Ferromagnetism in Colloidal Mn2+-Doped ZnO Nanocrystals - The

Raymond and Beverly Sackler Faculty of Exact Sciences, Tel Aviv University, Tel ... transition metals with the hope to integrate these materials w...
1 downloads 0 Views 266KB Size
Download PDF
20232

J. Phys. Chem. B 2005, 109, 20232-20236

Ferromagnetism in Colloidal Mn2+-Doped ZnO Nanocrystals Tal Meron and Gil Markovich* School of Chemistry, Raymond and BeVerly Sackler Faculty of Exact Sciences, Tel AViV UniVersity, Tel AViV 69978, Israel ReceiVed: July 19, 2005

High-temperature hydrolysis of Zn(II) and Mn(II) alkoxides in a high boiling point solvent in the presence of surfactants was used to prepare surfactant-coated Zn1-xMnxO nanocrystals with average size of 5.5 nm and x ) 0.04 ( 0.03. The magnetic properties of the nanocrystals were measured both for isolated particles diluted in a hydrocarbon matrix and for a nanocrystal powder. Nanocrystals of manganese oxide and ZnO coated with manganese oxide were prepared for comparison to the Zn1-xMnxO nanocrystals. We find that the manganese ions primarily substitute zinc ions in the hexagonal ZnO lattice, and part of them are ferromagnetically coupled up to room temperature even in isolated noninteracting nanocrystals. The rest of the ions are magnetically disordered or uncoupled. Surprisingly, these small Zn1-xMnxO nanocrystals poses relatively large low-temperature magnetic coercivity and relatively high blocking temperature in the isolated form, which indicate large magnetic anisotropy. In the nanocrystal powder the coercive field decreased significantly. This study highlights the advantages of working with noninteracting single domain particles of these intriguing materials.

Introduction Dilute magnetic semiconductors (DMS) are in the focus of magnetic materials research due to the possibility of using such materials for spintronic devices. A large effort has been dedicated to study semiconductors such as GaAs1 and InAs2 doped with various transition metals with the hope to integrate these materials with conventional semiconductor technology. Unfortunately, such types of DMS have low Curie temperatures, below room temperature.3 Recently, room-temperature ferromagnetism was reported for Co2+-doped TiO24 and Mn2+-doped ZnO,5 as well as for other doped wide band gap semiconductors.6 These discoveries arose the expectation of researchers in this field that DMS-based spintronic materials operating at room temperature are within reach but also brought up doubts regarding the source of the detected ferromagnetism in some of these materials.7 Significant progress has been achieved in the past few years in the development of methods to produce ferromagnetic ZnO films doped with various metal ions. In particular, most of the wet chemical synthesis techniques used were variants of nonaqueous precipitation of the doped oxides at room temperature in the presence of capping agents to form nanocrystalline films.8,9 The development of wet chemical routes to colloidal nanocrystalline DMS materials is important from several aspects: the use of wet chemical techniques may offer better control over material’s composition and uniformity and consequently over its magnetic properties,10 compared with various physical deposition techniques, which often require very high temperatures accompanied by phase segregation or other undesired effects. Studies of this form of magnetism in small well-separated nanoparticles may offer a key to better understand the physical mechanism of ferromagnetism in this family of materials in addition to the prospects of using such magnetic nanostructures for nanoscale spintronic or magneto-optical * Corresponding author. E-mail: [email protected].

