Magnetic Properties of FexPtyAu100_X_y Nanoparticles

Transcription

1 Copyright 20 American Scientific Publishers Journal of An rights reserved Nanoscience and Nanotechnology Printed in the United States of America Vol., , 20 Magnetic Properties of FexPtyAu0_X_y Nanoparticles Vikas Nandwana, Girija S. Chaubey, Yunpeng Zhang, and J. Ping Liu* Department of Physics, The University of Texas at Arlington, Arlington, Texas 76019, USA FexPtyAu1Oo_x_y nanoparticles of size 3.5 nm were prepared by polyol reduction of platinum acetylacetonate and gold acetate and the thermal decomposition of iron pentacarbonyl. The as-synthesized nanoparticles with disordered fcc structure were then heat treated to transform to the L1 0 structure with high magnetocrystalline anisotropy. By tuning the stoichiometry of the FexPtyAu1Oo_x_y nanoparticles, the phase transition temperature was reduced by mor~ than 200 DC. After the annealing 500 DC, for instance, the highest coercivity of 18 koe was obtained from the Fe51 Pt 36 Au 13 nanoparticles which is substantially higher compared to 2 koe for Fe51 Pt 9 nanoparticles annealed at the same temperature. In addition to the high coercivity, the saturation magnetization value obtained from Fe 51 Pt 36 Au 13 nanoparticles was 7 emug which is similar to that for the Fe51 Pt 9 nanoparticles, indicating that there is no trade-off between the coercivity and the saturation magnetization upon Au doping. Keywords: Nanoparticles, Permanent Magnets, Data Storage, Biomedical Applications, Chemical synthesis. INTRODUCTION Chemically synthesized FePt nanoparticles l have drawn considerable interest for future data storage media,2 high performance permanent magnets,3 and biological tlpplications. However, as-synthesized FePt nanoparticles are of disordered fcc structure and should be annealed at above 500 DC to convert to ordered LI o structure.1 The 'llildealing of the as-synthesized fcc nanoparticles results in particle sintering and grain growth, leading to wide size distributions and deterioration of the monodispersity and nanomorphology of the particles. 5 To overcome these difficulties many additives such as, Ag, Cu, Sb, V, Pd, and Au6-11 have been used as doping elements to decrease the transition temperature. Although the ordering temperature from fcc to Ll o FePt phase was decreased as high as 150 DC, the promoted grain growth due to the additives complicates the true effect of doping. 12 Moreover, due to the non-magnetic nature of the additives, the saturation magnetization of FePt nanoparticles drops significantly upon the doping. In this paper, we report a <;omposition controlled synthesis of Au doped FePt nanoparticles. Instead of just scanning Au composition while keeping Fe and Pt equiatomic, we have also done composition scan of all Fe, Pt and Author to whom correspondence should be addressed. Au in FexPtyAuIOO_x_y nanoparticles. It has been found that by tuning the concentration of Fe, Pt and Au in FexPtyAuIOO_x_y nanoparticles, the coercivity as high as 18 koe was obtained which is substantially higher than 2 koe of the equiatomic FePt nanoparticles annealed at the same temperature. As a result, the phase transition temperature was reduced by more than 200 dc. Also, the similar values of saturation magnetization of Au-doped FePt nanoparticles to equiatomic FePt nanoparticles suggest that Au doping enhances the coercivity without sacrifice in saturation magnetization if their compositions are well controlled. By comparing coercivity of FePt and FePtAu nanoparticles'of similar grain size, we also found that the increment of coercivity was primarily due to the Au doping and not due to promoted grain growth. 