Investigation of novel inverted NiO@NixCo1-xO core-shell nanoparticles

Inverse core-shell nanoparticles, comprised of an antiferromagnetic (AFM) core covered by a ferromagnetic (FM) or ferrimagnetic (FiM) shell, are of current interest due to their different potential application and due to the tunability of their magnetic properties. The antiferromagnetic nature of NiO and high Neel temperature (523 K) makes this material well suited for inverse core-shell nanoparticle applications. Our primary objective in this project has been to synthesize and characterize inverted core-shell nanoparticles (CSNs) comprised of a NiO (AFM) core and a shell consisting of a NixCo1-xO (FiM) compound. The synthesis of the CSNs was made using a two-step process. The NiO nanoparticles were synthesized using a chemical reaction method. Subsequently, the NiO nanoparticles were used to grow the NiO@NixCo1-xO CSNs using our hydrothermal nano-phase epitaxy method. XRD structural characterization shows that the NiO@NixCo1-xO CSNs have the rock salt cubic crystal structure. SEM-EDS data indicates the presence of Co in the CSNs. Magnetic measurements show that the CSNs exhibit AFM/FiM characteristics with a small coercivity field of 30 Oe at 5 K. The field cooled vs zero field cooled hysteresis loop measurements show a magnetization axis shift which is attributed to the exchange bias effect between the AFM NiO core and an FiM NixCo1-xO shell of the CSNs. Our ab initio based calculations of the NixCo1-xO rock salt structure confirm a weak FiM character and a charge transfer insulator property of the compound.Inverse core-shell nanoparticles, comprised of an antiferromagnetic (AFM) core covered by a ferromagnetic (FM) or ferrimagnetic (FiM) shell, are of current interest due to their different potential application and due to the tunability of their magnetic properties. The antiferromagnetic nature of NiO and high Neel temperature (523 K) makes this material well suited for inverse core-shell nanoparticle applications. Our primary objective in this project has been to synthesize and characterize inverted core-shell nanoparticles (CSNs) comprised of a NiO (AFM) core and a shell consisting of a NixCo1-xO (FiM) compound. The synthesis of the CSNs was made using a two-step process. The NiO nanoparticles were synthesized using a chemical reaction method. Subsequently, the NiO nanoparticles were used to grow the NiO@NixCo1-xO CSNs using our hydrothermal nano-phase epitaxy method. XRD structural characterization shows that the NiO@NixCo1-xO CSNs have the rock salt cubic crystal structure. SEM-EDS data indicates the p...


I. INTRODUCTION
Bimagnetic core-shell nanoparticles (CSNs) are of considerable interest due to their potential applications in magnetic spin valves, spintronics, magnetic random access memory, hyperthermia, MRI imaging, drug delivery and in other areas. [1][2][3][4] The magnetic properties of bimagnetic CSNs (without being embedded in a surrounding matrix) are tuned by adjustment of overall size, core vs shell size, core vs shell composition, and morphology. Core-shell nanoparticles are typically fabricated in a conventional configuration, having a ferromagnetic (FM) core and an antiferromagnetic (AFM) or ferrimagnetic (FiM) shell, or in an inverted configuration, having an AFM core and a FM or FiM shell. Due to their highly ordered AFM cores, the inverted bimagnetic CSNs have highly tunable coercivities, blocking temperatures, and other magnetic properties that are highly promising for device and medicinal applications. There has been particular attention paid to the exchange bias effect in bimagnetic CSNs, whereby a magnetic interaction between the core and shell has a direct bearing on the overall magnetic properties of the nanostructures. 5,6 By suitable tuning of the CSNs in terms of composition, size and morphology characteristics alluded to above, the exchange bias properties can potentially be exploited for various magnetic device and medicinal applications.