devices. So far, room-temperature ferromagnetism was observed only in bulk, thin films and aggregated nanocrystalline forms of these materials. In this work we have prepared organically capped colloidal ZnO nanocrystals doped with 4% Mn2+, on average, which exhibit ferromagnetic behavior up to room temperature. Isolated nanocrystal samples revealed high coercivity upon cooling. To prove that these results are due to the mixing of Mn ions within the ZnO lattice the magnetic properties of the Zn1-xMnxO nanocrystals were compared with reference samples of MnO (including Mn3O4) nanocrystals and with manganese oxide deposited on pregrown ZnO nanocrystals. Experimental Section The nanocrystal synthesis is based on a high-temperature hydrolytic process originally developed for BaTiO3 nanocrystal preparation by O’brien et al.11 and was recently used by us to produce CoFe2O4 nanocrystals.12 Colloidal nanocrystals of ZnO, ZnO doped with few percent of Mn2+, MnO, and MnO-coated ZnO were synthesized by similar procedures. In a typical ZnO nanocrystal synthesis 1 mmol of zinc tert-butoxide (Alfa-Aesar) was dissolved in 10 mL of dioctyl ether (99%, Sigma-Aldrich) in the presence of 1 mmol of oleic acid (99%, Alfa-Aesar) and heated. At 100 °C, 1 mL of concentrated hydrogen peroxide (30% in water, Merck) was injected into the mixture. The temperature of the reaction was immediately increased to 200 °C while flushing with nitrogen to remove water vapor from the system and kept at 200 °C for about 3 h under nitrogen. Mn2+ doping was achieved by addition of 0.1 mmol of manganese(II) methoxide (Alfa-Aesar) to the 1 mmol of zinc tert-butoxide precursor solution. After the solution was cooled to room temperature the nanocrystals were precipitated from the solution by centrifugation. The precipitate was dissolved in chloroform and centrifuged again to separate out aggregated nanocrystals, resulting in a transparent solution of ZnO nanocrystals or a clear orange-yellow solution of the Mn-doped nanocrystals. For the preparation of manganese oxide nano-

10.1021/jp0539775 CCC: $30.25 © 2005 American Chemical Society Published on Web 10/07/2005

Ferromagnetism in Mn2+-Doped ZnO Nanocrystals

J. Phys. Chem. B, Vol. 109, No. 43, 2005 20233

crystals only Mn-methoxide was used as a precursor, and for coating the ZnO nanocrystals with MnO, a Mn-methoxide solution (0.1 mmol dissolved in 1 mL of dioctyl ether) was injected at 100 °C, 5 min after the injection of the concentrated hydrogen peroxide. Powder samples for X-ray diffraction and magnetic measurements were prepared by precipitation of the nanocrystals from the chloroform by addition of excess ethanol and centrifugation. Isolated particle samples were produced by mixing a dilute nanocrystal chloroform solution with excess oleic acid and octadecane, running at 9000 rpm, 15 min, in a centrifuge to dispose of aggregates, and evaporating out the chloroform. The nanocrystals were characterized using transmission electron microscopy (TEM), including energy-dispersive X-ray spectroscopy (EDX) over individual nanocrystals to measure Mn/Zn atomic ratio distributions over the nanocrystals. X-ray diffraction (XRD) was used to determine the abundant crystal phases in the various samples as well as to estimate nanocrystallite size for the powder samples from the line widths. The magnetic properties of the nanocrystal samples were characterized using a Quantum Design MPMS XL5 SQUID magnetometer. Results and Discussion The Zn1-xMnxO nanocrystal data presented here correspond to one specific synthesis, but the results for several syntheses with small variations in average size and doping level were qualitatively similar. A somewhat different sample with larger particles was prepared using oleylamine as a cosurfactant (see Supporting Information). Figure 1a shows a TEM image of Mn2+-doped ZnO nanocrystals having an average dimension of 5.5 ( 1.3 nm. Highresolution TEM images of the as-prepared nanocrystals confirmed that the particles were single crystals. EDX measurements performed on individual nanocrystals in this sample provided Mn atom percentage out of the total Zn + Mn in the range of 0-28% Mn. A histogram of the EDX statistics over ∼40 particles is shown in Figure 1b but should be taken with caution as the average error of each individual particle measurement is also of the order of several percent. Almost half of the nanocrystals showed less than 1% doping level, while at higher doping levels the most probable atomic concentration was around 4-5%. The average concentration of manganese atoms measured by EDX over large groups of nanoparticles was 4 ( 3%, which is lower than the relative atomic concentration of the Mn(II) inserted into the solution (∼9%). The excess manganese ions have possibly formed small oxide clusters that did not precipitate out or, alternatively, were part of insoluble large oxide aggregates. Both of which were discarded in the purification steps after the synthesis. The XRD measurements of the Zn1-xMnxO nanocrystal powder displayed in Figure 2 revealed the standard Wurtzite structure of zinc oxide, with lattice parameters a ) 3.2485 Å and c ) 5.2051 Å, within 0.06% of that of the pure ZnO nanocrystals (a ) 3.2465 Å, c ) 5.2054 Å) and peak widths corresponding to nanocrystal size of 6.7 nm, in reasonable agreement with the TEM statistics. No other diffraction peaks were detected in this sample. The broadening of the XRD lines due to the small size of the nanocrystals makes the observed differences between the lattice parameters of doped sample compared with pure ZnO insignificant. The small difference between the ionic radii of the tetrahedrally coordinated zinc(II) and the manganese(II) (0.74 Å in Zn2+ compared to 0.80 Å in Mn2+) allows incorporation of Mn2+ into ZnO with only slight changes in lattice parameters.

