Synthesis and Characterization of Structure Controlled Nano-cobalt Particles

Shiqiang (Rob) Hui1, Mingzhong Wu1, Shihui Ge1, Dajing Yan1, Y.D. Zhang1 *, T.D. Xiao1, M. J. Yacaman2, M. Miki-Yoshida3, W. A. Hines4, and J. I. Budnick4,

1Inframat Corporation, Farmington, CT 06032

2Department of Chemical Engineering, University of Texas, Austin, TX 78712 3Texas Materials Institute, University of Texas at Austin, Austin, TX 78712-2201

4Physics Department and IMS, University of Connecticut, Storrs, CT 06269 ABSTRACT Nanostructured cobalt particles with and without a ceramic coating have been synthesized using a wet chemical method. The structure and magnetic properties of synthesized powder were characterized using x-ray diffraction (“XRD”), high-resolution transmission electron microscopy (“HRTEM”), and a Quantum Design (SQUID) magnetometer. The cobalt nanoparticles are of either face-centered cubic (“fcc”) and/or hexagonally close-packed (“hcp”) crystalline structures. The average grain size is ~14 nm for cobalt (either fcc or hcp) with an amorphous silica coating, and the average grain size is ~9 nm for hcp cobalt and 26 nm for fcc cobalt without a silica coating. The effect of annealing temperature on grain size and magnetic properties are addressed. INTRODUCTION

Cobalt has been known with allotropic forms including fcc, hcp, epsilon (“�”), and body-centered cubic (“bcc”). Hull first reported the existence of fcc and hcp-cobalt after analyzing different patterns of metallic powders prepared by several methods in 1921 [1]. The ε-cobalt, a complex cubic primitive structure (P4332), was recently recognized by Dinega et al. through a detailed structure analysis [2]. A non-equilibrium bcc structure, which does not naturally occur in bulk materials, was also obtained using epitaxial growth [3]. The fcc structure is thermodynamically preferred at higher temperatures and the hcp structure is favored at lower temperatures. The ε-cobalt can be converted to common hcp and fcc-cobalt by annealing at temperatures of 300oC and 500oC, respectively [4, 5]. However, the fcc structure appears to be stable phase even below room temperature when the particle sizes are smaller than 20nm [6]. Temperatures of about 200oC are enough to trigger atom diffusion and phase transformations for nanostructured cobalt crystals [7]. The fcc and hcp phases of cobalt are close-packed and nearly isoenergetic crystal structures that differ only in the stacking sequence of atomic planes in the [111] direction. Low activation energy for formation of stacking faults often results in the formation of both phases in the same sample under high-temperature crystallization techniques, such as melting-crystallization and evaporation-condensation. On the other hand, ε-cobalt is most often found in nanostructured particles prepared at low temperatures by solution phase

* Author to whom correspondence should be addressed; email: yzhang@inframat.com, Tel: 860-487-3838, Fax: 860-429-5911, Inframat Corp., Willington, CT 06279

DD5.20.1Mat. Res. Soc. Symp. Proc. Vol. 755 © 2003 Materials Research Society

chemical synthesis technique, which is generally not thermodynamically controlled and thus can allow the preparation of metastable phases [2, 4].