We have recently investigated a number of Cr 2 O 3 @M x Cr 2-x O 3-δ CSN systems and found these to have enhanced magnetic properties that show promise for potential magnetic applications. [7][8][9] However, the exchange bias effect in these CSNs has been observed to be considerably reduced at room temperature due to the Néel temperature of the AFM core occurring just above the RT value (T N = 308 K). Bulk NiO generally has AFM ordering with a Néel temperature considerably above room temperature (T N = 523 K), which makes it well suited for bimagnetic core-shell nanostructure applications. Remarkably, to our knowledge, with the exception of a very small number of other systems, the Ni@NiO CSNs appear to be the overwhelmingly predominant bimagnetic NiO-based core-shell nanostructure system that has been investigated to date. [10][11][12][13][14][15] The synthesis of Ni@NiO CSNs typically involves synthesis of the Ni nanoparticles followed by their oxidation, thus producing a NiO shell over the Ni core. These CSNs have been shown to exhibit high coercivities at room temperature 15 and, for ones coated with graphitic carbon, superparamagnetic blocking temperatures at above room temperature value. 13 In both cases, these effects were attributed to substantial exchange bias effects occurring at the interface between the FM Ni core and the AFM NiO shell. Recently, Skoropata et al. studied a series of iron oxide based core-shell nanostructures, including γ-Fe 2 O 3 @NiO CSNs. 16 Their conclusion was that the magnetic properties of the γ-Fe 2 O 3 @NiO CSNs were consistent with the formation of an inter-diffused Ni-doped Fe 2 O 3 layer at the core-shell interface, resulting in a trilayer nanostructure. Although most likely not bimagnetic, it has been reported that SiO 2 @NiO CSNs exhibit greater remanent and saturation magnetization values, but lower coercivities, than NiO nanoparticles. 17 Ponnusamy et al. have recently synthesized Co-doped NiO nanoparticles using a chemical reaction method. 18 The authors report that their magnetic measurements reveal ferromagnetic properties in the Co x Ni 1-x O nanoparticles. In this paper, we report on the synthesis and characterization of inverted NiO@Ni x Co 1-x O CSNs, where the core-shell nanostructure was made using our hydrothermal nanophase epitaxy method. [7][8][9]

II. METHODS
The synthesis of the CSNs was made as follows: NiO nanoparticles were first synthesized according to procedures outlined by El-Kemary et. al. 19 The nanoparticle synthesis was made using thermal decomposition of Ni(OH) 2 . A precursor was first prepared using nickel chloride hexahydrate (0.111 M) dissolved in ethanol which was then added to 6.73 ml of hydrazine monohydrate at a solution molar ratio of 5. KOH was used to adjust the pH to about 12 and the solution was stirred for ∼2 hours at 22 • C. Upon completion of stirring, the reaction product was rinsed with deionized water for removal of residues and then centrifuged in tolune. The Ni(OH) 2 .0.5H 2 O nanoparticles were dried in atmosphere. Finally, the nickel hydroxide nanoparticles were converted to NiO nanoparticles through thermal decomposition at 600 • C for ∼2 hours.
The NiO nanoparticles were next used as cores to synthesize the NiO@Ni x Co 1-x O CSNs using a hydrothermal method described previously. 20 First, HPLC grade water (pH = 7) was purged with N 2 gas for 15 minutes at 70 • C. Next, 0.237 g CoCl 2 .6H 2 O were added to the deoxygenated water to make a 0.05 M solution (pH = 5.6) and sonicated for ∼30 minutes. Several drops of HCl were added to reduce the pH of the solution to a value of 3.2. Subsequently, 0.33g of NiO nanoparticles was added to the solution, which was then sonicated for ∼30 minutes. Subsequently, the solution was inserted inside an autoclave and hydrothermal treated at ∼190 • C for 20 hours. The resulting NiO@Ni x Co 1-x O CSNs were then rinsed, centrifuged in DI water and subsequently dried in atmosphere.
The SEM imaging characterization was made using a JEOL SEM instrument operating at 2 kV with WD of 5 mm. The samples were mounted on carbon tape for SEM imaging. Preliminary elemental analysis was made using a SEM-EDS (Oxford Instruments) and a field emission gun. XRD data were measured from the nanopowder samles using a Bruker D8 Discover diffractometer operating at 40 kV and 40 mA. The diffractometer is operated using a Cu tube x-ray source (Cu Kα, λ= 1.54184 Å), a Göbel mirror and a 0.6 mm slit on the incident x-ray beam. A Linxeye 1-D Si strip detector was used for measurement of the XRD data. The XRD data were analyzed using Rietveld refinement, which was accomplished using the Bruker TOPAS software. The modelling of the background in the XRD patterns was made using a 5-th order Chebychev polynomial.
