Critical Analysis on the Structural and Magnetic Properties of Bulk and Nanocrystalline Cu-Fe-O

Nanocrystalline and bulk samples of “Fe”-doped CuO were prepared by coprecipitation and ceramic methods. Structural and compositional analyses were performed using X-ray diffraction, SEM, and EDAX. Traces of secondary phases such as CuFe 2O 4, Fe 3O 4, and α-Fe 2O 3 having peaks very close to that of the host CuO were identified from the Rietveld profile analysis and the SAED pattern of bulk and nanocrystalline Cu 0.98Fe 0.02O samples. Vibrating Sample Magnetometer (VSM) measurements show hysteresis at 300 K for all the samples. The ferrimagnetic Neel transition temperature ( 𝑇 𝑁 ) was found to be around 465°C irrespective of the content of “Fe”, which is close to the 𝑇 𝑁 value of cubic CuFe 2O 4. High-pressure X-Ray diffraction studies were performed on 2% “Fe”-doped bulk CuO using synchrotron radiation. From the absence of any strong new peaks at high pressure, it is evident that the secondary phases if present could be less than the level of detection. Cu 2O, which is diamagnetic by nature, was also doped with 1% of “Fe” and was found to show paramagnetic behavior in contrast to the “Fe” doped CuO. Hence the possibility of intrinsic magnetization of “Fe”-doped CuO apart from the secondary phases is discussed based on the magnetization and charge state of “Fe” and the host into which it is substituted.


Introduction
Oxides such as ZnO, TiO 2 , and SnO 2 are currently explored as host materials for preparing Diluted Magnetic Semiconductors (DMSs) [1][2][3][4][5][6][7]. Among these, diamagnetic ZnO (n-type) doped with "3d" transition metal ions is presently explored in detail ( [8][9][10][11][12][13] and references there in). Wide controversies exist in literature reports in explaining the observed results, and the role of preparation methods in deciding the property. CuO (Tenorite), a p-type semiconductor with a band gap of 1.2 eV, is an essential component in a variety of oxide superconductors and is also known for its catalytic and gas sensing properties [14][15][16]. CuO also being nontoxic and abundant, an attempt has been made to unravel its properties by doping with "Fe" and to find its potential for application in SPINTRONICS. "Cu" has three oxidation states, Cu + , Cu 2+ , and Cu 3+ , because of which both hole doping and electron doping are possible. "Fe" doping into CuO has been attempted by researchers [17][18][19][20][21] by different preparation methods and was found to have no definite evidence for "Fe" replacing the "Cu" site in CuO, as the two components are immiscible. Smith et al. have reported that introducing Fe 2 O 3 into CuO by ceramic method is tough and phase segregation results above 0.3% of "Fe" in CuO [17]. Few reports exist on CuO doped with "Mn" [22] and "Fe" [23,24] from DMS point of view. On the other hand, Cu 2 O (cuprite), also an eco-friendly p-type semiconductor with a direct band gap value of 2 eV [25], is useful as an energy converter for solar cell applications [26] and as humidity and gas sensor material [27,28]. Reports also exist on magnetic properties of "Mn-", "Co-" and "Fe-" doped Cu 2 O [25,29,30]. Comparison of "Fe-doped Cu 2 O" with "Fe-doped CuO" may throw some light on the understanding of the intrinsic magnetic contribution of the later.
In this work, paramagnetic CuO (antiferromagnetic below 230 K) doped with varying concentration of "Fe" 2 Advances in Materials Science and Engineering was characterized in detail for its structural, compositional, and magnetization behavior by X-Ray diffraction (XRD), Rietveld analysis, Scanning electron microscopy (SEM), Energy Dispersive X-Ray analysis (EDAX), Transmission Electron Microscopy (TEM), Vibrating Sample Magnetometer (VSM), Thermogravimetric analyzer (TGA), Mössbauer spectroscopy, and XRD at high-pressures using synchrotron radiation. Presence of other related phases like CuFe 2 O 4 /Fe 3 O 4 and Fe 2 O 3 , having coincidence with many of the major peak positions of CuO, was not categorized before through the powder X-ray diffraction in detail. Their presence is only discussed based on the Mössbauer spectra [20,21,23]. We performed a detailed powder Xray diffraction study using Rietveld refinement procedure and cross checked the presence of CuFe 2 O 4 from the Neel temperature (T N ) measurement using a TGA with a small applied field. High-pressure studies were carried out to identify the presence of any possible secondary phases. We have also discussed the effect of secondary phases in CuO and also compared with 1% and 2% "Fe"-doped Cu 2 O. The possibility of intrinsic magnetization of "Fe"-doped CuO [24] apart from the ferrite phases is discussed based on the magnetization measurements and the charge state of "Fe" in the host into which it is substituted.

