Ultrathin Interface Regime of Core–Shell Magnetic Nanoparticles for

Letter pubs.acs.org/NanoLett

Ultrathin Interface Regime of Core−Shell Magnetic Nanoparticles for Effective Magnetism Tailoring Seung Ho Moon,†,‡,§ Seung-hyun Noh,†,‡,§ Jae-Hyun Lee,†,‡,§ Tae-Hyun Shin,†,‡,§ Yongjun Lim,†,‡,§ and Jinwoo Cheon*,†,‡,§ †

Center for Nanomedicine, Institute for Basic Science (IBS), Seoul 03722, Republic of Korea Yonsei-IBS Institute, Yonsei University, Seoul 03722, Republic of Korea § Department of Chemistry, Yonsei University, Seoul 03722, Republic of Korea ‡

S Supporting Information *

ABSTRACT: The magnetic exchange coupling interaction between hard and soft magnetic phases has been important for tailoring nanoscale magnetism, but spin interactions at the core−shell interface have not been well studied. Here, we systematically investigated a new interface phenomenon termed enhanced spin canting (ESC), which is operative when the shell thickness becomes ultrathin, a few atomic layers, and exhibits a large enhancement of magnetic coercivity (HC). We found that ESC arises not from the typical hard−soft exchange coupling but rather from the large magnetic surface anisotropy (KS) of the ultrathin interface. Due to this large increase in magnetism, ultrathin core−shell nanoparticles overreach the theoretical limit of magnetic energy product ((BH)max) and exhibit one of the largest values of specific loss power (SLP), which testifies to their potential capability as an effective mediator of magnetic energy conversion. KEYWORDS: Hard−soft exchange coupling, core−shell magnetic nanoparticle, shell-thickness control, surface canted spins, magnetic anisotropy, magnetic energy product

O

We started our investigation with the synthesis of bimagnetic core−shell nanoparticles, composed of a hard magnetic CoFe2O4 core and a soft magnetic MnFe2O4 shell (CF@MF) with a controlled volume ratio, via a seed-mediated growth method (see Supporting Information (SI) section 1).13 The cube-shaped nanoparticles with an edge length of 30 nm are used as a core material to take advantage of their low energy facets, being more sensitive to changes in magnetic environments.14,15 CF@MF core−shell nanoparticles with varying shell thicknesses of 0.5, 1, 1.5, 2.5, and 5 nm are synthesized and represented as the shell volume fraction (fshell = Vshell/Vtotal) to reflect the magnetic atomic ratio of the shell in the nanoparticle. As the shell grows thicker, fshell increases to 0.09, 0.18, 0.25, 0.37, and 0.58, which are referred to as [email protected], CF@ MF0.18, [email protected], [email protected], and [email protected], respectively (Figure 1a). Figure 1b shows transmission electron microscopy (TEM) images of CF core and CF@MF nanoparticles with a 2.5 nm shell thickness and high homogeneity in size (σ ≈ 0.5; see SI Figure S1 for TEM images of other CF@MF nanoparticles). The atomic contents of the shell component are quantified by inductively coupled plasma atomic emission spectroscopy (ICP-AES) analysis (SI Figure S2 and Table S1). The HRTEM image clearly shows continuous atomic lattice

ver the past decade, the coupling interaction between magnetically hard and soft phases has gained considerable attention as an efficient way to rationally design nanoparticles to control their magnetism,1−7 as demonstrated in the cases of exchange bias and two-stage magnetization reversal.2,8,9 In particular, magnetic coupling of the hard−soft phase is beneficial for tuning nanoscale magnetism since overall magnetic properties can be effectively adjusted by modulating the core−shell parameters such as shape, size, and chemical composition.2,5 For core−shell magnetic nanoparticles, two distinct regimes of magnetism can exist as the shell thickness varies: (1) the exchange coupling (EC) regime and (2) the enhanced spin canting (ESC) regime (Scheme 1). The EC regime is where the typical hard−soft exchange coupling takes place (Scheme 1b(i)), as already well described in the literature when the shell thickness is roughly twice the width of the domain wall size of the magnetically hard core phase (δw,hard).2 In this study, we report a unique phenomenon of ESC, which occurs at the ultrathin shell thickness within a few atomic layers, similar to the size of a crystalline unit-cell (Scheme 1b(ii)). From the investigation of the ESC regime, we found that an exceedingly disoriented spin structure of the shell with structural strains restricts both crystallographic surface stabilization and magnetic relaxation,2,10,11 which consequently leads to a spin-glass-like surface and an increase in KS (see vide inf ra).12 © 2017 American Chemical Society