2. EXPERIMENTAL DETAILS FexPtyAuIOO_x_y nanoparticles were prepared by the polyol method reported by Kang et al. II However, no additional reducing agent was used in this case and Au precursor was reduced with heating in presence of surfactants. All the reactions were carried out using standard schlenk line technique. In brief, 0.5 rnmol of platinum acetylacetonate (Pt(acach) and designated amount of gold acetate (Au(ac)3) were added to 125 ml flask containing a magnetic stir bar and mixed with ml of phenyl ether. After J. Nanosci. Nanoechnol. 20, Vol., NO doi:l0.1166ijnn

2 Magnetic Properties of FexPtyAuIOO_x_y Nanoparticles purging with argon for 30 minutes at room temperature, the flask was heated up to 80 0c. Then, 0.5 mmol oleic acid, 0.5 mmol oleyl amine, and designated amount of iron pentacarbonyl (Fe(CO)s) were added and then the solution was refluxed for 30 minutes before cooling down to room temperature. There was no additional reducing agent added in the reaction since the reduction potential of Au and Pt is too low, therefore Pt(acac)z and Au(ac)3 can be reduced by oleyl amine itself After being washed in ethanol three or more times, the as-synthesized FexPtyAuIOO_x_y nanoparticles were dispersed in hexane and stored in glass bottles under refrigeration. Samples for magnetic characterization were prepared by depositing a drop of the final hexane dispersion on a silicon substrate and carbon coated Cu transmission electron microscopy (TEM) grid, evaporating the solvent at room temperature and further drying in vacuum, which led to the formation of FexPtyAuIOO_x_y nanoparticle assemblies. The samples were then annealed at 300 to 500 C for I hour under the flow of forming gas (Ar +7% Hz) in a tube furnace. The TEM images were recorded on a leol 1200 EX electron microscope at an accelerating voltage of 120 ky. Powder X-ray diffraction (XRD) spectra were recorded with a Cu Ka X-ray source (.A = 505 A). The magnetic hysteresis measurements were carried out by using superconducting quantum interference device (SQUID) magnetometer with magnetic field up to 7 T. The composition analysis was done by energy dispersive X-ray spectroscopy (EDX). 3. RESULTS AND DISCUSSION Table I shows the synthetic conditions of the Au-doped FePt nanoparticles. The molar concentration of Fe and Pt precursors were kept constant while molar concentration of Au precursor was varied. Effort was made to control the atomic ratio of Fe to Pt close to unity in FexPtyAuIOO_x_y nanoparticles. The as-synthesized FePt nanoparticles with different atomic percent of Au additive were annealed at 500 C and their corresponding coercivity and saturation magnetization was measured. As shown in the table, initially the coercivity of FexPtyAuIOO_x_y nanoparticles increased with increasing concentration of Au and the maximum coercivity of 13 koe was obtained Fig. Nandwana et at. TEM image of as-synthesized Fe2 Pt1 AU 17 nanopartic1es. for FePt nanoparticles having 17 atomic percent of Au. In comparison, pure FeS1 Pt9 nanoparticles yielded only 2 koe of coercivity when annealed at the same temperature. However, the coercivity of the FexPtyAuIOO_x_y nanoparticles was found to decrease when Au concentration was increased above 17 atomic percent. On the other hand, unlike the coercivity trend, the magnetization monotonously decreased with increase of Au content. It is clearly seen from the table that, FeZPtIAu17 nanoparticles have the highest coercivity, and hence was selected for further structural characterization. Figure I shows a TEM image of the as-synthesized FeZPt1 Au 17 nanoparticles of 3.5 nm. The TEM image reveals that the particles were monodisperse with a narrow size