Sample Preparation and Experimental Details
Nanocrystalline samples of Cu (1−x) Fe x O (x = 0.01, 0.02, 0.03, 0.04, and 0.05) were prepared from the stoichiometric solutions of copper nitrate and iron nitrate in deionized water by coprecipitation method after refluxing for 20 hour, 30 hour, 40 hour and 50 hour. Precipitation was done using aqueous NH 4 OH and peptized with deionized water to remove by-products. The filtrates were then dried, powdered and annealed at various temperatures. Nanocrystalline CuFe 2 O 4 and bulk Cu (1−x) Fe x O (x = 0.01, 0.02, 0.03, 0.04, 0.05 and 0.07) were also prepared for comparing the properties. For the preparation of bulk samples, stoichiometric mixtures of oxides were ground in isopropyl alcohol medium, pelleted and sintered at 900 • C/12h in air. Chemical coprecipitation method with subsequent reduction of the precipitates was employed to prepare the samples of Cu  Table 1. This overlap of reflections presented the problem of identifying hidden impurity phases.

Rietveld
Analysis. XRD measurement with a step size of 0.01 • with an exposure time of 8 seconds was performed on the as-mixed powders of CuO : Fe 2 O 3 (2%) and on the same mixtures but sintered at 900 • C/12h in air. Interestingly, from the close scan XRD spectrum, the subtle merged peaks belonging to the ferrimagnetic cubic CuFe 2 O 4 phase were identified. The full-scale view of the close scan XRD pattern of as-mixed CuO : Fe 2 O 3 (2%) presents peaks of CuO and Fe 2 O 3 . On sintering at 900 • C/12h in air, all the lines of Fe 2 O 3 vanished indicating the substitution of "Fe" into the copper site and the CuO main phase lines were alone evident. The phase quantification by Rietveld refinement using a program called MAUD [32], which combines the Rietveld method and a Fourier transform analysis, also indicated the existence of 0.2% of CuFe 2 O 4 in the samples. The results of refinement are given in Table 2. The observed pattern matches with that    Figure 2. The square root of intensity is plotted so as to clearly depict the closeness of the peaks and also to show the presence of less intense impurity peaks.

SEM and EDAX Results
. SEM measurements were performed on nanocrystalline Cu 0.95 Fe 0.05 O samples obtained after refluxing for 40 hour. The image corresponding to the as-prepared sample shows particles with rod-like and disc-shaped morphologies (Figure 3), and their thickness increases when annealed at 500 • C/2h in air; however it was less than a micrometer. The thickness and length of the rod shaped particles alone were measured. A Gaussian fit was given and the average value of thickness was found to be 78 nm with an average length of 0.366 μm.
As the temperature was raised to 700 • C/2h, the rod-like morphology vanishes and the particles grow in size. At 900 • C/2h the particle morphology transforms to platelets due to temperature-induced agglomeration and growth (not shown here). Compositional analysis by EDAX confirmed the desired composition of "Fe" in the entire batch of samples (not presented here). The nominal variation in composition is due to the random alloying of the magnetic dopants. The observation of two different particle morphologies, rod-like and platelet-like may be an indication of existence of two different phases. Rod-and platelet-like morphology may be the undoped CuO and CuFe 2 O 4 phases, respectively. This argument is supported by the magnetization results which are discussed in the following sections. The samples annealed at lower temperature are found to have paramagnetic component [24] without saturation superimposed with the hysteresis behavior (bottom inset of Figure 10). This is because the temperature is not sufficient for the CuFe 2 O 4 to crystallize out. Hence, after annealing at 900 • C in air, the magnetization curve becomes smooth and the corresponding SEM micrograph presented only platelet-like particles (not presented here).  Table 3. This is also confirmed from SAED patterns for samples taken from three different locations (not shown here). From the above result, it is seen that Cu 0. 95 The TEM micrograph of nanocrystalline Cu 0.95 Fe 0.05 O (900 • C/2h) shown in Figure 5 has agglomerates of smaller particles due to annealing of the sample. The SAED pattern (inset of Figure 5) shows well-defined spots indicating formation of highly crystalline particles on annealing. The d hkl and hkl values are tabulated in Table 4. All the d hkl values match with those of the parent CuO except for the 3.310Å d hkl , which has good coincidence with that of (012) plane of α-Fe 2 O 3 . Hence the nanocrystalline Cu 0.95 Fe 0.05 O sample also has the possible secondary phases as observed in the SAED of bulk Cu 0.95 Fe 0.05 O. The equation representing the sample formation with possible secondary phases is given as follows: (1)

High-Pressure X-Ray Diffraction Using Synchrotron Radiation.