Received: September 26, 2016 Revised: December 23, 2016 Published: January 3, 2017 800

DOI: 10.1021/acs.nanolett.6b04016 Nano Lett. 2017, 17, 800−804

Letter

Nano Letters

Scheme 1. Schematic Illustrations of Magnetic Spin Configurations and Their Magnetism in the EC Regime and ESC Regimea

a

(a) Magnetic spin state of a hard magnetic core nanoparticle. The core nanoparticle has aligned spins (black arrows) in its inner core and canted spins (orange arrows) on its surface. (b) Magnetic spin interactions in hard−soft bimagnetic core−shell nanoparticles. (i) In the EC regime, soft magnetic spins at the interface are pinned by hard magnetic spins. These spins are called exchange-coupled spins (blue arrows) and result in decreased HC and increased MS. (ii) In the ESC regime, where the shell thickness is similar to the size of a crystalline unit cell (∼1 nm), the population of canted spins (orange arrows) increases. These disoriented interfacial spins enlarge HC while maintaining MS.

fringes and successful epitaxial growth of a shell layer in the [email protected] nanoparticle (Figure 1c). The EDS-mapped image (Figure 1d(i−iv)) and line scan profile (Figure 1d(v)), in which cyan, red, and green colors correspond to the atomic distribution of manganese, cobalt, and iron, respectively, clearly show the core−shell structure. The detailed elemental distribution of nanoparticles is further identified by a depth profile using angle-resolved X-ray photoelectron spectroscopy (AR-XPS) with an argon-ion beam etching method (see SI Figure S3). The correlation of different shell volume fractions with magnetic properties is examined by measuring field-dependent magnetization from −2 to +2 T at 5 K using a superconducting quantum interference device (SQUID). As shown in Figure 2a, all nanoparticles display smooth hysteresis curves with no kinks, implying that the two magnetic phases are completely coupled into single-phase magnetism. The CF core shows an HC value of 3.3 kOe (Figure 2a,b). When the fshell of CF@MF is small and at the very thin layer of the ESC regime, [email protected], [email protected], and [email protected] exhibit the unusual phenomenon of significantly increased HC of 7.2, 5.3, and 4.5 kOe, respectively, which are at most 2.2 times higher than that of the CF core. As the shell grows thicker, [email protected] and CF@ MF0.58 show decreased HC of 0.9 and 0.8 kOe, respectively, which correspond to typical hard−soft EC properties (see SI Figure S4 for the more thick-shelled [email protected]). The blocking temperature (TB) is also measured from the peaks of zero-field-cooling (ZFC) curves and found to be 495, 477, and 470 K for [email protected], [email protected], and [email protected], respectively. All of these values are higher than that of the CF core (377 K) (Figure 2c), while [email protected] and [email protected] show TB of 375 and 370 K, similar to the CF core. These results indicate that HC and TB are dependent on the shell thickness in the hard−soft core−shell nanoparticles. Unlike the typical core−shell nanoparticles in the EC regime (i.e., [email protected]

and [email protected]), thin-shelled core−shell nanoparticles possess larger anisotropic energy barriers than the hard magnetic phase of the core, and these magnetic properties do not follow typical hard−soft exchange coupling.16 These phenomena are demonstrated to be still effective at higher temperatures of 150 and 300 K as well as for different shape (i.e., cuboctahedron) and size (i.e., superparamagnetic regime) (SI Figures S5−S7). To test the generality of these observations for ESC, various metal ferrites (XFe2O4, X = Mn, Fe, Co, and Ni) have been examined as ultrathin shell components. Here, core−shell nanoparticles of CoFe2O4@MnFe2 O4, CoFe2O4 @Fe3O4 , CoFe2O4@CoFe2O4, and CoFe2O4@NiFe2O4 with fshell of 0.09 are referred to as [email protected], [email protected], [email protected], and [email protected], respectively. The magnetization hysteresis loops and corresponding HC and MS values of these nanoparticles are measured at 5 K (Figure 3a,b). The HC values of [email protected], [email protected], [email protected], and [email protected] are 7.2, 6.9, 5.8, and 4.9 kOe, respectively. These are all higher than that of the CF core (3.3 kOe) and are consistent with the number of unpaired d-electrons of shell atoms (i.e., Mn, 5; Fe, 4; Co, 3; Ni, 2), while the MS values remain constant. Nearly unchanged MS values are regarded as a result from the negligible impacts of magnetic spins in the shell on net magnetization due to their randomly aligned orientation in the ESC regime.10 These results indicate that a substantial increase of HC is a general phenomenon that occurs at the ESC regime of bimagnetic core−shell nanoparticles, regardless of shell composition. Among the potential anisotropic factors, such as magnetocrystalline (Kmag), shape (Kshape), and surface (KS) anisotropy,17−19 KS induced by canted spin is known as a major factor for enhancing the effective magnetic anisotropy (Keff),20 which has a proportional relationship with the HC.19 To examine the effect of the spin structure discrepancy between the surface and core of a nanoparticle (i.e., KS) on the HC and KS values of the CF core, various CF@MF and CF@XF (X = M(Mn), I(Fe), 801

