. Such films with assemblies of high-coercivity magneticparticles are attractive for realizing new types of ultra-high-density data storage devices andmagneto-composites.

1. Introduction

FePt and CoPt nanoparticles with the chemically orderedfct structure have been considered as building blocks forfabricating next-generation information storage media [1, 2].The high magnetocrystalline anisotropy [3–5] of fct FePtand its high chemical stability helps decrease the size limitfor the onset of superparamagnetism of nanoscopic magneticbits. However, several challenges remain to be overcomein order to exploit nanostructures of these materials forpractical applications. As-synthesized FePt nanoparticlestypically have the chemically disordered face centered cubic(fcc) structure, which leads to low magnetic coercivity [6].Obtaining control over the particle size [7] and dispersity [1],and forming ordered arrays of nanoparticles with tunablecoercivity, are a combination of attributes that are yet to berealized. Furthermore, the high temperature anneals between400 and 650 ◦C necessary to transform the disordered fcc FePtphase to the fct phase not only disrupt the positional orderof the nanoparticles in assemblies [8], but also cause particlecoalescence [7, 9].

Several strategies have been explored to obviate particlecoalescence upon annealing. Examples include decreasingthe fct ordering temperature by impurity doping [10–12],

depositing SiO2 film by PVD onto FePt nanoparticle films [13],polymer templating [14, 15] and synthesis of core–shellnanoparticles [16–18]. However, success has been limited inpreserving the overall positional order of the self-assembledFePt particle assemblies upon annealing to achieve chemicalordering and high coercivity. For example, althoughthere are several ways to protect the FePt nanoparticlesfrom coalescence during annealing by forming FePt–silicacore–shell nanoparticles [17] and NaCl mixing [19], theseferromagnetic particles cannot easily form an assembly dueto strong interparticle magnetic dipole interaction as observedin our experiments. This indicates that simply protecting theFePt particles instead of their assemblies during the annealingprocess still cannot achieve the recording media applicationpurpose. In this case, embedding magnetic nanocrystals intoinorganic matrices [20] is an attractive approach to make thenanoparticle assembly both thermally and mechanically stableapart from being amenable to thin film processing via sol–geland spin-coating routes.

Here, we report the formation of assemblies of FePtnanoparticles embedded in silica thin films with controllableFe:Pt ratio and nanoparticle loading. The FePt–silica thinfilms yield high coercivity Hc ∼ 500 mT upon annealing at650 ◦C due to partial fct phase formation without disrupting

0957-4484/09/025609+05$30.00 © 2009 IOP Publishing Ltd Printed in the UK1

Nanotechnology 20 (2009) 025609 Q Yan et al

Figure 1. (a) Bright-field TEM micrograph and (b) a diffraction pattern from an as-prepared assembly of FePt nanoparticles capped with oleicacid and oleyl amine, dispersed on a carbon-coated TEM grid. (c) A representative TEM micrograph from an assembly of FePt nanoparticlescapped with a second surfactant layer of CTAB. (d) The size distribution of the particles shows an average particle size of 5.4 nm. The solidline is Gaussian curves connecting the data.

the positional order of the particles in the assembly. Suchthermally stable FePt–silica thin film composites comprisedof assembled high-coercivity magnetic nanoparticles could befurther developed and improved for novel memory devices andmagneto-composites.

The FePt nanoparticles were synthesized by decomposingFe(CO)5 and Pt(acac)2 by refluxing in phenyl ether in aflowing Ar ambient [1] as reported recently. The as-preparedFePt nanoparticles, passivated with oleic acid and oleyl amine,were rendered water soluble by coating the particles witha second surfactant layer of tetraoctylammonium bromide(CTAB) [20, 21]. Silica films were formed from tetraethylorthosilicate (TEOS) by hydrolysis and condensation. Wecarried out hydrolysis in an acidic environment of pH ∼ 5with a gelation speed similar to that observed at pH ∼2 [22]. We offset the Fe depletion during TEOS exposureby including an excess Fe precursor in our synthesis, andlimiting the exposure time of FePt nanoparticles to hydrolyzedTEOS.

2. Experimental details

The FePt nanoparticles were incorporated into the hydrolyzedTEOS solution before spin coating at a speed of 1000 rpm toform the FePt–silica composite thin films.