The magnetization data were measured from the samples using a SQUID MPMS/XL magnetometer (Quantum Design) and a PPMS-VSM instrument at CCMR, Cornell University (Quantum Design). The nanoparticle samples were packed in a gelatin capsule and then inserted inside the dewar of the SQUID magnetometer. The magnetization data were measured in the 5 -300 K temperature (T) range. An applied field of 500 Oe was used to make the magnetization vs T measurements in the field-cooled (FC) case. The magnetic hysteresis curves were measured at 5 K in the -10,000 to 10,000 Oe applied field range. The samples were cooled at 20,000 Oe in the case of FC hysteresis loop measurements whereas 0 field was applied in the case of zero-field-cooled (ZFC) measurements. The TEM analysis was made using a Titan 80-300 instrument for which the field emission gun was operated at 300 keV. The TEM samples were prepared on lacey carbon grids.
The first principles calculations were made using the local spin density approximation method (LSDA) implemented in Quantum Espresso. 21 A rhombohedral primitive unit cell of NiO with a space group of Fm3m (COD 4329325) was used for the self-consistent field calculations. The Ni 0.88 Co 0.12 O structure was constructed by replacement of one of the Ni (out 8 cation sites) with Co in the 1 x 2 x 2 supercell. Calculation convergence was achieved with a net antiferromagnetic configuration (+-+-) for the NiO structure. The Brillouin zone integral 6 x 6 x 6 point grid was used based on the Monkhorst-Pack scheme. 22 The cutoff energy for the plane waves was 150 eV. Structural relaxation (volume) was accomplished for all of the structures and the relaxed structural parameters were used in the final calculation. Density of states (DOS) and the projected density of state (pDOS) were obtained by using the tetrahedron method with an 8 x 8 x 8 k-point grid mesh. Perdew-Burke-Ernzerhof exchange-correlational norm conserving pseudopotentials were used for all the atoms. 23 Additionally, in order to describe the electron-electron correlation characteristics of local electrons in transition metal atoms, we have performed the GGA+U type calculation where U is the on-site Coulomb energy, U and the exchange interaction (J) were used. Fig. 1(a) shows an SEM image of synthesized CSNs. A Gaussian fit of the histogram plot of particle size distribution shown in Fig. 1(b), as determined from TEM imaging, gives an average particle size of 29.7(2) nm for the NiO@Ni x Co 1-x O CSNs. This is in very good agreement with our XRD Scherrer equation analysis yielding a CSN size of 29.39(6) nm. The SEM-EDS data shown in Fig. 1(c) show that Co, along with Ni and O, are present in the sample. Fig. 1(d) shows a highresolution TEM (HRTEM) image of an isolated CSN. As shown in the TEM image, the core and shell are separated by an interface which is slightly more structurally disordered than either the core or the shell of the isolated CSN. The SEM and TEM imaging indicate presence of faceted as well as pseudospherically shaped CSNs. The shell region of the nanoparticle shown in Fig. 1(d) is not completely uniform as the core NiO nanoparticles are pseudospherical in shape. The approximate thickness of the shell, estimated from the TEM images, is ∼2 nm. The shell and core regions of the CSNs, respectively, have identical crystallographic symmetry, as evidenced from our TEM and XRD results (Figs. 1(d and e)), providing evidence of epitaxial growth of the shell. Fig. 1(e) shows the powder x-ray diffraction pattern measured from our CSNs at room temperature. XRD analysis indicates that the CSNs possess the FCC structure with space group Fm3m with a small contribution from a minor cubic phase having the F43m structure: The minor phase is consistent with nanocrystalline Co 3 O 4 . The values for the ionic radii of Ni 2+ and Co 2+ (high spin) are 0.69 and 0.745 Å for six fold coordination, respectively. 