Cobalt nanocrystals display a wealth of size-dependent structural, magnetic, electronic, and catalytic properties. There has been a considerable amount of research involving the preparation, structure, and properties of magnetic cobalt nanoparticles in the past decade [8-17]. The high magneto crystalline anisotropy of hcp Co has spurred intensive studies of Co-based nanostructures for magnetic storage purposes [18]. Cobalt nanoparticles coated with insulators have been prepared and studied for the applications in AC electrical and electronic devices [19-21]. With the growing interest in building advanced materials using nanoscale building blocks, there is a need to control the sizes, shapes, and structures. Different crystalline structures provide some practical benefits. Symmetrically structured fcc-cobalt provides higher saturation moment and lower magnetic anisotropy, which is suitable for linear applications such as converter. On the other hand, asymmetric hcp-cobalt has been a basic element for high temperature permanent magnets, and is also suitable for microwave applications such as in circulator. To date, making magnetic cobalt nanocrystals has been difficult and required costly size-selective precipitation methods. In practice, cobalt nanoparticles often possesses mixed structures in which low energy stacking faults introduce a combination of fcc and hcp character, making the synthesis of single structured cobalt a challenging task. Puntes et al achieved size and shape-controlled cobalt nanorods as well as spherically shaped nanocrystals [22]. In this work, we reported the successful preparation of structure-controlled cobalt nanoparticles with or without coating. EXPERIMENTAL Cobalt nanoparticles with or without an amorphous SiO2 coating were synthesized by a wet chemical approach [23]. The main procedures included: (1) preparing the starting precursors containing cobalt and silicon, (2) atomizing the precursors to make colloidal solutions, (3) annealing the solutions to form a pre-composite complex, and (4) low temperature calcination to form cobalt or SiO2-coated cobalt nanoparticles. The cobalt concentration in the synthesized nanoparticles was found to be dependent mainly on the precursor concentration and the calcination temperature. In this paper, the nanoparticle samples were prepared by using the same precursor but different calcination temperatures of 600, 700, 800 and 900 °C. The volume fraction of cobalt in these samples was targeted at 50%. The structure of the Co nanoparticles was determined by powder XRD analysis using Cu Kα1 radiation. Morphology and microstructure was analyzed using HRTEM. The analysis was performed in a JEOL-2010F high resolution field emission electron microscope, with a point to point resolution of 0.19 nm. The microscope was also coupled with an EDS system to perform nanobeam analysis. Magnetic properties were studied using a SQUID magnetometer.

RESULTS AND DISCUSSION A series of cobalt samples have been prepared and their XRD patterns were shown in Fig. 1. It is evident that the as-synthesized nanopowders correspond to either hcp or fcc structure of

DD5.20.2

cobalt depending on the controlled processing conditions, including starting materials, additives, and heat-treatment temperatures [23]. The average grain size of the synthesized cobalt nanoparticles was estimated by the Scherrer formula from the (101) and (111) diffraction peak for hcp and fcc-cobalt, respectively [24]. A value of 9nm for hcp-cobalt and 26nm for fcc-cobalt was obtained. It illustrates clearly from the XRD pattern that the (101) hcp-cobalt diffraction peak was further broadened than other reflection peaks. This is most likely due to the formation of stacking faults, leading to energetically growth along the b-axis, indicating that the synthesized nanoparticles might be in the shape of rods rather than other shapes. A tempt to directly observe the detailed microstructure by TEM was not successful. It is difficult to make isolated nanoparticles of cobalt for TEM observation because of large attractive van der Waals and magnetic forces between the particles. Similarly, silica-coated cobalt (Co:SiO2 50:50 in vol) nanoparticles were prepared under different conditions and the XRD patterns for the as-synthesized samples were shown in Fig. 2. The as-synthesized samples can be signed as either hcp or fcc-cobalt with estimated being approximately the same average grain size of 14nm from the XRD peak broadening analysis. There were no peaks corresponding silica, indicating that the silica was most likely in its amorphous state. HRTEM studies were carried out to study the nanostructure of these samples. A typical nanoparticle image for sample 121401 and 52024 was shown in Fig. 3. (a) and (c), respectively. In addition, an enlarged FFT filtered image of squared zone was shown in Fig. 3. (b) for sample 121401, which shown atomic resolution images of planes (111) and (-111), corresponding to the cubic phase of Co (PDF #15-0806). Fig. 3. (c) shows characteristic images of a Co particle surrounded by SiO2. A core-shell structure is clearly evident in the HRTEM image. The enlarged zone in Fig. 3. (d) shows FFT filtered image of the atomic resolution of planes (101) and (002), corresponding to a hexagonal structure (PDF # 05-0727) for sample 52024. The average particle size is about 18nm from the analysis of TEM particle distribution for both samples, which is consistent with the calculation from XRD. The slight

20 30 40 50 60 70 80 90 100

311

112

201

004

111

200

220

222

002

101

110

103

100

Inte

nsity

2θ

(b) #43022

(c) #424021

(a) #32602

20 30 40 50 60 70 80 90 100

222

311

22020

0

111

100

(a) #52024

2θ

Inte

nsity

(b) #121401

102

20111

2

10311

0

101

002

Fig. 2. X-ray diffraction patterns for as-synthesized nanocobalt particles coated with silica (Co:SiO2 50:50 in vol): (a) hcp, (b) fcc.