Figure 1. (a) TEM image of the Zn1-xMnxO nanocrystals, x ) 0.04. Inset: high-resolution image of a relatively large nanocrystal revealing the single-crystal nature of these particles. The lattice fringes shown correspond to the a-axis spacing of ZnO. (b) Histogram of Mn percentage out of the total Mn + Zn content measured over about 40 individual nanoparticles using EDX. One measurement yielded 28% and was omitted from the histogram.

The manganese oxide sample showed a more complex diffraction line pattern that was primarily identified as a mixture of cubic MnO and Mn3O4 but with fcc MnO lattice parameter (4.41 Å) that is smaller than the bulk value by 0.8% and a different intensity pattern. The apparent uniformity of the observed manganese oxide nanocubes and the relatively large change in the MnO lattice parameter and intensity pattern hint that the two phases coexist within the same nanocrystals, as also proposed in ref 13. The TEM images of the ZnO nanocrystals, which were coated with manganese oxide (see Supporting Information), revealed relatively large nanocrystals with irregular shapes and large variations in Mn/Zn atomic ratios/nanocrystal (0-80%). Using high-resolution TEM of some of the smaller nanocrystals it was possible to observe a nonuniform coating. The narrow XRD lines for this sample fitted exactly the ZnO wurtzite structure with nanocrystal size over 50 nm and no trace of manganese related phase. Thus, with combination of the XRD, TEM, and EDX data, it is concluded that there was a nonuniform manganese oxide coating deposited on most of the ZnO nanocrystals in the form of a thin polycrystalline and/or amorphous film. The dried Zn1-xMnxO nanocrystal powder had a strong response to passing a small permanent magnet nearby at room temperature while the manganese oxide and manganese oxide coated ZnO samples did not show any visible response. The magnetization behavior of the different nanocrystal samples was investigated by SQUID magnetometry. Figure 3a displays a comparison between the magnetization curves of the

20234 J. Phys. Chem. B, Vol. 109, No. 43, 2005

Meron and Markovich

Figure 2. XRD of (a) Zn1-xMnxO nanocrystals, (b) ZnO nanocrystals coated with MnO, and (c) manganese oxide nanocrystals.

Zn1-xMnxO nanocrystals powder, the pure manganese oxide and the manganese oxide coated ZnO nanocrystals, all taken at 100 K. The magnetization scale is given in units of µB/manganese ion. It is clear from the sharp increase of the magnetization around zero field and partial low field saturation in the curve of the Zn1-xMnxO sample that at least part of the sample behaves as a ferromagnet with an average magnetic moment of the order of 0.1-0.2 µB/Mn2+ (at the partial saturation, 100 K) that is of the order of the values measured in bulk films.5 The two reference samples, manganese oxide and manganese oxide coated ZnO, yielded a straight line with an order of magnitude lower magnetic moment/Mn values, indicative of the weak paramagnetism typical of antiferromagnetic MnO. It should be added, however, that below 40 K the MnO samples did show ferromagnetic like hysteretic behavior similar to previous observations on manganese oxide nanocrystal samples.14,15 This is probably related to the presence of ferrimagnetic Mn3O4, which has a transition temperature of about 40-46 K. The combination of the structural XRD and TEM data, EDX analysis, and magnetization curves leads to the conclusion that the synthesized Zn1-xMnxO nanocrystals are primarily DMS like with Mn2+ embedded in the ZnO lattice. The high-temperature hydrolysis scheme appears to be particularly useful for efficient doping of the oxide nanocrystals, probably due to the fast kinetics of the hydrolysis at 100 °C, which makes reaction rates for the Zn-O-R and Mn-O-R hydrolysis and condensation into the oxide structure comparable. Other colloidal synthesis schemes based on high-temperature decomposition of metalcarboxylate salts, used recently to produce various oxide nanocrystals,16,17 were not able to produce efficient Mn doping of ZnO nanocrystals, possibly due to variance of decomposition rates between different metal-carboxylates.