2.1. Preparation of FePt in water solution

The FePt nanoparticles, capped by oleic acid and oleylamine in chloroform solution, were synthesized using reportedprotocols, e.g. solvothermal [1] method or microemulsionmethod [17]. 1-octodecanethiol (50 mg, Aldrich) was added tothe FePt nanoparticles (50 mg) in chloroform solution (1 ml,Aldrich) and sonicated for 10 min. Subsequently, the FePtsolution was added into deionized water (3 ml) containingCTAB (45 mg, Aldrich) under vigorous stirring and heated at60 ◦C to remove the chloroform solution, which results in thetransfer of the hydrophobic FePt nanoparticles into deionizedwater by CTAB encapsulation [20, 26].

2.2. Preparation of silica precursor

TEOS (1 g, Aldrich) was added to deionized water (6 ml) thathas a pH = 5 adjusted by adding a small amount of HCl. Theaqueous solution was heated at 50 ◦C for 4–5 h to hydrolyzethe TEOS.

2.3. Preparation of FePt–silica superlattice film

The as-prepared aqueous FePt solution was mixed with theabove aqueous TEOS solution. The FePt–silica superlatticefilms were formed by spin casting at 1500 rpm or drop casting

2

Nanotechnology 20 (2009) 025609 Q Yan et al

100 nm 50 nm

100 nm 50 nm

Figure 2. Bright-field TEM images from as-prepared FePt–silica films. (a) Low- and (b) high-magnification images of samples preparedusing XFT = 1:15. (c) and (d) show the corresponding images for samples prepared with XFT = 1:5. The insets in (a) and (c) show Fouriertransform patterns indicating random, and short-range positional order of the FePt particles, respectively.

onto silicon substrate. The volume ratio between aqueous FePtsolution and TEOS solution can be varied to change the FePtloading in the silica films.

2.4. Characterization

The as-prepared FePt–silica films were annealed in a 4 ×10−6 Torr vacuum at 650 ◦C for 1 h. The as-preparedand annealed FePt–silica films were scraped from the siliconsubstrate with a blade, dispersed into ethanol solution anddrop-coated onto carbon-coated 200 mesh copper grids. TheFePt–silica film microstructure was examined by using aPhilips CM 12 transmission electron microscope (TEM). ASCINTAG/PAD-V x-ray diffractometer with Cu Kα radiationwas used to characterize the assembly order of FePt–silica films and the FePt phase transformation after heating.In the latter case, drop-coated thick FePt–silica sampleswere used. The magnetic properties were characterized atroom temperature, in a Lake Shore 7400 vibrating samplemagnetometer (VSM) instrument using applied magnetic fieldsup to 2 T.

3. Result and discussion

The as-prepared ∼5 nm diameter FePt nanoparticles (seefigure 1) have a narrow size distribution with a standarddeviation of less than ±10%. The interparticle spacing is ∼4–5 nm, maintained by the capping agents oleic acid and oleylamine (see figure 1(a)), consistent with previous reports [1, 2].Selected area electron diffraction patterns from the as-preparedFePt nanoparticles (see figure 1(b)) reveal a disordered fccstructure. Coating the FePt nanoparticles with CTAB—thesecond surfactant layer—renders the particles water soluble,but does not produce any detectable change in the interparticledistance (see figure 1(c)). The domain size of the nanoparticleassembly is small, e.g. 30–100 nm, probably due to fast solventremoval (e.g., <60 min), and could be improved through theuse of Langmuir–Blodgett methods [23, 24] and decreasing thesolvent evaporation rate [16].

We mixed the hydrophilic nanoparticles with an aqueoussolution of hydrolyzed TEOS in order to embed FePtnanoparticles into a silica thin film (see figure 2). The FePtnanoparticles are uniformly dispersed in the silica matrix, anddo not form clusters. The microstructure of the FePt–silica

3

Nanotechnology 20 (2009) 025609 Q Yan et al

(d) (e)

100 nm 100 nm

Figure 3. Bright-field TEM images of FePt–silica films after annealing at 650 ◦C for 1 h for (a) XFT = 1:15 and (b) XFT = 1:5. (c) Arepresentative electron diffraction pattern from FePt–silica films with XFT = 1:5 showing peaks corresponding to the fct structure. Low-anglex-ray diffractograms from FePt–silica composite films before and after annealing for samples with (d) XFT = 1:5 and (e) XFT = 1:15.