Fig. 1. X-ray diffraction patterns for as synthesized cobalt nanoparticles with different structures: (a) hcp, (b) mixed hcp and fcc, and (c) fcc.

DD5.20.3

10 nm

(a)

(111)

(-111) (b)

10 nm

(c)

(101)

(002)

(d)

Fig. 3. HRTEM images for: (a) sample 121401 and (c) sample 52024. Enlarged zone shows atomic resolution images of planes: (b) (111) and (-111) for sample 121401 and (d) (101) and (002) for sample 52024. difference between these two values is due to the methodology difference between XRD and TEM.

The silica-coated cobalt nanoparticles with fcc structure were prepared at different

temperatures form 600 – 900 oC in order to study the property-temperature relationship. From the (111) diffraction peak, the average grain size of the inner cobalt core in the synthesized Co/SiO2 nanoparticles was estimated (see Fig. 4). The cobalt grain size increased slightly with temperature when the calcination temperature was 800 °C or below. At above 800 oC, the cobalt grain size increased significantly. This indicates that when the calcination was conducted at 800 °C or below, the SiO2 coating remained as an unbroken shell and prevented the coarsening of the

DD5.20.4

cobalt nanoparticles. The SiO2 coating no longer acted as a barrier to effectively to prevent abnormal cobalt grain growth at 900 °C.

Magnetic properties of the synthesized cobalt nanoparticles were also studied for these samples prepared at different temperatures. Fig. 4 shows the coercivities measured at 10 K and 300 K as a function of calcination temperature. The coercivity of the cobalt/silica nanocomposite material decreases with increasing calcination temperature. Since the cobalt phase in the nanoparticles calcined at lower temperature has a smaller effective size due to the presence of the inner Co-oxide core, it can be inferred from Fig. 4 that the coercivity decreased with increasing particle size. This result is consistent with the theoretical analysis given in reference [25], where R. C. O’Handley argued that the coercivity went as braH c −≈ 2 (r, particle size) for non- interacting single-domain fine particles. The cobalt nanoparticles in our measured samples are free from interaction, and their sizes are smaller than the single-domain critical size (76 nm) of fcc cobalt particles [17].

Fig. 4 also reveals that the coercivities measured at 10 K are notably higher than those measured at 300 K for the nanoparticles calcined at 600, 700, and 800 °C. This is believed to be due to the Co/CoO exchange coupling, which occurs at 10 K and disappears at 300 K. Besides, although the nanoparticles calcined at low temperatures exhibited high coercivities up to 984.3 Oe, the nanoparticles calcined at 900 °C exhibited low coercivities of 126.7 Oe at 10 K and 63.3 Oe at 300 K. This indicates that, using the engineered wet chemical approach, the magnetic softness of the nanoparticles can be readily controlled by the calcination temperature. CONCLUSION

Cobalt nanoparticles with or without silica coating were synthesized using the engineered wet chemical approach. Their structure was found to vary with preparation conditions and could be controlled as either hcp or fcc structure. For silica-coated cobalt nanoparticles, their gain size

0

500

1000

1500

500 600 700 800 900 1000

Calcination temperature (C)

Coe

rciv

ity (

Oe)

20

30

40

50

Particle size (nm

)

Coercivity at 10 KCoercivity at 300 KParticle size (nm)

Fig. 4. Change of magnetic property and grain size as a function of temperature for silica-coated nanoparticles of cobalt.

DD5.20.5

increased slightly when the calcination temperature was lower than 800 °C. The silica shell as observed by HRTEM hindered the grain growth of cobalt nanoparticles during the synthesis.