Figure 3. (a) Magnetization curves measured at 100 K on powder samples of Zn1-xMnxO nanocrystals (blue), manganese oxide nanocrystals (red), and ZnO nanocrystals coated with manganese oxide (black). Magnetization units are µB/Mn ion. (b) Magnetization curves measured at 100 K on two samples of Zn1-xMnxO nanocrystals: (1) aggregated powder, red curve; (2) diluted in octadecane, black curve. Magnetization units are arbitrary. The inset shows the raw magnetization curve for the isolated particle sample, before subtraction of the octadecane’s diamagnetic response.

Figure 3b compares the magnetization curve of the Zn1-xMnxO nanocrystals powder to that of the isolated nanocrystals dispersed in octadecane and measured at 100 K. The magnetization scale here is arbitrary since it is difficult to obtain a good calibration of nanocrystal concentration in the dilute samples. The significant diamagnetic negative slope contributed by the octadecane matrix (seen in the inset of Figure 3b) was measured on a dilute ZnO nanocrystals sample (-7.5 × 10-8 emu/Oe) and removed from the isolated Zn1-xMnxO nanocrystals magnetization data. Note the large difference in coercive field between the two samples: over 3000 Oe in the isolated nanocrystals compared with ∼200-250 Oe in the aggregated ones. Thus, the low coercivity in the powder samples should be attributed to the strong interparticle interactions of dipolar nature and possibly even exchange type due to nanocrystal aggregation, which tends to reduce the coercivity.18 Studies of magnetization versus temperature on the various samples were conducted to complete the set of magnetic property data. Figure 4a displays raw data of magnetization versus temperature zero-field-cooled curves (ZFC) measured under a 100 Oe field for two samples in powder form: Zn1-xMnxO and MnO. These two curves as well as the one measured for the ZnO coated with MnO have a feature around 40 K, either a sharp rise or a peak in the magnetization which can be attributed

Ferromagnetism in Mn2+-Doped ZnO Nanocrystals

Figure 4. (a) Raw ZFC and FC magnetization vs temperature curves measured with 100 Oe for the Zn1-xMnxO nanocrystals powder compared with the ZFC curve measured for the manganese oxide nanocrystal powder. The ZFC curve for the isolated nanocrystals is also displayed for comparison. The curves have a negative baseline due to a diamagnetic background. Inset: ZFC imaginary AC susceptibility vs temperature measured under zero field at several frequencies for the Zn1-xMnxO sample. (b) Coercivity vs temperature for the powder and isolated Zn1-xMnxO nanocrystal samples.

to a phase transition in Mn3O4 and was previously observed in manganese oxide colloidal syntheses.14,15 Thus, any mention of MnO throughout the paper refers also to Mn3O4. Interestingly, a similar feature around 40 K appears in the ZFC data for the recently reported ferromagnetic Zn1-xMnxO bulk material.19 Apart from this feature, at higher temperatures, the ZFC curves of the MnO and ZnO coated with MnO powders remained low without any significant changes while the Zn1-xMnxO nanocrystals exhibited a strong broad peak at about 300 K after which a slow decrease in magnetization with temperature followed. The peak in the ZFC curve for the isolated Zn1-xMnxO nanocrystals appeared shifted to a lower temperature at ∼250 K. The black curve on the plot for the Zn1-xMnxO nanocrystals is the field-cooled (FC) magnetization vs temperature curve of the powder sample, which merges with the ZFC curve after the peak at 300 K. This implies that the 300 K magnetization peak is a superparamagnetic-like blocking temperature of the nanocrystals in the powder and explains the strong response of the powder to an external magnetic field gradient at room temperature. In addition, the inset in Figure 4a displays the imaginary ac susceptibility vs temperature curves, where the peaks shift toward higher temperatures with increasing frequencysagain typical of the slowing down of the magnetization switching occurring at the blocking temperature for superparamagnetic particles. Figure 4b displays the dependence of the coercive field on temperature for both Zn1-xMnxO nanocrystals samples (isolated and powder), where similar temperature dependencies