films is sensitive to the FePt:TEOS ratio XFT. Films with alow loading of FePt nanoparticles, e.g. XFT = 1:15 show a lowdegree of assembly order (see figures 2(a) and (b)), which ismanifest as a diffuse ring in the Fourier transform of the image.Increasing XFT to 1:5 leads to formation of 3D close-packedparticle assembly in SiO2 film (e.g., see figures 2(c) and (d)).Fourier transform patterns from films with XFT = 1:5 exhibitspot patterns with six-fold symmetry. The pattern contrast isweak, probably due to the small assembly domain size, whichcan presumably be improved by adjusting the solidificationtime.

Annealing the FePt–silica films at 650 ◦C in vacuumtransforms the nanoparticles to the chemically partially orderedfct phase that yields high coercivity, without obvious particlecoalescence (see figure 3). We attribute the retention ofthe particle shape and size at such high temperatures (atwhich uncapped FePt particles are known to coalesce) [27] tocurtailed atomic transport resulting from particle encapsulationby the silica host matrix. The FePt nanoparticles transformfrom a disordered fcc structure to the partially ordered fctphase, as testified by the emergence of the (001) and (110)electron diffraction spots (see figure 3(c)).

Low-angle x-ray diffractograms from FePt–silica filmswith a high FePt loading of XFT = 1:5 show two sets of Braggreflections: weak broad peaks at 2θ = 1.30◦ and 2.15◦ andtwo strong sharp peaks at 2θ = 3.41◦–6.82◦ (see figure 3(d)).

The narrow diffraction peaks are attributed to the formation ofordered pores in the silica films [20] due to excess surfactant.Assuming a two-dimensional hexagonal pore structure [25],the sharp reflections can be indexed as (100) and (200),respectively, with d100 = 2.61 nm. The broad peaks arisefrom the short-range positional order of the FePt nanoparticlesin the silica matrix, and can be indexed as (111) and (220),respectively. Assuming a close-packed fcc structure [20, 26],the lattice parameter of the nanoparticle assembly is 11.7 nm.Silica films with a low FePt nanoparticle loading, e.g. XFT =1:15, show only the sharp peaks (see figure 3(e)). The diffusepeaks are absent, consistent with the lack of short-range order.Annealing at 650 ◦C disrupts the mesoporous silica phase inboth types of sample, as indicated by the disappearance of thestrong narrow peaks. However, the retention of the broad peaksat 2θ = 1◦–2.5◦ (see figure 3(d)) indicates that short-rangepositional order of the FePt nanoparticles in the silica matrixcan be preserved after annealing, which is consistent with ourTEM results.

The XRD diffractogram of as-prepared FePt-SiO2 filmsshows the chemically disordered fcc structure of FePt, asshown in figure 4(a). The FePt nanoparticles transformedinto partially ordered fct phase after annealing at 650 ◦C, asindicated by the emergence of (001), (110) and (201) Braggpeaks in the high-angle x-ray diffractograms. The maximumordering parameter is S = 0.78, as estimated from the integral

4

Nanotechnology 20 (2009) 025609 Q Yan et al

Figure 4. (a) X-ray diffractograms from FePt–silica composite samples before and after annealing, depicting the annealing-inducedfcc → partial fct phase transformation. The estimated grain size for FePt particles annealed at 650 ◦C is calculated from the (111) peak widthto be 4 nm. (b) Room-temperature hysteresis loops obtained from FePt–silica films drop-coated on a Si substrate.

intensity ratio between (110) and (111) reflections [28]. Priorto annealing the thin film composite is superparamagnetic, withHc ∼ 0 mT, as expected. The fcc → fct transformation in thenanoparticles increases Hc to 630 mT (see figure 4(b)), whichis also consistent with the partial fct ordering indicated by theXRD results.

4. Conclusion

In summary, we demonstrate the incorporation of a FePtnanoparticle assembly into silica to form FePt–silica thin films.The short-range positional order of the FePt nanoparticlescan be preserved upon annealing at 650 ◦C while theroom-temperature Hc increases to ∼630 mT. Achievingsuch high-coercivity FePt–silica thin films provides a newpotential approach to be further developed for realizing FePtnanoparticle patterned media for data storage applications.