distribution. Figure 2(a) shows the XRD patterns of as-synthesized FeS1Pt9 and FeZPt1Au17 nanoparticles. The peak positions and intensity ratio of both samples match well with disordered fcc structure of FePt. It is seen that the (Ill) peak of FeZPt1 Au17 nanoparticles was shifted to low angle in comparison to Fe SI Pt 9 nanoparticles, indicating the lattice expansion of FePt nanoparticles after the addition of Au. The resulted lattice expansion of FeZPt1Au17 nanoparticles was due to the larger atomic volume of Au (.2 cm 3 mol) than that of Fe (7.1 cm 3 mol) and Pt (9.1 cm 3 mol).is The particle size calculated for FeZPt1 Au 17 nanoparticles using Scherrer's formula 16 is consistent with the TEM image. Table I. Composition and magnetic properties of FexPtyAUloo_x_y nanopartic1es annealed at 500 'C for I hour. Coercivity after Saturation Magnetization Fe(CO)s Pt(acac)2 Au(ac)3 Composition of annealing at after annealing (mmol) (mmo!) (mmol) FexPtyAuIOO_x_y 500 'C (koe) at 500 'C (emug) FeSl Pt Fe<iPtSA~ Fe..Pt3 Au Fe2 Pt1 AU Fe 38 Pt 37 Au Fe3SPt3SAu J. Nanosci. Nanotechnol., 297~2983, 20

3 Nandwana et a (8) 30.a eo 2 Theta (deg.) - 3OO C C --QO C (111) -- eoo C Au (200) (001) (1) (111) (002) (201) 0 eo 2 Theta (deg) Fig. 2. XRD patterns of (a) as-synthesized Fe S,Pt9 and Fe2Pt,Aul7 nanoparticles (b) Fe 2 Pt,AU 17 nanoparticles after annealing at temperas from 300 'C to 500 'C. The disordered fcc Fe2Pt1Au \7 nanoparticles have low magnetocrystalline anisotropy, and hence were '!;~Juperparamagnetic at room temperature. To obtain 'C c;6,igh uniaxial magnetocrystalline anisotropy, Fe2Pt 1 Au\7 'nanoparticles were annealed at temperatures between ~OO e to 500 e. Figure 2(b) illustrates the XRD patterns Magnetic Properties of FexPtyAuIOO_X_y Nanoparticles of Fe2Pt1 Au \7 nanoparticles annealed at different temperatures. The phase transition from disordered fcc to ordered LIo can be seen as emergence of the superlattice (001) and (1) peaks. As shown in the figure, the appearance of weaker superlattice peaks starts at temperature as low as 350 e, suggesting that partial ordering of fcc to LI o in Fe2Pt1 Au \7 nanoparticles began around this temperature. It can be also seen that with increasing annealing temperature, the intensity of the superlattice peaks became stronger; indicating that increase of annealing temperature favors the phase transition. The appearance of (Ill) peak of Au was also observed with increasing annealing temperature as reported previously6, which became more clear with increasing annealing temperature. The separate Au peak was detected for particles with all the different Au compositions. It appears that Au atoms left the FePt lattice during the annealing, and created lattice vacancies. These vacancies increased the mobility of Fe and Pt atoms, thus enhanced the kinetics of the ordering process. To complete the loss in saturation magnetization caused by Au doping, Fe concentration in FexPtyAuIOO_x_y nanoparticles was increased. Table 11 shows composition and magnetic properties of the samples in which the molar concentration of Pt and Au precursors were kept identical to that used to synthesize Fe2Pt1Au\7 nanoparticles in Table I while the molar concentration of Fe precursor was varied. The coercivity and saturation magnetization values given in Table 11 were obtained from FexPtyAuIOO_x_y nanoparticles annealed at 500 0c. It was interesting to observe that along with the saturation magnetization, the coercivity of FexPtyAuIOO_x_y nanopar~ ticles also increased with increasing concentration of Fe and maximum coercivity of 18 koe was obtained from FeSl Pt36Au13 nanoparticles. However, it is not clear how the increase in