ACKNOWLEDGMENTS This work was supported by DARPA through contract No. F7-6AM945-X05. REFERENCES 1. A.W. Hull, Phys. Rev. 17, 571 (1921). 2. D.P. Dinega and M.G. Bawendi, Angew. Chem. Int. Ed. Engl. 38, 1788 (1999). 3. S. Sun, C.B. Murray, and H. Doyle, in Advanced Hard and Soft Magnetic Materials, eds. M.

Coey, L.H. Lewis, B.M. Ma, T. Schrefl, L. Schultz, J. Fidler, V.G. Harris, R. Gasegawa, A. Inoue, and M. McHenry, Met. Res. Soc. Symp. Proc. 577, Warrendale, PA, p.385 (1999).

4. S. Sun and C.B. Murray, J. Appl. Phys. 85, 4325 (1999). 5. A. Y. Liu and D. J. Singh, Bcc cobalt: metastable phase or forced structure? J. Appl. Phys.

73, 6189 (1993). 6. O. Kitakami, H. Satao, Y. Shimada, F. Sato and M. Tanaka, Phys. Rev. B 56, 13849 (1997). 7. M.H. Yang and C.P. Flynn, Phys. Rev. Lett. 62, 1476 (1989). 8. J.F. Loffler, H.B. Braun, W. Wagner, G. Kostorz, and A. Wiedenmann, Materials Science

and Engineering A 304, 1050 (2001). 9. S. Gangopadhyay, G.C. Hadjipanayis, C.M. Sorensen, and K.J. Klabunde, IEEE

Transactions on Mangnetics 28, 3174 (1992). 10. O. Kitakami, H. Sato, and Y. Shimada, Physical Review B 56, 13849 (1997). 11. C.J. Choi, X.L. Dong, and B.K. Kim, Scripta Materialia 44, 2225 (2001). 12. M. Jamet, V. Dupuis, C. Thirion, W. Wernsdorfer, P. Melinon, and A. Perez, Scripta

Materialia 44, 1371 (2001). 13. S. Ram, Materials Science and Engineering A 304, 923 (2001). 14. A.H. MacDonald and C.M. Canali, Solid State Communications 119, 253 (2001). 15. E.E. Carpenter, C.T. Seip, and C.J. O’Connor, J. Applied Physics 85, 5184 (1999). 16. W. Wernsdorfer, E. Bonet Orozco, K. Hasselbach, A. Benoit, B. Barbara, N. Demoncy, A.

Loiseau, H. Pascard, and D. Mailly, Physical Review Letters 78, 1791 (1997). 17. Y.D. Zhang, J.I. Budnick, W.A. Hines, S.A. Majetich, and E.M. Kirkpatrick, Applied Physics

Letters 76, 94 (2000). 18. K. O,Grady and H. Laidler, J. Magnetism and Magnetic Materials 200, 616 (1999). 19. Y.D. Zhang, S.H. Wang, D.T. Xiao, J.I. Budnick, and W.A. Hines, IEEE Transactions on

Magnetics 37, 2275 (2001). 20. M. Wu, Y. D. Zhang, S. Hui, S. Ge, W.A. Hines, and J.I. Budnick, J. Applied Physics 92,

491 (2002). 21. M. Wu, Y. D. Zhang, S. Hui, S. Ge, W.A. Hines, and J.I. Budnick, Applied Physics Letter 20,

4404 (2002). 22. V.F. Puntes, K.M. Krishnan, A.P. Alivisatos, Science 291, 2115 (2001). 23. Y.D. Zhang, S.H. Wang, and D.T. Xiao, Co-based magnetic nanocomposite materials and the

synthesis method, U.S. Patent Appl. No. 60/243,649 (October 26, 2000). 24. H. Lipson and H. Steeple, Interpretation of X-Ray Powder Diffraction Patterns (St Martin’s

Press, New York, 1970), p. 256. 25. R.C. O’Handley, Modern Magnetic Materials (John Wiley & Sons, New York, 2000), p.

435.

DD5.20.6