Nanostructure and magnetic properties of the MnZnO system, a room temperature magnetic semiconductor?

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/51465616

Nanostructure and magnetic properties of the MnZnO system, a room temperature magnetic semiconductor? Article in Nanotechnology · February 2005 DOI: 10.1088/0957-4484/16/2/006 · Source: PubMed

CITATIONS

READS

49

45

8 authors, including: José Luis Costa-Krämer

F. Briones

Spanish National Research Council

Spanish National Research Council

140 PUBLICATIONS 2,308 CITATIONS

223 PUBLICATIONS 2,774 CITATIONS

SEE PROFILE

SEE PROFILE

Jose F. Fernández

Marisel Diaz

Institute of Ceramics and Glass

Venezuelan Institute for Scientific Research

359 PUBLICATIONS 4,294 CITATIONS

46 PUBLICATIONS 324 CITATIONS

SEE PROFILE

All content following this page was uploaded by Jose F. Fernández on 20 January 2017.

The user has requested enhancement of the downloaded file.

SEE PROFILE

INSTITUTE OF PHYSICS PUBLISHING

NANOTECHNOLOGY

Nanotechnology 16 (2005) 214–218

doi:10.1088/0957-4484/16/2/006

Nanostructure and magnetic properties of the MnZnO system, a room temperature magnetic semiconductor? J L Costa-Kr¨amer1 , F Briones1 , J F Fern´andez2, A C Caballero2 , M Villegas2 , M D´ıaz3 , M A Garc´ıa4 and A Hernando4 1

Instituto de Microelectr´onica de Madrid, CNM-CSIC, Isaac Newton 8 PTM, 28760 Tres Cantos, Madrid, Spain 2 Instituto de Cer´amica y Vidrio, CSIC, 28049 Madrid, Spain 3 Centro de F´ısica, Instituto Venezolano de Invest. Cient., Apartado 21827, Caracas 1020A, Venezuela 4 Instituto de Magnetismo Aplicado ‘Salvador Velayos’ RENFE UCM Las Rozas, PO Box 155, 28230 Madrid, Spain

Received 30 September 2004, in final form 12 November 2004 Published 7 January 2005 Online at stacks.iop.org/Nano/16/214 Abstract The magnetic properties of the system MnZnO prepared by conventional ceramic procedures using ZnO and MnO2 starting powders are studied and related to the nanostructure. Thermal treatment at 500 ◦ C produces a ferromagnetic phase, although this temperature is not high enough to promote proper sintering; thus the thermally treated compact shows brittle characteristics of unreacted and poorly densified ceramic samples. Scanning electron microscopy and x-ray analysis reveal the appearance of a new phase, most probably related to the diffusion of Zn into MnO2 oxide nanocrystals. The magnetic properties deviate considerably from what would be expected of an unreacted mixture of ZnO (diamagnetic) and MnO2 particles (paramagnetic above 100 K and anti-ferromagnetic below that temperature), exhibiting a ferromagnetic like behaviour from 5 to 300 K and beyond mixed with a paramagnetic component. The ferromagnetic phase seems to be originated by diffusion at the nanoscale of Zn into MnO2 grains. The Curie temperature of the ferromagnetic phase, once the paramagnetic component has been subtracted from the hysteresis loops, is measured to be 450 K. EPR resonance experiments from 100 to 600 K confirm a ferromagnetic to paramagnetic like transition above room temperature for these materials.

1. Introduction In recent years there has been a growing interest in semiconductor materials that exhibit ferromagnetism above room temperature, RT [1–3]. These materials are key ingredients for the development of spintronic devices such as non-volatile memories, spin valve transistors, and ultrafast optical switches. Since Dietl et al [4] in 2000 predicted the existence of ferromagnetic semiconductors at RT, there have been several experimental works reporting the appearance of ferromagnetism above 300 K, like Mn:ZnO [5, 6], Co:ZnO [6, 7], Co:TiO2 [8, 9], GaMnN [10] GaMnP [11]. Among these systems, those based on ZnO are particularly 0957-4484/05/020214+05$30.00 © 2005 IOP Publishing Ltd

interesting due to the fact that ZnO is a wide band semiconductor and is transparent in the visible part of the spectrum. Sharma et al [5] have recently reported ferromagnetism at RT in the Mn:ZnO system both in pellets fabricated by mixing appropriate amounts of ZnO and MnO2 and in thin films grown by pulsed laser deposition using the pellets as targets. Although these reports show the appearance of a ferromagnetic phase, the mechanism responsible for the ferromagnetism still remains unclear. There are several proposals to account for this magnetic behaviour. For these materials to have application in spintronic devices it is essential that they are homogeneous, and that the magnetic properties do not arise from phase segregation.