Fe-content enhanced the coercivity. Nevertheless, further increase of Fe concentration resulted in decrease of coercivity as shown in Table 1 Figure 3(a) shows the dependence of coercivity on annealing temperature of Fes 1Pt 9, FePt3Au 13 and Fes1Pt36Au13 nanoparticles. Similar concentration of Au was chosen in both the AU-doped FePt nanoparticles for comparison. It can be clearly seen from the figure that at the same annealing temperature, the coercivity of both FePt3Au13 and FeSlPt36Au13 nanoparticles was higher II. Variation of coercivity and saturation magnetization of FexPtyAulOo_x_y nanoparticles annealed at 500 'C for I hour with increasing qoncentration of Fe. Coercivity after Saturation Magnetization Pt(acac)2 Au(ac)s Composition of annealing at after annealing (mmol) (mmol) FexPtyAuIOO_x_y 500 'C (We) at 500 'C (emug) Fe 2 Pt,AU Fe6Pt'9Au,S Fes, Pt s6 Au Fe SS Pt 33 Au' FeooPtsoAu lo

4 Magnetic Properties. of FexPtyAuIOO_x_y Nanoparticles (a) 28,-;:::==========::::; , 11 1 r ~ 12 (b) 2 o -.- FeflPt.. -e- Fe Pt "'--1S -.- Fe flo Pt...'.6uis o SOl 00 lioo Ann..llng Tempenrtu,-. ('C) Nandwana et ai. nanoparticles while the coercivity was significantly higher than both FePt3Au13 and FeSt Pt9 nanoparticles. These results clearly show that by tuning the stoichiometry of the FexPtyAulOO_x_y nanoparticles. Au doping enhances the coercivity of FePt nanoparticles without sacrifice of saturation magnetization. It has been reported previouslyl2 that addition of Au in FePt nanoparticles decreases the ordering temperature; however it promotes the grain growth during heat treatments due to the low surface energy of Au. Since grain growth of FePt nanoparticles during annealing also promotes the chemical ordering, it was not clear whether the decrease in ordering temperature was due to Au doping or due to the enhanced grain growth during annealing. Figure (a) shows the grain size of Fest Pt9 and FeStPt36Aul3 nanoparticles as a function of annealing temperature. The grain size was calculated from the XRD curves of annealed nanoparticles using Scherrer's FIeld (koe) Fig. 3. (a) Coercivity dependence of Fes' Pt9' Fe..Pt3Au ll,. and FeSlPt36Aull nanoparticles on annealing temperature. (b) Hysteresis loops of Fes' Pt9, Fe..Pt3 Aull and FeS,Pt36Aull nanopartic1es after annealing at 500 C for I hour in forming gas. than Fest Pt9 nanoparticles. It can also be seen that by keeping atomic percent of Au constant, increasing FelPt atomic ratio from I to resulted in a significant increase in coercivity. The coercivity of Fest Pt36Au13 nanoparticles after annealing at 300 C was 2.6 koe while the coercivity of Fe sl Pt9 nanoparticles was negligible at this temperature. To achieve similar coercivity in Fes, Pt9 nanoparticles, the particles have to be annealed above 500 DC, t which suggests that for the FeSt Pt36Au 13 nanoparticles, the phase transition temperature from fcc to Ll o structure was reduced by more than 200 DC. Figure 3(b) shows the hysteresis loops of FeSI Pt9, FePt3Aul3' and FeStPt36Au13 nanoparticles after annealing at 500 C for I hour in forming gas. As shown in the figure, the coercivity of FePt3Au13 nanoparticles is higher than that of Fest Pt9 nanoparticles; however the saturation magnetization is lower. In comparison, the saturation magnetization of FeSt Pt36Au13 nanoparticles was close to Fest Pt (a) 1 is (b) 20 1 ~12 f: -.-Fe Pt Fe PtaaAu >.1 1S _ ~---. A I Ann..llng Tempenrtu,-. (C) -.-Fe fl Pt Fe. 1 PtaaAu 13 A.'" ~ ~--. 0.~ Oraln Size (nm) 12 1 Fig., (a) Grain size dependence of FeSI Pt9 and FeS,Pt36Aul3 nanoparticles on annealing temperature. (b) Correlation of grain size and coercivity for Fes' Pt9 and Fes' Pt 36 Au 13 nanoparticles.