Pressure is a fundamental thermodynamic degree of freedom available for basic investigations as well as for processing of materials. The effects of pressure are a result of changes in the band structure and of the energy content of a material produced by the reduction of inter atomic spacing. The study of the behavior of materials at high-pressures has been useful in the observation of new features of the physical and chemical properties. In order to produce appreciable effects on the properties of condensed phases, high-pressures of the order of hundreds or thousands of bars are required. Synchrotron radiation is a photon light source generated by high-energy electrons that are centripetally accelerated to relativistic velocities in the magnetic fields of a storage ring and are emitted in a narrow cone tangent to the orbit. Synchrotron radiation, 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24   because of unique properties like high brightness, wide energy spectrum, high degree of polarization, and high collimation has allowed extensive investigation of a variety of materials. The remarkable qualities of diamond such as high yield strength and transparency to radiation over a wide range of wavelengths have made it an obvious choice for the anvil materials making it possible to conduct highpressure X-ray diffraction measurements on materials at few hundreds of giga Pascal (GPa) using synchrotron radiation. Presence of secondary phases can be evidenced from the high-pressure X-ray diffraction measurements due to the fact that the elastic properties of CuO, Fe, and CuFe 2 O 4 are different. Hence, we are motivated to study the sample at high-pressures to identify the presence of any hidden secondary phases.
High-Pressure XRD measurements were conducted in angle dispersive geometry using Mao-Bell type Diamond Anvil Cell. Experiment was performed using the synchrotron source at beamline 16IDB, High-Pressure Collaboration Access Team (HPCAT), Advanced photon source, Chicago, with a wavelength of 0.4564Å. Bulk CuO : Fe (2%) powder was loaded along with a few grains of platinum which acts as a pressure standard [33]. A stainless steel gasket of 50 μm thickness indented using diamond anvils of culet size 300 μm, wherein a 100 μm hole drilled serves as the sample chamber. The size of the incident beam was 10 × 10 μm 2 and the diffracted beam was recorded on a Mar 3450 image plate. The diffraction measurements were carried out from 0.5 GPa to 44 GPa at room temperature and are shown in Figure 6. Each pattern was fitted and refined to monoclinic structure of CuO with space group C2 1 /c using the software GSAS [34] with EXPGUI [35]. Though no structural transitions had been reported up to 100 GPa in monoclinic CuO [36][37][38][39][40], the study helps in identifying the hidden phases based on the variation in compressibility.
In the case of 2% Fe-doped CuO, application of pressure causes a shift in the diffraction peaks towards  higher angle indicating reduction in cell volume. Relative variation of refined lattice parameters as a function of pressure is shown in Figure 7. The overall trend on application of pressure on the "b" and "c" lattice parameters is to decrease them with increasing pressure and that "a" increases gradually (Figure 7). The cell volume obtained from the refinement is found to decrease with pressure ( Figure 8) and no abrupt change is Advances in Materials Science and Engineering 7 observed indicating absence of structural transition. Literature reports indicate absence of structural transition from high-pressure electrical resistivity measurements in monoclinic CuO up to 100 GPa [36,40]. Upon increasing the pressure, at 3.98 GPa, a peak clearly separates out from the (200) plane. Hence, when CuO gets compressed and shifts towards higher angle, an unknown peak becomes evident. With further increasing pressure, we observed that the (−111) peak of monoclinic CuO shifts towards higher angle and tends to overlap with the unknown peak. As the pressure is further increased, the intensity of the unknown peak gradually decreases and completely vanishes at 26.02 GPa before overlapping with the (−111) plane of CuO. Similar behavior has been observed in pure CuO itself (the results will be published elsewhere). Reimann and Syassen [41] observed changes in the Raman modes of pure CuO in high-pressure Raman measurements up to 34 GPa which they explain as the changes in the internal structure parameter. Reimann and Syassen also report Malinowski to observe a structural anomaly in his X-ray diffraction measurement around 10 Gpa (referred as private communication in back reference of [41]). Highpressure neutron diffraction studies by Ehrenberg et al. [38] suggest Jahn Teller type structural distortion seen below 8 GPa in pure CuO. Hence, appearance and suppression of an additional peak in the high-pressure XRD of 2% Fe-doped CuO sample is also characteristic of pure CuO. The phase transformation is similar to Jahn Teller-like distortion which occurs at high-pressure related to the variation of the bond angle with pressure. In addition to this observation, the absence of any other additional peaks evidences absence of secondary phases of them to be present below the limit of detection. The conclusion was drawn based on the variation of the bulk modulus of CuO and various secondary phases.