DOI: 10.1021/acs.nanolett.6b04016 Nano Lett. 2017, 17, 800−804

Letter

Nano Letters

Figure 1. CF@MF core−shell nanoparticles with tunable volume fraction. (a) Schematic illustrations of 30 nm CF core and CF@MF core−shell nanoparticles with various shell thicknesses of 0.5, 1, 1.5, 2.5, and 5 nm, which correspond to shell volume fractions of 0.09, 0.18, 0.25, 0.37, and 0.58, respectively. (b) TEM images of (i) 30 nm CF core and (ii) 35 nm [email protected] nanoparticles. (c) High-resolution TEM image of [email protected] nanoparticles. (d) Elemental mapping analyses of [email protected] nanoparticles. (i) Mn, (ii) Co, and (iii) Fe false-colored with blue, red, and green, respectively. (iv) Merged image of Mn, Co, and Fe. (v) Line-scanned EDS profile of Mn, Co, and Fe for a single [email protected] nanoparticle.

Figure 2. Magnetic properties of CF@MF core−shell nanoparticles with various shell volume fractions. (a) Field-dependent magnetization curves of CF core and various CF@MF nanoparticles at 5 K and (b) their HC values. (c) ZFC temperature-dependent magnetization curves of CF core and CF@MF nanoparticles. The peaks of curves (black arrows) indicate the TB of each nanoparticle. (d) Calculated values of KS for CF core and CF@MF nanoparticles with different shell volume fractions.

spin inversion process of CF core, [email protected] (i.e., ESC regime), and [email protected] (i.e., EC regime) nanoparticles, in which the spin angle is presented by color (+180°, red color; −180°, blue color). All spins of nanoparticles are initially aligned in the +180° direction, and then, the external magnetic field (B0) is applied in the opposite direction (−180°) to investigate spin angle deviations. The thin-shelled core−shell of [email protected] requires the highest magnetic field strength of 21.5 kOe for complete spin inversion, which is ca. 3.1- and 2.0-fold higher than those for [email protected] (7.0 kOe) and the CF core (10.7 kOe). These results show that the bimagnetic core−shell in the ESC regime is most resistant to the spin inversion, which is consistent with the trend of enhanced HC values of the nanoparticles (Figure 2b). Such interesting magnetism in the ESC regime can be utilized for versatile magnetic applications with high performance. For example, the maximum energy product, (BH)max, which is magnetostatic energy stored in magnetic materials and defined as the area in the second quadrant B−H loop, can be optimized with the ultrathin-shelled core−shell nanoparticles.22,23 The B−H curves and calculated (BH)max values for the CF core, [email protected], and [email protected] are illustrated in Figure 4a, where [email protected] represents the theoretical limit of (BH)max calculated from a conventional hard−soft exchange coupling, based on the Stoner−Wohlfarth approximation (see SI section 5). The highest-energy product, 1791 J/m3,