J. Phys. Chem. B, Vol. 109, No. 43, 2005 20235 can be observed despite the order of magnitude difference in its absolute value. The initial increase in coercivity as the temperature was raised above 50 K was due to the contribution of the ferrimagnetic Mn3O4 to the magnetization curves below this temperature. As the nulling of the coercivity roughly coincides with the ZFC curve peak, it provides support for the superparamagnetic behavior of the nanocrystals. The magnetic measurements of the Zn1-xMnxO nanocrystal samples yielded two surprising results: (1) unprecedented high coercivity at low temperatures for the isolated nanocrystals, which is an order of magnitude higher compared with previous work and with the present powder sample; (2) high blocking temperature peak (∼250-300 K), which, considering the small size of the particles, indicates a very high magnetic anisotropy value, higher than highly magnetically anisotropic CoFe2O4 particles of comparable size, for example.12 Assuming that this anisotropy comes from magnetocrystalline anisotropy, one could use the relation KV ∼ 25kBTb to estimate the magnitude of K, the magnetocrystalline anisotropy constant, where V is the average nanocrystal volume, kB, the Boltzmann constant, and Tb, the blocking temperature. Thus, K ∼ 1 × 107 erg/cm3, while, for CoFe2O4, K ∼ (2-3) × 106 erg/cm3.20 Using an estimate for the zero temperature saturation magnetization value, MS(0), in the Zn1-xMnxO nanocrystals, one could calculate the anisotropy field (HK) of ideal noninteracting single domain magnetic particles with uniaxial magnetic anisotropy according to the Stoner-Wolfarth model: HK ) 2K/MS(0). HK serves as an upper boundary for the coercive field measured on a randomly oriented single domain particle sample, while the true coercive field should be some fraction of it. MS(0) was roughly estimated by fitting the magnetization curves of the powder sample at several temperatures to a combination of a fast saturating component and a paramagnetic component using a function of the form M(H) ) a + bH + c/H + d/H2 ,21 and extrapolating the constant a to T ) 0. This constant represents MS(T) of the ferromagnetically coupled clusters. Although the above function could not fit well the M(H) curves, its use provided a rough estimate of MS(0) ∼ 0.35 emu/g, and consequently, one gets HK ∼ 1 × 107 Oe, which is about 3.5 orders of magnitude larger than the coercive field measured at 50 K for the isolated particles sample. Thus, it is clear that the Zn1-xMnxO nanocrystals cannot be considered ideal single domain magnetic nanocrystals, either in the sense that their magnetization orientation does not rotate coherently or that instead of being dominated by magnetocrystalline anisotropy, their magnetic anisotropy is mainly contributed by surface anisotropy. We believe that both reasons are interrelated and are responsible for the abnormal superparamagnetic-like behavior of these nanocrystals. One difference that distinguishes the Zn1-xMnxO nanocrystals from regular anisotropic ferro-/ferrimagnetic nanocrystals is the slow approach to saturation in the Zn1-xMnxO sample at high fields. This indicates that part of the sample is not ferromagnetically coupled but more weakly interacting, possibly magnetically disordered. Trials to accurately fit the high field part of the Zn1-xMnxO magnetization curves in the powder or isolated nanocrystal samples to various physically reasonable functions did not succeed. The high-field part of the magnetization curves did not fit a Brillouin function as should be in the case of purely paramagnetic noninteracting Mn2+ ions.9 Consequently, the shape of the magnetization curves could be interpreted as representing a combination of two major parts in the sample: (1) ferromagnetically coupled Zn1-xMnxO clusters, which include the minority of the Mn2+ ions in the sample; (2) smaller