The (−111) peak which was the maximum intensity peak at 0.5 GPa loses its intensity gradually with increasing pressure, and at the same time, the (200) peak gain intensity. This may be due to the pressure-induced preferred orientation. This small internal structural distortion [37,38,41] is found to be reversible upon releasing the pressure. However, there is a slight shift towards higher angle in the peak positions along with broadening, also, after releasing the pressure, the (200) plane becomes the peak with maximum intensity instead of (−111) plane which may be due to the pressure-induced lattice strain and preferred orientation. However, pressureinduced substitution of a fraction of "Fe" at the "Cu" site might have also resulted which can be understood in more detail from high-pressure neutron diffraction measurement which is more sensitive to the magnetic moment of the elements in the sample when performed in the similar pressure range. The PV-equation of state fitted using the third-order Birch-Murnaghan equation of state (BM-eos) is given by where K 0 is bulk modulus, K 0 is derivative of bulk modulus with pressure, V 0 is zero pressure volume, and V is volume at pressure P. The value of K 0 set to 4 shown in Figure 8 indicates a steady decrease in volume on increasing the pressure with no abrupt change indicating absence of phase transition. Hence, absence of secondary phases is evident from the high pressure XRD data and even if present, it is well below the level of detection.
2.6. TGA Results. To identify the effect of "Fe%" doping on the T N of the samples, the TGA measurements were performed with a small applied field (≈20 Oe) to determine the T N as a function of temperature. The measurement was performed in N 2 atmosphere with a heating rate of 10 • C/min up to 600 • C. For the 50 hour refluxed nanocrystalline Cu 0.95 Fe 0.05 O sample, the T N was found to increase with decreasing crystallite size ( Figure 9). The increase in magnetization prior to T N in the TGA curves observed for Cu 0.95 Fe 0.05 O is due to Hopkinson effect which is also reported in the case of NiFe 2 O 4 [42]. The bulk Cu 0.95 Fe 0.05 O had a T N of 458 • C, whereas a T N of 479 • C (900 • C/12h) for the 49 nm sample and that of 486 • C (900 • C/2h) for the 25 nm sample were observed. Thus the decrease in crystallite size increases the T N . However, this variation of T N may be due to the size effect of the copper ferrite phase that has formed within the sample. To avoid the effect of particle size on T N , we studied the bulk samples prepared by ceramic method to trace the original trend of variation of T N with "Fe" concentration in CuO. The T N observed for "Fe"doped CuO bulk samples is shown in Table 5. There was no noticeable trend in the variation of T N . The T N remained around 460 • C irrespective of the "Fe%", indicating that one major secondary phase had formed within the sample. This T N value has good coincidence with the T N of cubic CuFe 2 O 4 , which is 465 • C [43], whereas the T N of cubic Fe 3 O 4 is 585 • C. This observation is a clear experimental evidence that the observed hysteresis behaviour is from the phase-segregated CuFe 2 O 4 . This confirms that the "Fe" does not get substituted into the substitutional site of CuO even for 1% of "Fe" in CuO to effect any change in the property of CuO; rather it favors the formation of cubic CuFe 2 O 4 as segregates within the sample whereas Park et al. [24] do not report any ferrite impurity phase; they instead discuss intrinsic ferromagnetic behaviour in 2% " 57 Fe"-doped CuO mediated by carriers localized around the oxygen vacancies studied by Mössbauer measurements. For "Fe"-doped bulk CuO, the T N is found to be around 460 • C. The T N of Fe 3 O 4 is 585 • C and is higher than the T N observed for bulk Cu 0.95 Fe 0.05 O whereas the T N of cubic CuFe 2 O 4 is 465 • C [43]. The T N of bulk CuO doped with Fe to various extents was very close to that of cubic CuFe 2 O 4 . Hence no appreciable variation in T N was found with varying "Fe" concentration. Hence, even the 1% and 2% "Fe"-doped CuO is inevitably found to possess copper ferrite phase. CuFe 2 O 4 when quenched or rapidly cooled (from above 760 • C), takes cubic structure with a = 8.37Å. If cooled slowly, it attains tetragonal phase with a = 8.22Å and c = 8.70Å [43]. Since all the samples were furnace cooled from  [23]. Park et al. [24] have reported ferromagnetic hysteresis behavior for annealing above 500 • C; however they explain it as intrinsic behavior due to indirect coupling among the Fe 3+ ions via a localized carrier at oxygen vacancy. Treatment at 500 • C is not sufficient to completely crystallize out CuFe 2 O 4 . We could also observe a paramagnetic-like behaviour for the samples annealed at 500 • C/2h from VSM (bottom inset of Figure 10) and the loops did not saturate but were found to show hysteresis at room temperature when annealed at 900 • C/2h. Hence, after annealing at 900 • C in air, the magnetization curve becomes smooth and the corresponding SEM micrograph also presented only plateletlike particles (not shown here). The M max values were found to increase with increasing refluxing hour (top inset of Figure 10) and also found to increase with increasing "Fe%" in nanocrystalline CuO samples refluxed for 50 hour and annealed at 900 • C/2h in air ( Figure 11). Inset of Figure 11 shows the comparison of the magnetization data of all the samples with pure copper ferrite; note the large difference in magnetization. This may be presumably due to increasing volume fraction of CuFe 2 O 4 within the samples. The nonsaturation of loops may be due to the following reasons: (a) presence of clusters of undoped paramagnetic CuO (at 300 K), and (b) incomplete formation of ferrite secondary phases, which are responsible for the observed dominant magnetic behaviour. In our case of "Fe"-doped CuO, ferrimagnetic correlations among "Fe 2+ " atoms substituting the "Cu" site may lead  do more work to explain from the standpoint of "Fe 2+ " substituting the "Cu" site by delineating the contribution from the segregated CuFe 2 O 4 phase.
The magnetization measurements on bulk 1% and 2% Fe-doped CuO performed at 77 K showed distorted hysteresis loops ( Figure 12). This is because the measurement temperature is well below the Neel temperature (T N ) of CuO. The noncollinear spin arrangement due to partial Fe substitution at the Cu site now weakens and the spin alignment tend towards (perfect anticollinear) antiferromagnetic arrangement. The distorted hysteresis ( Figure 12) showed a loop shift away from the M = 0 (inset of Figure 12) which is known as exchange bias [45]. Exchange bias is commonly observed in a mixture of antiferromagnetic (AFM) and ferromagnetic (FM) type compounds [46]. Susceptibility studies also show three dimensional AFM behavior below 212 K [47].
The 1% "Fe"-doped CuO showing ferrimagnetic behavior at 300 K is found to present distorted signal at 77 K. At 77 K, CuO is antiferromagnetic; hence the spins tend towards anticollinear arrangement leading to decrease in magnetization. However, if reasonable quantity of CuFe 2 O 4 is present, then, an increase in magnetization at 77 K should have been observed, but, to the contrary it deteriorates. At 77 K, the sample becomes an antiferromagneticferrimagnetic composite mixture, whereas at 300 K, it is a paramagnetic-ferrimagnetic mixture. At 77 K, the loss of hysteresis behavior could be explained based on the antiferromagnetic transition occurring in CuO in which spins

Results and Discussion
Based on the above given experimental observations, formation of cubic CuFe 2 O 4 is confirmed. However one could not discard the candidature of CuO to be an efficient material from applications point of view. This is because, even if CuFe 2 O 4 phase persists in CuO, the material can be utilized for the fabrication of magnetic tunnel junction (MTJ-) like structures and also there may be an intrinsic magnetic behaviour due to "Fe" substitution in the "Cu" site.  [55]. The coercive field of the two magnetic electrodes in an MTJ should be well separated and the magnetization reversal of these layers should be abrupt. Hence without attaining single phase one may try to use this material for multilayer switching applications considering it as a composite material of "pinned" type magnetization; the discussion in this point of view is given in this section.