C(Co), and N(Ni)) nanoparticles are obtained from the relationship of measured TB and Keff (see SI section 3 and Table S2 for the calculation of KS). For CF@MF, KS tends to increase as the volume fraction of the MF shell decreases (Figure 2d). It is noteworthy that this trend is the same as that observed in the HC changes (Figure 2b). Similarly, KS of CF@ MF0.09, [email protected], [email protected], and [email protected] (Figure 3d) also exhibits the same trend as observed in their HC changes (Figure 3b). All of these results indicate the significant effects of KS on HC at the ESC regime. The KS dependency on different shell compositions could be related to the unpaired d-electrons in the octahedral lattice sites of XFe2O4 (X = Mn, Fe, Co, and Ni), which contribute to the total numbers of disoriented spins on the surface,10 acting as another important factor of the ESC effect. It should be noted that Kmag and Kshape are highly likely minor contributions to HC, if any, as compositional changes of the hard core and soft shell do not follow the expected trend of Kmag, and all tested nanoparticles have identical Kshape with the same shape and aspect ratio. The magnetic behavior of core−shell nanoparticles in the ESC regime are further supported by numerical simulation based on the standard Landau−Lifshitz−Gilbert (LLG) theory using the object-oriented micromagnetic framework (OOMMF, National Institute of Standard and Technology, NIST).21 Crosssectioned views of nanoparticles (Figure 3e) display the magnetic 802

DOI: 10.1021/acs.nanolett.6b04016 Nano Lett. 2017, 17, 800−804

Letter

Nano Letters

Figure 3. Compositional effects of ultrathin shell on magnetic properties and theoretical display of the magnetic spin inversion process. (a) Hysteresis loops at 5 K for the CF core and CF@XFe2O4 (X = Mn, Fe, Co, and Ni) core−shell nanoparticles with a shell volume fraction of 0.09 and (b) their HC and MS values. (c) Literature values of Kmag for various metal ferrites.19 (d) Calculated values of KS for the CF core and various CF@ XFe2O4 (X = Mn, Fe, Co, and Ni). (e) Simulated magnetic spin states of CF core, [email protected], and [email protected] nanoparticles using the OOMMF program. The color map displays the angle of spin deviation over the B0. Initially, the magnetic spins of nanoparticles are set to be aligned toward the +180° direction (red color). The B0 values required for complete magnetic spin inversion (−180°, blue color) of (i) CF core, (ii) [email protected], and (iii) [email protected] nanoparticles are estimated to be 10.7, 21.5, and 7.0 kOe, respectively.

example, the specific loss power (SLP), which is the amount of thermal energy generated from magnetic nanoparticles under the application of an alternative magnetic field (AMF), is examined. The SLP values of various nanoparticles, including ultrathin-shelled core−shell nanoparticles, are measured from the heating profile tested in toluene under an AMF of 30 kA/m at 500 kHz (Figure 4b,c; see SI section 6). [email protected] core− shell nanoparticles display the highest SLP value of 10810 W/g, which is approximately 1.7 times higher than the value of 6230 W/g for the CF core. This value is among the largest SLP values observed in nanoparticle systems and is also an order of magnitude higher than that of previously reported exchangecoupled nanoparticles with hard−soft exchange coupling.14,24 In this study, we investigated the effects of shell thickness on the modulated magnetism of the core−shell nanoparticles. Conventional hard−soft exchange coupling is inapplicable for explaining the unexpected large increase of coercivity for thinshelled core−shell nanoparticles. Instead, surface anisotropy plays a significant role here to shift overall magnetic anisotropy values and related magnetic properties. By taking advantage of such magnetic features, enhanced energy product and magnetic heating efficiency are demonstrated. In addition to existing EC strategies,3 such ultrathin-shelled nanoparticles can provide new opportunities for further tailoring nanoscale magnetism.

Figure 4. (BH)max and SLP of various CF@MF core−shell nanoparticles. (a) (i) Second quadrant B−H curves and (ii) BH product curves for the CF core, [email protected], and [email protected]. (b) The measured magnetic heating profile as a function of time for CF core and various CF@MF nanoparticles under an AMF of 30 kA/m at 500 kHz, and (c) the resulting SLP values.

■

ASSOCIATED CONTENT

S Supporting Information *

is acquired for the [email protected] core−shell nanoparticles, is 4.6 times higher than the 390 J/m3 of the CF core, and exceeds the theoretical limit of 1417 J/m3 of [email protected]. As another

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.6b04016. 803

DOI: 10.1021/acs.nanolett.6b04016 Nano Lett. 2017, 17, 800−804

Letter

Nano Letters

■

(23) Sakuma, N.; Ohshima, T.; Shoji, T.; Suzuki, Y.; Sato, R.; Wachi, A.; Kato, A.; Kawai, Y.; Manabe, A.; Teranishi, T. ACS Nano 2011, 5, 2806. (24) Zhang, Q.; Castellanos-Rubio, I.; Munshi, R.; Orue, I.; Pelaz, B.; Gries, K. I.; Parak, W. J.; del Pino, P.; Pralle, A. Chem. Mater. 2015, 27, 7380.

