Superparamagnetic and Perfect-Paramagnetic Zinc Ferrite Quantum Dots from Microwave-Assisted Tunable Synthesis.

The properties of quantum dot (QD)-size material depend directly upon its unit cell structure. Spinel zinc ferrite QD powder is produced via a one-pot microwave-assisted hydrothermal synthesis for just 5 min. Varying initial pH values of the preparation sol from 6 to 12 enlarges the Zn/Fe atomic ratio (by ca. 10%), unit cell volume (by ca. 0.5%), particle size (3.5-4.5 nm), and degree of inversion. This leads to a change in the magnetic behavior of the QD-size zinc ferrite from a superparamagnetic to a perfect-paramagnetic type. This novel finding points that the significant changes in the inherent structural parameters of spinel ZnFe2O4 QDs (Zn/Fe ratio and degree of inversion) induced by the systematic pH change of the preparation sol are exclusively responsible for the observed unique magnetic behavior instead of mere QD (single domain) nanosizes.

Introduction

Zinc ferrite (ZnFe2O4) is used in multiple forms in various fields of photocatalysis, spintronics, sensors, energy and information storage, etc.[1−10] The performance of this material in these applications depends critically upon an interplay of a number of intrinsic (surface/bulk defects, valence, vacancies, Zn/Fe ratio, and inversion parameter) and extrinsic (morphology and size/shape) properties.[2−10] In particular, it is always interesting to prepare a nanosize entity of this face-centered cubic (FCC) material through safe, simple, rapid, and cost-effective synthetic approaches. More importantly, these approaches must be amenable to controlling one or two of the intrinsic variables mentioned above. One preferable synthesis approach with such attributes is direct dielectric heating of a liquid dispersion, known as a microwave-assisted hydrothermal method.[11−13] This method has gained attention not only due to its simplicity and short synthesis time but also because it could bring up reproducible synchronous tailoring of the intrinsic variables quickly via the specific role of synthesis parameters (salt, concentration, solvent, time, temperature, binder, and nature of pH of the adjusting agent).[11−13]

Herein, we report a rapid (ca. 5 min) synthesis of superparamagnetic (SP) cum perfect-paramagnetic (PP) ZnFe2O4 powder nanomaterial of a quantum dot (QD) size, utilizing a simple sol dispersion whose initial pH (as the single parameter) was precisely controlled using a non-complexing NaOH solution. A number of powder samples were synthesized under a wide range of pH environments (2 to 12), while strictly maintaining all other synthesis parameters to be invariant.

Results and Discussion

The first direct eye observation was unique and with differentiable colors (e.g., brown and/or tan, etc.) of the individual powder samples, which are clearly noticed in optical photographs in Figure S1. Such differentiated colors of the powder materials could be associated with their distinct inherent properties. Remarkably, using the current recipe of sol dispersion, the impurity-free QD-size ZnFe2O4 structure was formed only when the initial pH value of the sol dispersion was set above 5 (see XRD patterns of all samples in Figure S2 and structural data in Table S1). Since no binder or additive is used in the synthesis, this unwanted phase formation at a pH of less than 6 is perhaps caused by a solubility difference of metal hydroxo complexes at a specific pH and reaction temperature, which leads to the main byproduct hematite (Fe2O3) and other stable oxide impurities.[14−18] Without paying much attention to the detailed mechanism of the phase formation, here, we focus on the comparative room-temperature magnetic properties of spinel ZnFe2O4 samples with two QD sizes synthesized under mildly acidic (pH = 6) and strongly basic (pH = 12) conditions.

Powder X-ray diffraction (XRD) patterns of these ZnFe2O4 QD samples are displayed in Figure . Both samples show broad symmetric XRD patterns with all peaks directly indexed to the face-centered cubic (FCC) ZnFe2O4 crystal structure (space group Fd3̅m/227), indicating that they are single-phase nanocrystalline materials.[4−6,9] However, a unique effect of different pH environments is revealed in a 2θ shift of ca. 0.35° in all XRD peaks, indicating a ca. 0.5% larger lattice constant (LFCC) for the sample synthesized at pH = 12 than that of the sample prepared at pH = 6.[19] The crystallite size and unit cell parameters of synthesized samples were estimated by applying Scherrer’s formula on the (311) XRD peak and the equation L = d(h2 + k2 + l2)1/2, where d = the interspacing of two nearest hkl planes, which is applicable to a cubic symmetry. The LFCC is calculated by averaging values of all major peaks that appeared in the XRD pattern.