Pinned Type Composite Magnetic
Material. The VSM measurements on CuO doped with varying "Fe%" show the M max value to increase. Since CuFe 2 O 4 present in the sample is mainly responsible for the observed magnetic behavior, then with increasing "Fe%" in CuO, the volume fraction of CuFe 2 O 4 has to increase; that is, the samples should tend towards more soft nature, with high M max and low H C . Instead, we observe a linear increase in M max and an increase in H C with increasing "Fe%" instead of decrease in H C ( Figure 13) for 1, 4, and 7% "Fe"-doped CuO, which may be due to domain wall pinning effects. The spontaneous magnetic moment per "Fe" atom in CuO for 1, 4, and 7% are 0.0057 μ B , 0.0168 μ B , and 0.0276 μ B , respectively. Besides the enhancement of H C , the squareness ratio, M R /M S also increases with "Fe%" in CuO. The value of squareness ratio of 1, 4 and 7% "Fe" doped CuO is 0.249, 0.448, and 0.519, respectively. This is an indication that the "Fe"-doped CuO tend towards "hard" nature in spite of the presence of soft magnetic CuFe 2 O 4 .
The H C value of 7% "Fe"-doped bulk CuO (620 Oe) is three times greater than the value of 1% "Fe"-doped bulk CuO (202 Oe) (Figure 13 inset) indicating that the layer of "Fe" composition-varied CuO thin films may show exchange coupling. High H C CuO layer may be used as a "pinned" layer in MTJ structures. The H C of bulk and nanocrystalline Cu 0.95 Fe 0.05 O was found to be 236 Oe and 1225 Oe, respectively. The H C increased five times in magnitude in the nanocrystalline sample ( Figure 14). The M max value of nanocrystalline CuO : Fe (1% to 5%) is found to increase with "Fe" concentration ( Figure 14 inset). These aspects could induce exchange coupling which is desirable for device fabrication if one could stabilize a single phase Cu-Fe-O solid solution without any ferrite formation by adopting different preparation techniques or by preparing a sample with homogenously dispersed CuFe 2 O 4 in CuO.
By comparing the virgin magnetization curve in the hysteresis loop, one recognizes two qualitatively different kinds of behavior: (i) nucleation type and (ii) pinning type [56]. In a nucleation type magnet, the virgin curve is steep and saturation is reached under fields much lower than the saturation loop coercive field. Domain walls are present in the virgin state and the fact that the material can be easily saturated shows that walls are free to move and do not experience any important pinning effects. Once the virgin domain structure has been swept away, the formation of reversed domains becomes a difficult process, and the demagnetization curve is characterized by a substantial coercivity.
In a pinning type magnet, fields of the order of the saturation loop coercive field are required to saturate the material even when one starts from the virgin state. This indicates that domain wall pinning is the main mechanism responsible for coercivity. Domain wall motion is the dominant mechanism and the loop width is mainly determined by wall pinning effects [56]. The impediments to wall motion will determine the coercive field. The physical picture summarized by the term wall pinning is one where the impediments to wall motion arise from some form of structural disorder; dispersed phases with magnetic properties different from those of the matrix, as well as nonmagnetic inclusions or cavities, may also be important sources of pinning. The strength of the pinning effect may critically depend on the dimensionality of the pinning sources. The strength, density, and dimensionality of pinning centers can affect the value of H C . Though the solubility of "Fe" in CuO is very much limited, a very small quantity of any immiscible component will always dissolve in another component, as this increases the configurational entropy and lowers the free energy of the crystal [57]. Hence, though the formation of CuFe 2 O 4 is more favorable while trying to dope "Fe" in CuO, a fraction of "Fe" also enters into the Cu 2+ site as substitutional impurity. Hence contribution from intrinsic magnetization from the fraction of Cu-Fe-O substitutional solid solution formed within the sample apart from the composite like behavior of paramagnetic CuO dispersed with CuFe 2 O 4 particles cannot be disregarded. Hence, one has to really ascertain its effect delineating the major magnetic contribution from the segregated CuFe 2 O 4 phase. The intrinsic magnetic behavior could result from the spin alignment induced by the fraction of "Fe" substituting the "Cu" site. One could also argue that as the "Fe%" in CuO increases, the fraction of "Fe" substituting the cationic site also increases reasonably. However to cross check this argument, neutron diffraction measurements at ambient and at high-pressures are necessary.    For the formation of AB 2 O 4 spinel structure, A is required to be a dipositive ion, but in Cu 2 O, "Cu" is in +1 oxidation state and hence formation of spinel CuFe 2 O 4 is not favored. Due to the absence of this spinel impurity phase, "Fe" doped Cu 2 O does not show a dominant ferromagnetic property as it is observed in the case of 1% Fe-doped CuO.