20236 J. Phys. Chem. B, Vol. 109, No. 43, 2005 magnetic moments, probably corresponding to magnetically disordered clusters of exchange coupled Mn ions, accounting for the majority of the manganese ions. Part of these nonferromagnetically coupled ions are probably free paramagnetic ions. It is difficult to provide a good estimate of the percentage of Mn2+ ions involved in the ferromagnetic part of the sample. An upper limit for this number can be obtained by assuming that the high-field slopes of the magnetization curves are only contributed by free Mn2+ ions with each having a spin-only magnetic moment of about 5.9 µB.22 The result of such an estimate is that about 40% of the Mn2+ contributes to the ferromagnetism in the powder of Zn1-xMnxO nanocrystals. In reality this number may be below 10%, as might be the case if the disordered magnetic parts of the sample are mostly antiferromagnetically coupled, for example. This would also imply that the ferromagnetically coupled Mn2+ ions may contribute an order of magnitude higher magnetic moment (compared with Figure 3a), of the order of 1-2 µB/ion. The anomalous magnetic behavior of the Zn1-xMnxO nanocrystals is somewhat similar to that reported for NiO nanocrystals by Berkowitz and co-workers,23 where the NiO, which is antiferromagnetic in bulk crystal form, became ferromagneticlike in small clusters. These nanocrystals showed relatively high coercivity and blocking temperature and a magnetic moment/ Ni ion that is much larger than the one observed in the bulk. This phenomenon was explained by a surface-induced change in the magnetic order of the nanocrystals. One difference between the NiO and the Zn1-xMnxO systems is that in the Zn1-xMnxO nanocrystals the magnetization curves were always symmetric with respect to H ) 0, while loop shifts were observed for NiO nanocrystals, indicating the existence of exchange bias between antiferromagnetic parts and ferromagnetic parts of the nanocrystals. Thus, we conclude that in the present case the nonferromagnetically coupled parts of the sample are not antiferromagnetic but disordered, spin-glass like, as reported previously for certain preparations of the bulk material.24 Another indication to the existence of a spin glass in addition to the ferromagnetic part is the slight irreversibility that occurred in the magnetization curves at high fields, as can be seen in the curve of the powder Zn1-xMnxO sample in Figure 3b. It is intriguing to try to understand the ferromagnetism observed in this work in the light of models invoked to explain magnetism in DMS materials.25,26 As many of the models assume a high density of charge carriers in the conduction, valence, or some impurity band within the band gap to obtain room-temperature ferromagnetism, it is possible that the synthesis described here creates such a condition. Traditionally, defects in ZnO were considered responsible for n type doping which could mediate the long-range exchange interactions between the paramagnetic metal centers. It is yet to be determined whether the two magnetic phases (ferromagnetic and disordered) appear in different nanocrystals, possibly with different doping levels and/or coexist within individual nanocrystals in the form of ferromagnetic core and magnetically disordered shell. Support for the occurrence of the latter configuration can be found in the anomalously high magnetic anisotropy measured in the isolated nanocrystals, which most probably arises from large surface magnetic anisotropy. Such anisotropy is often explained by spin disorder at the surface.27 Conclusion It was shown that it is possible to produce nanocrystals of Zn1-xMnxO using a novel colloidal chemistry technique. Such