Figure 1

XRD (indexed) patterns and TEM images (insets; indexed diffraction ring patterns) of spinel ZnFe2O4 QD powder samples as synthesized within 5 min of microwave irradiation at 165–170 °C under different initial pH environments. The curves are vertically offset by a constant factor. The estimated QD sizes (in parentheses) by XRD are also indicated.

As shown in Table , the average crystallite sizes estimated from XRD (3.5 and 4.5 nm for samples prepared at a pH of 6 and 12, respectively) are consistent with the mean particle sizes (3.5 and 4.8 nm, respectively), those obtained from transmission electron microscopy (TEM) images (Figure and Figure S3). A narrow and homogeneous particle size distribution with a minimal significant variation in the spheroid shape is evident in TEM images in both of the pristine samples (Figure S3). The brightness of the selected area electron diffraction (ED) ring pattern in insets of microscopic images and a high-resolution TEM image also support the crystalline nature of the powder samples (Figure S3). The optical band gaps (Figure S4; UV–vis spectra) of powder samples, namely, ca. 2.15 eV (3.5 nm) and ca. 2.19 eV (4.5 nm), are much higher than those typically reported (1.8–2.0 eV) for much larger sizes of ZnFe2O4 materials.[3,30] All the above results clearly indicate that the pH of the synthesis sol dictates the structural properties of the as-synthesized ZnFe2O4 QD powder samples.

Table 1

Inherent Properties, Raman Modes [A1g and F2g(2)], and Magnetic Data (at 300 K) of Quantum Dot (QD)-Size Spinel ZnFe2O4 Samples

sample	average crystallite/particle sizea (nm)	LFCC (VFCC) (Å; Å3)	binding energy (eV) of Zn (Fe) [O]	element (%) Zn (Fe) [O]	Zn/Fe ratiob	Raman (cm–1) A1g [F2g(2)]	Ms (Hc)c (emu/g, Oe)
pH = 6	3.5 (3.5)	8.41902 (596.7)	1021.34 (711.34) [529.94]	15.1 (15.0) [69.9]	1.006 (0.49)	584 [635]	19.5 (8.9)
pH = 12	4.5 (4.8)	8.45452 (604.3)	1021.46 (711.46) [529.96]	16.9 (15.3) [67.8]	1.101 (0.59)	381 [333]	4.7 (6.7)

By XRD (TEM in parenthesis). VFCC = unit cell volume.

By XPS (ICP-MS in parenthesis).

Ms (saturation magnetization at 100 kOe) and Hc (coercive force).

Figure 2

(i) High-resolution XPS spectra of Zn and (ii) superparamagnetic (sigmoidal; Zn/Fe ratio ≈ 0.49) and perfect paramagnetic (linear; Zn/Fe ratio ≈ 0.59) profiles (at 300 °K) of QD-size (parentheses) spinel ZnFe2O4 samples (numerical data in Table ).

The high-resolution X-ray photoelectron spectroscopy (XPS) Zn spectra are displayed in Figure i (Figure S5 for Fe and O spectra). The binding energy (BE) was calibrated with respect to the reference C 1s peak (284.6 eV) of adventitious carbon in the sample. The BE analysis of XPS peaks indicated that both samples consisted of Zn2+, Fe3+, and O2– ions, and the BEs of cations are very close to that of a standard FCC spinel zinc ferrite structure (Zn ≈ 1021.40 eV; Zn/Fe ratio = 0.5, degree of inversion = 0).[4,5,9] However, the pH = 12 sample (4.5 nm) has a ca. 10% higher Zn/Fe atomic ratio and ca. 0.12 eV higher BE values for Zn2+ and Fe3+ cations relative to its counterpart pH = 6 sample (3.5 nm). This extra Zn/Fe ratio is consistent with the result of a ca. 0.5% larger LFCC caused by the larger ionic size of Zn2+ (0.74 pm) than that of Fe3+ (0.63 pm),[2,19,20] but the Zn/Fe ratios determined by XPS are much larger than the ideal stoichiometry of ZnFe2O4 (Zn/Fe = 0.5). However, the overall Zn/Fe ratios determined using an inductively coupled plasma mass spectrometry (ICP-MS) technique (Table and Figure S5) are close to the stoichiometry. Still, the ICP-MS results also confirm a ca. 10% higher Zn/Fe atomic ratio for the pH = 12 sample (0.59) than for the pH = 6 sample (0.49), which is consistent with the XPS results. Thus, the unexpectedly high Zn/Fe ratios from the XPS analysis are caused by the surface-sensitive XPS technique and indicate that Zn ions prefer to reside near the surface of QD-size particles.[21−23] This segregation of a particular ion to the surface is known as the self-purification effect, which is common in ultrasmall structures due to their small interior volume.[21−23]