Printed in the UK

214

Nanostructure and magnetic properties of the MnZnO system, a room temperature magnetic semiconductor?

MnO2

Mn2O3

ZnO

Intensity (a.u.)

500°C 12h milled disc

500°C 12h disc as sintered

400°C 8h

10

20

30

40

50

60

70

Figure 1. XRD spectra for a mixture of ZnO and MnO2 powders thermally treated at 400 ◦ C for 8 h, as thermally treated discs at 500 ◦ C 12 h, and powders obtained by grinding the same discs.

If the ferromagnetic properties arise from a segregated phase within the semiconducting matrix, this material would not have the desired properties for spintronic devices such as spin injection, etc. This demands a very thorough investigation of the nanostructure to ensure that the magnetic and the semiconducting properties are due to the same phase. In this work, the magnetic properties of MnZnO prepared by conventional ceramic procedures using ZnO and MnO2 starting powders are studied and correlated with the nanostructural characterization.

2. Experimental procedure Materials with different Mn concentrations were fabricated following the procedure of Sharma et al [5]. High purity (>99.99%) ZnO and MnO2 raw powders were used for sample preparation. In a first step, the powders were mixed and prereacted at 400 ◦ C for 8 h in air. After that, the calcined powders were attrition milled with zirconia balls in an ethanol medium, dried at 60 ◦ C overnight, pressed into discs at 90 MPa and exposed to a second thermal treatment at 500 ◦ C for 12 h in air. Two different sets of samples, (ZnO)1−x (MnO2 )x , with x = 2 and 10% were fabricated. X-ray diffraction (XRD) spectra were obtained on thermal treated discs and powder was obtained from these discs by softly grinding them in an agate mortar. The XRD spectra were recorded in a Siemens D5000 diffractometer with a scan rate of 1 2
◦ min−1 , by using the Cu Kα line. TEM analysis was carried out on powder samples. Magnetic properties were measured on discs thermally treated at 500 ◦ C for 12 h by using a SQUID up to RT and a high T VSM above RT. EPR experiments were performed at the X-band (n ∼ 9.5 GHz) with a Bruker EMX spectrometer in the temperature range between 100 and 600 K.

3. Results and discussion XRD spectra of the 90% ZnO–10% MnO2 sample are shown in figure 1. For the powders treated at 400 ◦ C for 8 h, no evidence of bulk reaction is found; only ZnO and MnO2 characteristic peaks are evidenced. The same result is observed for the discs treated at 500 ◦ C for 12 h when measured directly on the surface of the pellet. However, an interesting effect appears when

the sample is ground and measured again. In this case, the observed phases are ZnO and Mn2 O3 . The energy injected into the system by the grinding process of the pellet is very low, but it is enough to promote the MnO2 transformation. On the other hand, in a pure manganese oxide the Mn2 O3 phase is not expected to stabilize at such a low temperature. This indicates that there is a certain degree of interaction between MnO2 and ZnO, although no extensive reaction is observed in the sample. The TEM analysis of ZnO and MnO2 starting powders mixture (figure 2) reveals the presence of dense crystalline ZnO particles in contrast to MnO2 particles that show higher porosity and defects. However, for the disc thermally treated at 500 ◦ C 12 h in air, the morphology of the manganese oxide shows marked differences. No information on the Mnoxidation state can be extracted with this technique. Moreover, EDS analysis of the particle clearly shows the presence of Zn cations well inside the grain. The ratio Mn/Zn measured by EDS is 2.7, but there is a lack of homogeneity and not all the Mn particles show the same ratio. Meanwhile the spinel, ZnMn2 O4 , has a ratio of 2 for Mn3+ ; the EDS results show a higher proportion of Mn cations. In contrast, no diffusion of Mn cations into the bulk of ZnO particles was observed. Therefore diffusion of Zn2+ into the MnO2 lattice takes place, and its presence seems to promote the MnO2 transformation to Mn2 O3 observed in XRD measurements. This phenomenon might be favoured by the use of highly reactive particles, i.e. an excess of interfacial energy due to small particle sizes favours the onset of the diffusion mechanisms. In fact what is actually observed can be described as Zn-doped MnO2 or Mn–Zn–O phase instead of Mn-doped ZnO. Figure 3 shows the thermal dependence of the magnetization (under a field of 10 kOe) of the discs thermally treated at 500 ◦ C for 12 h. A sample made of 100% MnO2 subjected to the same thermal treatment is also shown for comparison. This 100% MnO2 sample exhibits the typical curve for an antiferromagnetic material with a N´eel transition temperature to a paramagnetic phase of 85 K, in agreement with previous studies [12]. Above this temperature, the fieldinduced magnetization follows the Curie–Weiss law with C = 3.74×10−2 emu kg−1 Oe−1 and θ = −160 K. For the samples containing 2% and 10% MnO2 the value of the magnetization is 215