23,24 Thus, Co2+ can readily substitute for Ni 2+ ions in the rock salt structure of NiO. The higher intensity of the (200) peak (located at 43.18 • ) compared to the usual ratio of peak intensities for bulk NiO suggests somewhat oriented nanocrystalline structures. Rietveld refinement of the XRD data measured from the CSN sample was made in order to further analyze their structural properties. Two Fm3m structure CIF files were used in the refinement, one for the NiO core and another for the Ni x Co 1-x O shell, and a F43m structure CIF file for the impurity phase. The lattice parameters and the unit cell volume were found to be increased slightly for the Ni x Co 1-x O shell due to Co doping in the shell region, which is consistent with the ionic radii difference between Ni 2+ and Co 2+ . Therefore, Co-incorporation results in a modification of the structure of the shell region of the CSNs, which can indirectly affect the magnetic properties of the Ni x Co 1-x O shell. The unusual magnetic behavior may be due to grain size reduction and breaking of large number of exchange bonds. The presence of small magnetic clusters on the surface and lattice imperfection increases the uncompensated spin values. 18 The magnetization (M) as a function of applied field (hysteresis loop) measured at 5 K is shown in Fig. 2(a). Our preliminary measurements clearly show that the magnetization of the CSNs shows a high retentivity value and does not reach a saturation value at 1000 Oe. The hysteresis shows a coercive field of about ∼528 Oe for the ZFC and ∼538 Oe for the FC case at 5 K. Ferromagnetism has been reported previously in Co doped NiO nanoparticles by Ponnusamy et al. 18 The hysteresis loops show a shift of the FC hysteresis loop relative to the ZFC loop along the negative H axis and positive M axis directions. This is consistent with an exchange bias between the core and shell regions of the CSNs. 3 The value of the exchange bias is ∼34 Oe between FC and ZFC curves based on the formula H e = (H ZFC+ -H FC+ -H FC-+ H ZFC-)/2 where the +/-indicate positive/negative H values when M = 0. 25 It is clear that introduction of Co 2+ ions for Ni 2+ ions in the NiO structure results in most likely FiM properties within the shell region of the CSNs. The magnetic data for our NiO and Co 3 O 4 NPs indicates that these exhibit AFM properties (Fig. 2(c) and (d)). Fig. 2(b) shows the ZFC and FC magnetization vs temperature data measured from our NiO@Ni x Co 1-x O CSNs. The FC magnetization shows a peak at ∼20 K followed by a rapid decrease with increasing temperature to ∼174 K and a very shallow reduction trend thereafter to 300 K. The ZFC magnetization curve increases sharply from 5 K, peaks at ∼20 K and then declines in a similar manner to 300 K as the FC curve but shows a shoulder near ∼90 K. The peak at ∼20 K is potentially a manifestation of a spin order/disorder transition in the CSNs. The shoulder occurring at ∼90 K may be associated with the superparamagnetic blocking temperature of the CSNs. The bifurcation point between the ZFC and FC magnetization curves is at ∼100 K.

III. RESULTS AND DISCUSSION
The supplementary material shows the calculated density of states (DOS) and the partial DOS

IV. CONCLUSIONS
We have used a two-step process, involving thermal decomposition and hydrothermal synthesis, to fabricate NiO@Ni x Co 1-x O CSNs. Our characterization confirms Co incorporation in the shell region of the CSNs. XRD and TEM analysis shows that both core and shell regions have the FCC structure. In addition, the TEM analysis confirms the core-shell structure and epitaxial growth of the shell region over the core region of the NiO@Ni x Co 1-x O CSNs. The magnetometry results are consistent with the presence of an AFM core and an FiM shell in the CSNs exhibiting an exchange bias effect between the two components as observed by a shift along the applied field and magnetization axes. Our DFT based ab-initio calculations confirm that Co introduction in NiO and formation of Ni x Co 1-x O results in weak FiM properties and in the formation of a Mott-Hubbard insulator.

SUPPLEMENTARY MATERIAL
See supplementary material for DOS-pDOS plots and a table of calculated atomic magnetic moments.