Since the BE of ionic species depends on the bonding strength and surrounding environment, a significant change in the BE indicates that the initial pH of the preparation sol plays a pivotal role that forces some Zn2+ and Fe3+ to change their occupancy preference between two interlattice subsites (tetrahedral ↔ octahedral) in the synthesized ZnFe2O4 QD samples. Due to a non-stoichiometric Zn/Fe ratio, the degree of inversion might be greater than 0 in either of the QD-size powder samples.[6,19,20,24,25] This difference in the inversion state is further verified by measuring the static magnetic response. Thus, Figure ii shows the room-temperature static magnetization profiles of ZnFe2O4 QDs, which demonstrates an immense difference in the magnetic behavior of the two QD size samples (Table ). The 3.5 nm ZnFe2O4 (nearly stoichiometric Zn/Fe ratio = 0.49 by ICP-MS) reveals a sigmoidal superparamagnetic (SP) profile, whereas the 4.5 nm QD sample (non-stoichiometric Zn/Fe ratio = 0.59 by ICP-MS) shows a linear perfect-paramagnetism (PP) behavior. This observation is surprising as the SP particle (3.5 nm) has a smaller (by ca. 0.5%) unit cell size/volume than that of the PP sample (4.5 nm). Therefore, for comparison purpose, any of the minor magnetic contributions (coercivity, magnetization, and magnetic profile) arising from extrinsic variables (size, shape, surface area, volume, and rotation of a single domain; domain boundary effects; and canting and anisotropy of surfaces) and intrinsic unit cell factors (crystallinity, directional order, and LFCC or VFCC) could be safely neglected at first sight in both spinel ZnFe2O4 QD samples.[10,19,20,24−26]

A striking difference in the relative degree of inversion between the two pristine ZnFe2O4 QD powder samples is clearly visible in the high-resolution A1g and Fe2g(2) Raman modes in Figure (data in Table ). The lower (584 cm–1) and higher (635 cm–1) wavenumber positions of the A1g peak (tetrahedral group) in 3.5 and 4.5 nm-size samples resemble higher and lower numbers of tetrahedron ZnO4 units, respectively, thus reflecting a lesser inversion state in the 3.5 nm ZnFe2O4 (synthesized at pH = 6) sample than that of 4.5 nm ZnFe2O4 (synthesized at pH = 12).[6,27−30] Also, the F2g(2) (octahedral group) peak position of the 3.5 nm sample (381.6 cm–1) is higher than that of the 4.5 nm sample (333.2 cm–1); therefore, this higher side shift is cognate with the lower number of octahedron FeO6 units, which again indicates a lesser inversion parameter of the 3.5 nm sample than that of the 4.5 nm sample.[27−30]

Figure 3

High-resolution Raman spectra of superparamagnetic (3.5 nm; pH = 6) and perfect-paramagnetic (4.5 nm; pH = 12) spinel ZnFe2O4 QD samples (see Figure S6 and Table S2 for all Raman active modes).