Meron and Markovich isolated nanocrystals with average size of 5.5 nm and about 150 Mn2+/particle exhibited ferromagnetic properties, which survived up to room temperature for part of them. The observed ferromagnetism cannot be attributed to foreign manganese oxide related phases. It is demonstrated here that DMS in the form of nanocrystals may have an advantage over the analogous bulk phase in the form of high coercivity. Further refinement of the synthesis technique with control over doping level and doping uniformity is expected to contribute significantly to the understanding of this form of magnetism. Acknowledgment. This work was supported by Israel Science Foundation Grant No. 208/03 and the James Frank program. Supporting Information Available: High-resolution TEM, temperature-dependent magnetization curves, and magnetization data for 12 nm diameter nanocrystals with x ) 0.05. This material is available free of charge via the Internet at http:// pubs.acs.org. References and Notes (1) Tanaka, M.; Higo, Y. Phys. ReV. Lett. 2001, 87, 026602. (2) Koshihara, S.; Oiwa, A.; Hirasawa, M.; Katsumoto, S.; Iye, Y.; Urano, C.; Takagi, H.; Munekata, H. Phys. ReV. Lett. 1997, 78, 4617. (3) Akinaga, H.; Ohno, H. IEEE Trans. Nanotechnol. 2002, 1, 19. (4) Matsumoto, Y.; Murakami, M.; Shono, T.; Hasegawa, T.; Fukumura, T.; Kawasaki, M.; Ahmet, P.; Chikyow, T.; Koshihara, S.; Koinuma, H. Science 2001, 291, 854. (5) Sharma, P.; Gupta, A.; Rao, K. V.; Owens, F. J.; Sharma, R.; Ahuja, R.; Osorio Guillen, J. M.; Johansson, B.; Gehring, G. A. Nat. Mater. 2003, 2, 673. (6) Pearton, S. J.; Heo, W. H.; Ivill, M.; Norton, D. P.; Steiner, T. Semicond. Sci. Technol. 2004, 19, R59. (7) Norton, D. P.; et al. Appl. Phys. Lett. 2003, 83, 5488. (8) Schwartz, D. A.; Norberg, N. S.; Nguyen, Q. P.; Parker, J. M.; Gamelin, D. R. J. Am. Chem. Soc. 2003, 125, 13205. (9) Norberg, N. S.; Kittilstved, K. R.; Amonette, J. E.; Kukkadapu, R. K.; Schwartz, D. A.; Gamelin, D. R. J. Am. Chem. Soc. 2004, 126, 9387. (10) Schwartz, D. A.; Gamelin, D. AdV. Mater. 2004, 16, 2115. (11) O’Brien, S.; Brus, L.; Murray, C. B. J. Am. Chem. Soc. 2001, 123, 12085. (12) Meron, T.; Rosenberg, Y.; Lereah, Y.; Markovich, G. J. Magn. Magn. Mater. 2005, 292, 11. (13) Park, J.; Kang, E.; Bae, C. J.; Park, J. G.; Noh, H. J.; Kim, J. Y.; Park, J. H.; Park, H. M.; Hyeon, T J. Phys. Chem. B 2004, 108, 13594. (14) Lee, G. H.; Huh, S. H.; Jeong, J. W.; Choi, B. J.; Kim, S. H.; Ri, H. C. J. Am. Chem. Soc. 2002, 124, 12094. (15) Park, J.; Kang, E.; Bae, C. J.; Park, J. G., Noh, H. J.; Kim, J. Y.; Park, J. H.; Park, H. M.; Hyeon, T. J. Phys. Chem. B 2004, 108, 13594. (16) Yin, M.; Gu, Y.; Kuskovsky, I. L.; Andelman, T.; Zhu, Y.; Neumark, G. F.; O’Brien, S. J. Am. Chem. Soc. 2004, 126, 6206. (17) Jana, N. R.; Chen, Y. F.; Peng, X. G. Chem. Mater. 2004, 16, 3931. (18) El-Hilo, M.; Chantrell, R. W.; O’Grady, K. J. Appl. Phys. 1998, 84, 5114. (19) Sharma, P.; Gupta, A.; Owens, F. J.; Inoue, A.; Rao, K. V. J. Magn. Magn. Mater. 2004, 282, 115. (20) Buschow, K. H. J., Handbook of Magnetic Materials; NorthHolland: Amsterdam, 1995; Vol. 8. (21) Respaud, M. J Appl. Phys. 1999, 86, 556. (22) du Tremolet de Lacheisserie, E.; Gignoux, D.; Schlenker, M. Magnetism I-Fundamentals; Kluwer: Dordrecht, The Netherlands, 2002; Chapter 7. (23) Kodama, R. H.; Makhlouf, S. A.; Berkowitz, A. E. Phys. ReV. Lett. 1997, 79, 1393. (24) Kolesnik, S.; Dabrowski, B.; Mais, J. J. Appl. Phys. 2004, 95, 2582. (25) Coey, J. M. D.; Venkatesan, M.; Fitzgerald, C. B. Nat. Mater. 2005, 4, 173. (26) Dietl, T.; Ohno, H.; Matsukura, F.; Cibert, J.; Ferrand, D. Science 2000, 287, 1019. (27) Iglesias, O.; Labarta, A. Physica B 2004, 343, 286.