Detailed synthetic procedures, TEM images, ICP-AES results, XPS results, SQUID data, spin structure simulation, and calculation methods (PDF)

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Jinwoo Cheon: 0000-0001-8948-5929 Author Contributions

S.H.M. and S.-h.N. contributed equally to this work. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was financially supported by the Institute for Basic Science (IBS-R026-D1).

■

REFERENCES

(1) Liu, F.; Hou, Y.; Gao, S. Chem. Soc. Rev. 2014, 43, 8098. (2) Lopez-Ortega, A.; Estrader, M.; Salazar-Alvarez, G.; Roca, A. G.; Nogues, J. Phys. Rep. 2015, 553, 1. (3) Zeng, H.; Li, J.; Liu, J. P.; Wang, Z. L.; Sun, S. H. Nature 2002, 420, 395. (4) Sun, X. L.; Huls, N. F.; Sigdel, A.; Sun, S. H. Nano Lett. 2012, 12, 246. (5) Song, Q.; Zhang, Z. J. J. Am. Chem. Soc. 2012, 134, 10182. (6) Teng, X. W.; Yang, H. Nanotechnology 2005, 16, S554. (7) Redl, F. X.; Black, C. T.; Papaefthymiou, G. C.; Sandstrom, R. L.; Yin, M.; Zeng, H.; Murray, C. B.; O’Brien, S. P. J. Am. Chem. Soc. 2004, 126, 14583. (8) Del Blanco, L.; Fiorani, D.; Testa, A. M.; Bonetti, E. J. Magn. Magn. Mater. 2005, 290, 102. (9) Soares, J. M.; Galdino, V. B.; Machado, F. L. A. J. Magn. Magn. Mater. 2014, 350, 69. (10) Cornell, R. M.; Schwertmann, U. The Iron Oxides: Structure, Properties, Reactions, Occurrences and Uses; Wiley, Weinheim, 2006. (11) Santos-Carballal, D.; Roldan, A.; Grau-Crespo, R.; de Leeuw, N. H. Phys. Chem. Chem. Phys. 2014, 16, 21082. (12) Ong, Q. K.; Lin, X. M.; Wei, A. J. Phys. Chem. C 2011, 115, 2665. (13) Lee, J.-H.; Jang, J. T.; Choi, J. S.; Moon, S. H.; Noh, S.-h.; Kim, J. W.; Kim, J. G.; Kim, I. S.; Park, K. I.; Cheon, J. Nat. Nanotechnol. 2011, 6, 418. (14) Noh, S.-h.; Na, W.; Jang, J. T.; Lee, J.-H.; Lee, E. J.; Moon, S. H.; Lim, Y.; Shin, J. S.; Cheon, J. Nano Lett. 2012, 12, 3716. (15) Carter, C. B.; Norton, M. G. Ceramic Materials: Science and Engineering; Springer, New York, 2007. (16) Tanaka, T.; Matsuzaki, J.; Kurisu, H.; Yamamoto, S. J. Magn. Magn. Mater. 2008, 320, 3100. (17) Chen, R.; Christiansen, M. G.; Anikeeva, P. ACS Nano 2013, 7, 8990. (18) Peddis, D.; Mansilla, M. V.; Morup, S.; Cannas, C.; Musinu, A.; Piccaluga, G.; D’Orazio, F.; Lucari, F.; Fiorani, D. J. Phys. Chem. B 2008, 112, 8507. (19) Coey, J. M. D. Magnetism and Magnetic Materials; Cambridge University Press, New York, 2010. (20) Yanes, R.; Chubykalo-Fesenko, O.; Kachkachi, H.; Garanin, D. A.; Evans, R.; Chantrell, R. W. Phys. Rev. B: Condens. Matter Mater. Phys. 2007, 76, 064416. (21) Zhang, S.; Li, Z. Phys. Rev. B: Condens. Matter Mater. Phys. 2001, 65, 054406. (22) Lopez-Ortega, A.; Lottini, E.; Fernandez, C. D.; Sangregorio, C. Chem. Mater. 2015, 27, 4048. 804

DOI: 10.1021/acs.nanolett.6b04016 Nano Lett. 2017, 17, 800−804