Figure S7i shows the temperature-dependent magnetic profiles of almost identical QD-sized particles where a clear drastic difference is easily observed between the magnetization behavior and thermal fluctuations of particles. The absence of particle interactions (negligible coercivity) suggests that the dissimilar structural parameters (Zn/Fe ratio, inversion state, and lattice parameter) together with uncompensated surface spins (paramagnetic type) of QD-size particles ultimately control the magnetization of powder samples.[8,19,31−33] As a result, at a constant low temperature (e.g., 50 °K; Figure S7ii), the effect of the dissimilar core (data in Table and Table S3) becomes more pronounced in a near-zero applied magnetic field, and a sharp ferrimagnetism (FiM) response is observed in a nearly stoichiometric (3.5 nm) sample as opposed to an extremely poor FiM in a Zn-enriched (4.5 nm) sample. The higher inversion and excessive nonmagnetic Zn causes a magnetic frustration (disorder) in a non-stoichiometric (4.5 nm) sample by disrupting the magnetic coupling of lattice-site interactions that eventually leads to perfect paramagnetism at room temperature.[19] Hence, the analyses (XRD, XPS, ICP-MS, TEM, and Raman) consistently confirm that the single sol parameter (pH) is sufficient to control the inherent structural parameters (phase purity, particle size, and its narrow size distribution, Zn/Fe ratio, LFCC, and degree of inversion) of the spinel-type ZnFe2O4 nanomaterial.

Additionally, the degree of inversion in this powder material is accompanied by the unique SP and/or PP evolution. To further confirm this, an inversion was boosted in the SP ZnFe2O4 sample (3.5 nm) by heating it at an elevated temperature (3 °C/min, 460 °C, 1 h, air) inside a laboratory furnace.[29] As shown in Figure S8 (data in Table S3), although this heating promoted better crystallinity and formed a larger particle (3.5–7.2 nm), it also drastically reduced the SP profile and the saturation magnetization (19.5 to 9.1 emu/g), which is expected owing to its relatively higher degree of inversion rather than its bigger QD size (characteristic length scale). This result also leads to the conclusion in that one intrinsic unit cell parameter (Zn/Fe ratio and/or a resulted degree of inversion) of each pristine QD size particle (3.5 or 4.5 nm) is exclusively responsible for the unique magnetic behavior (SP and/or PP) of spinel ZnFe2O4 QDs, not merely their nanosizes.

Conclusions

In summary, for the first time, to the best of our knowledge, a dual magnetic behavior (superparamagnetism and/or perfect paramagnetism) is observed in a QD-size spinel ZnFe2O4 nanomaterials, which is synthesized via a rapid (5 min) and facile microwave-assisted hydrothermal synthesis process under varying pH values of the preparation sol. It is interesting that the flipped magnetism is achieved by simply varying the initial pH of the sol dispersion, which modifies the Zn/Fe ratio and inversion parameter of the QD-size ZnFe2O4 powder. Thus, the observed drastic transition of the unique magnetic behavior and the magnetic properties is due to significant changes in the intrinsic structural parameters of spinel ZnFe2O4 (Zn/Fe ratio and inversion parameter), instead of merely QD (single domain) sizes.

Experimental Section

As-purchased reagent-grade chemicals (Samchun Chemical Co., Korea) were used. To prepare the ZnFe2O4 QD powder nanomaterial, a simple sol dispersion was made by mixing appropriate amounts of Zn/Fe nitrate hexahydrates (1–2 mole ratio) in a 60 mL triple-deionized water (Millipore, resistivity of ∼18 MΩ–1) while closely controlling its pH by dropwise addition of a non-complexing NaOH solution under constant mild magnetic stirring (ca. 20 min). The sol was then kept inside a sealed digestive Teflon vessel (ca. 100 mL) for only 5 min at 165–170 °C (ramp of 10 °C/min.) under microwave irradiation (MDS-2000, 2.45 GHz, 1200 W, CEM Corporation, USA). The vessel was cooled naturally, and the powder samples (weight yield of ca. 85%) were obtained after excessively (three times) washing the filtrates with distilled water and then overnight vacuum drying in a laboratory oven.

The optical images of powder samples were captured using an ordinary digital camera (Sony). The phase, particle, structural composition, optical band-gap characteristics, and magnetic profiles/properties of samples were determined using X-ray diffraction (XRD, Cu Kα, PRO-MPD, Philips), transmission electron microscopy (TEM, Cs-corrected JOEL, JEM-2200FS), X-ray photoemission spectroscopy (XPS, HR-XPS; VG Scientific ESCA LAB 220-IXL), ICP-MS system (Agilent, USA), Raman (alpha300R WITec, excitation of ∼532 nm), diffuse reflectance spectra (300–800 nm, integrated sphere method, Shimadzu), a vibrating sample magnetometer (VSM, Lake Shore), and Quantum Design PPMS VersaLab.