5 Narulwana et al. formula. 16 It can be seen that the grain size of Fe51Pt36Aul3 nanoparticles is larger than that of Fe51 Pt 9 nanoparticles at similar annealing temperatures which indicates that presence of Au promoted the grain growth in Fe51Pt36Au13 nanoparticles. To study the true effect of Au doping on the phase transition, coercivity versus grain size was plotted for Fe51 Pt 9 and Fe51Pt36Aul3 nanoparticles. It is seen from the Figure (b) that the coercivity of Fe51 Pt36Au13 was found higher than that for Fe 51 Pt 9 particles for similar grain size which suggested that even though Au promoted the grain growth, the decrease in ordering temperature in Fe51 Pt 36 Au 13 nanoparticles was primarily due to the Au doping.. CONCLUSION We have prepared FexPtyAuIOO_x_y nanoparticles by polyol reduction of platinum acetylacetonate and gold acetate and thermal decomposition of iron pentacarbonyl. The as-synthesized nanoparticles have their size of 3.5 nm with disordered fcc structure. Upon annealing, the particles transform to ordered Llo structure with high magnetocrystalline anisotropy. After annealing the samples at 500 DC, the highest coercivity of 18 koe was obtained from the Fe51 Pt36Au13 nanoparticles which is substantially higher compared to 2 koe from Fe 51 Pt 9 nanoparticles annealed at the same temperature. It is found that with right stoichiometry of the FexPtyAuIOO_X_y nanoparticles, the phase transitjon temperature can be reduced by more than 200 dc. The similar values of saturation magnetization of Fe51Pt36Au13 and Fe 51 Pt 9 nanoparticles imply that Au doping enhances the coercivity without sacrifice in saturation magnetization. The higher coercivity of Fe51 Pt36Au13 than that of Fe51Pt 9 Magnetic Properties of FexPtyAuIOO_x_y Nanoparticles particles for similar grain size suggests that the increase in coercivity was primarily due to Au doping. Acknowledgments: This work was supported by Center of Nanostructured Materials and Characterization Center for Materials and Biology at University of Texas at Arlington. References and Notes S. Sun, C. B. Murray, D. Weller, L. Folks, and A. Moser, Science 287, 1989 (2000). 2, D. Weller and A. Moser, IEEE Trans. Magn. 35, 23 (1999). 3. H. Zeng, J. Li, P. Liu, Z. L. Wang, and S. Sun, Nature 20, 395 (2002).. Q. A. Pankhurst, J. Connolly, S. K. Jones, and Dobson, J. Phys. D: Appl. Phys. 36, RI67 (2003). 5. Z. R. Dai, S. Sun, and Z. L. Wang, Nano Lett. 1,3 (2001). 6. S. Kang, J. W. Harrell, and D. E. Nikles, Nano Lett. 2, 33 (2002). 7. X. Sun, S. Kang, J. W. Harrell, and D. E. Nikles, Z. R. Dai, Li, and Z. L. Wang, J. Appl. Phys. 93, 7337 (2003). 8. Q. Yan, T. Kim, A. Purkayastha, P. G. Ganesan, M. Shima, and G. Rarnanath, Adv. Mater. 17,2233 (2005). 9. Luo, L. Han, N. N. Kariuki, L. Wang, D. Molt, C. J. Zhong, and T. He, Chern. Mater. 17,5282 (2005).. S. Kang, Z. Jia, D. E. Nikles, and W. Harrell, J. Appl. Phys. 95, 67 (200). 1 S. Kang, Z. Jia, D. E. Nikles, and J. W. Harrell, IEEE Trans. Magn. 39, 2753 (2003). 12. Z. Jia, S. Kang, S. Shi, D. E. Nikles, and W. Harrell, J. Appl. Phys. 97, 13 (2005). 13. V. Nandwana, K. E. Elkins, and J. P. Liu, Nanotechnology 16, 2823 (2005). S. Pyrpassopoulos, D. Niarchos, G. Nounesis, N. Boukos, I. Zafiropoulou, and V. Tzitzios, Nanotechnology 18,8560 (2007) B. D. Cullity, Introduction to Magnetic Materials, Addision-Wiley, London (1972). Received: 19 February Accepted: 18 June