Comparison of 1% Fe
The coordination number between Cu and O atoms in tenorite CuO (2+) and cuprite Cu 2 O (1+) is believed to play a crucial role in the observed difference in their magnetic properties with "Fe" doping. In monoclinic CuO, the Cu atoms are coordinated to four coplanar oxygen situated at the corners of an almost rectangular parallelogram [17] with two more distant apical "O" atoms. A distorted octahedron is formed because of large Jahn-teller effect. Depending on the valence state (1+, 2+ or 3+), a "Fe" ion has up to 5, 4 and "3d" electrons with a total spin of s = 2 1 2 , s = 2 and s = 1 1 2 , respectively, due to Hund's rule. A Cu ion has a total spin of 1/2 and hence "Fe" ions doped into CuO have extra moments which are coupled to the lattice of Cu spins via superexchange. CuO at 300 K is paramagnetic and has one unpaired spin due to the 3d 9 electronic configuration. The fraction of Fe 3+ ions substituted at the "Cu" sites in CuO has 5 unpaired electrons, which might also contribute to the observed magnetic property of "Fe"-doped CuO by forcing the "Fe" magnetic dipoles to align ferromagnetically by interacting with the unpaired spins of the CuO which is at paramagnetic state at 300 K. Extending the analogy in CuO : Mn [22], we suppose that a fraction of "d" electron states are available for hopping between the "Fe" sites and hence mediating ferromagnetism through the Hund's rule coupling. Hence we attribute the observed magnetic behavior of CuO : Fe to have a dominant hidden intrinsic nature apart from the segregated CuFe 2 O 4 phase whereas this is not in the case of Cu 2 O, where we have Cu + ions coordinated linearly to two "O" ions, due to its 3d 10 configuration. There are no unpaired spins and the spin value is s = 0. Hence on substituting with 1% Fe, irrespective of the charge state of "Fe", the Cu 2 O matrix could not mediate a ferromagnetic ordering through the substituted "Fe" ions. Hence, the spin configuration of the host matrix into which "Fe" is doped is also responsible for the observed magnetic behavior.

Conclusion
The XRD, TEM, and TGA results imply that the observed magnetic properties of "Fe"-doped CuO have contribution from the impurity phases, the nature of variation of H C and M max with increasing "Fe%", and the role of substitutional "Fe" and non-observation of ferromagnetic behavior in "Fe"-doped Cu 2 O invites exclusive investigation on CuO. Achieving ferromagnetic behavior from a paramagnetic/antiferromagnetic host by creating a spin imbalance seems to be relatively easier and advantageous than achieving it from diamagnetic oxides like ZnO, TiO 2 , and so on. The size dependent magnetic properties of "Fe"-doped CuO with controlled formation of CuFe 2 O 4 are another interesting aspect of this material which could be exploited for applications by suitably varying the size of the particles. The above said properties could be exploited for device applications provided if the "Fe" composition in CuO and the preparatory method and processing conditions are stabilized so as to result in a good single phase Cu-Fe-O solid solution with intrinsic magnetization and high reproducibility. Variation of H C in "Fe"-doped CuO, though CuFe 2 O 4 phases are present, extends from "soft" magnetic behaviour to "hard" with variation of "Fe%" in the samples studied. This wide variation in H C of such a DMS-like composite material could be thought of for fabricating MTJs with low H C -"free" layer and high H C -"pinned" layer separated by a thin insulating layer. The advantage of MTJ's made up of such all oxide DMS like layers with high M max , T N , and H C over the metallic Fe/AlO x /NiFe and CoFeB/AlO x /CoFeB type of MTJs is that these are less complicated and cheap and will have very good lattice matching and least strain between the "pinned" and "free" layers.