Acknowledgments

We gratefully acknowledge funding from NSF through DMR0519081, NY State through the Focus Center, a HondaResearch Initiation Grant and NSFC 10704035 Grant.

References

[1] Sun S H, Murray C B, Weller D, Folks L and Moser A 1989Science 2000 287

[2] Sun S H 2006 Adv. Mater. 18 393–403[3] Coffey K R, Parker M A and Howard J K 1995 IEEE Trans.

Magn. 31 2737[4] Christodoulides J A, Bonder M J, Huang Y, Zhang Y,

Stoyanov S, Hadjipanayis G C, Simopoulos A andWeller D 2003 Phys. Rev. B 68 054428

[5] Willoughby S D, MacLaren J M, Ohkubo T, Jeong S,McHenry M, Laughlin D E, Choi S J and Kwon S J 2002J. Appl. Phys. 91 8822–4

[6] Barmak K, Kim J, Shell S, Svedberg E B and Howard J K 2002Appl. Phys. Lett. 80 4268–70

[7] Ding Y and Majetich S A 2005 Appl. Phys. Lett. 87 022508[8] Dai Z R, Sun S H and Wang Z L 2002 Nano Lett. 505 325–35[9] Jia Z Y, Kang S, Shi S, Nikles D E and Harrell J W 2005

J. Appl. Phys. 97 10J310[10] Kang S, Harrell J W and Nikles D E 2002 Nano Lett. 2 1033[11] Kang S S, Jia Z Y, Nikles D E and Harrell J W 2003 IEEE

Trans. Magn. 39 2753–7[12] Yan Q Y, Kim T, Purkayastha A, Ganesan P G, Shima M and

Ramanath G 2005 Adv. Mater. 17 2233–7[13] Ding Y, Majetich S A, Kim J, Barmak K, Rollins H and

Sides P 2004 J. Magn. Magn. Mater. 284 336–41[14] Mizuno M, Sasaki Y, Yu A C C and Inoue M 2004 Langmuir

20 11305–7[15] Sun S H, Anders S, Thomson T, Baglin J E E, Toney M F,

Hamann H F, Murray C B and Terris B D 2003 J. Phys.Chem. B 107 5419–25

[16] Liu C, Wu X W, Klemmer T, Shukla N and Weller D 2005Chem. Mater. 17 620

[17] Yan Q, Purkayastha A, Kim T, Kroger R K, Bose A andRamanath G 2006 Adv. Mater. 18 2569

[18] Lee D C, Mikulec F V, Pelaez J M, Koo B and Korgel B A2006 J. Phys. Chem. B 110 11160–6

[19] Li D R, Poudyal N, Nandwana V, Jin Z Q, Elkins K andLiu J P 2006 J. Appl. Phys. 99 08E911

[20] Fan H Y, Yang K, Boye D M, Sigmon T, Malloy K J, Xu H F,Lopez G P and Brinker C J 2004 Science 304 567–71

[21] Fan H Y, Leve E, Gabaldon J, Wright A, Haddad R E andBrinker C J 2005 Adv. Mater. 17 2587

[22] Brinker J and Scherer G W 1990 Sol–Gel Science: The Physicsand Chemistry of Sol–Gel Processing (Boston, MA:Academic)

[23] Iakovenko S A, Trifonov A S, Giersig M, Mamedov A,Nagesha D K, Hanin V V, Soldatov E C and Kotov N A1999 Adv. Mater. 11 388–92

[24] Sachan M, Walrath N D, Majetich S A, Krycka K and Kao C C2006 J. Appl. Phys. 99 08C302

[25] Cohen Y, Landskron K, Tetreault N, Fournier-Bidoz S,Hatton B and Ozin G A 2005 Adv. Funct. Mater. 15 593–602

[26] Fan H Y, Wright A, Gabaldon J, Rodriguez A, Brinker C J andJiang Y B 2006 Adv. Funct. Mater. 16 891–5

[27] Dai Z R, Sun S H and Wang Z L 2001 Nano Lett. 1 443–7[28] Shima T, Takanashi K, Takahashi Y K and Hono K 2004 Appl.

Phys. Lett. 85 2571–3

5