J L Costa-Kr¨amer et al

Figure 2. Microstructural analysis by TEM of (a) starting mixture of ZnO and MnO2 , and (b) discs treated in a second thermal treatment at 500 ◦ C 12 h in air. The inset shows the EDS spectra obtained at the MnO2 particle showing the presence of Zn cations and at the ZnO particles without Mn cations. 1,5

2%

2%

1,4

10% 100%

10%

2

1,2

M (emu/gMnO )

M (emu/gMnO2)

1,3

1

1,1 1,0 0,9 0,8

0

-1

0,7 0,6 0,5

-8000

0

8000

H (Oe) 0,4 0

50

100

150

200

250

300

T (K)

Figure 3. The thermal dependence of the magnetization (after ZFC) under an applied field of 10 kOe for 100% MnO2 powders and mixtures (ZnO)1−x (MnO2 )x , with x = 10 and 2% thermally treated at 500 ◦ C for 12 h in air.

lower (partially due to the presence of diamagnetic ZnO phase), and a significant increase of the magnetization for temperatures below 50 K is observed in the curve. Additionally, the sample with 2% MnO2 shows two maxima at 65 and 80 K. The loops from the MnO2 –ZnO samples exhibit a ferromagnetic component superimposed on a paramagnetic one. This is shown in figures 4 and 5. Any contribution to the ferromagnetic phase from the pure MnO2 phase can be discarded, as the loops from pure MnO2 show no hysteresis in the whole temperature range, from 5 to 300 K. The coercive 216

Figure 4. Hysteresis loops for MnO2 (2 and 10%)–ZnO samples at 5 K.

field for the 2% MnO2 sample is about 180 Oe at 5 K, decreasing up to 60 Oe at 300 K, and being about half for the 10% MnO2 sample. The relative presence of ferromagnetic phase is larger for the 2% MnO2 sample than for 10% MnO2 one. The value of the saturation magnetization, MS , at 300 K is 2.9 × 10−2 emu g−1 (MnO2 ) and 3.6 × 10−3 emu g−1 (MnO2 ) for the 2% and 10% MnO2 samples respectively. The ferromagnetic (FM) phase presents a Tc at about 450 K, (see figure 6), in agreement with results previously reported [5] for MnO2 –ZnO pellets. This is, however, the first time to our knowledge that a value of the Tc is measured, and not given based on an extrapolation from below room temperature data. The coercive field and MS values are very similar to those reported in [5].

Nanostructure and magnetic properties of the MnZnO system, a room temperature magnetic semiconductor?

2%

10%

10%

) 2

M (emu/gMnO

2

)

0,4

M (emu/gMnO

2%

0,03

0,0

0,00

-0,4 -0,03 -8000

0

-8000

8000

0

H (Oe)

8000

H (Oe)

Figure 5. Left: hysteresis loops of MnO2 (2 and 10%)–ZnO samples at 300 K. Right: ferromagnetic component of the loop after subtracting the paramagnetic contribution. 0,04

3300

0,01

3200

1600

∆Hpp (Gauss)

0,02

HR (Gauss)

M (emu/gMnO

2

)

0,03

3100 0,00 100

150

200

250

300

350

400

450

1200

800 400

T (K) 100

Figure 6. The thermal dependence of the saturation magnetization (after ZFC) under an applied field of 10 kOe above 100 K from the ferromagnetic phase (obtained from the 2% MnO2 sample), showing a Tc at about 450 K. Data from SQUID below 300 K and from high T VSM above.

Hence we found that ZnO–MnO2 samples exhibit two additional magnetic features with respect to pure MnO2 samples that have been subjected to the same treatment: the appearance of an FM phase up to room temperature and a sharp decrease of magnetization as the temperature increases from 5 to 50 K. These results point out that the phase responsible for this ferromagnetic behaviour is associated with the interaction of MnO2 with ZnO. However, TEM measurements confirmed that after annealing the discs in air at 500 ◦ C for 12 h the only reaction between both phases was the diffusion of Zn into MnO2 grains. Thus, this region with an Mn–Zn–O compound is the origin of the FM phase and the low temperature magnetization increase, whereas the paramagnetic component is mainly due to unreacted MnO2 . It is evident that this ternary phase must be Mn-rich, and it cannot be considered as Mn-doped ZnO but more likely a Zn–Mn spinel. Actually, ZnMn2 O4 has been reported to be ferromagnetic, with TC greatly above room temperature [1]. One experimental fact that also supports this idea is that using MnO instead of MnO2 powders resulted in a totally different magnetic behaviour of the reacted material, with a ferromagnetic to paramagnetic transition at about 45 K, as will be shown below. EPR studies in the same materials confirm these results. A single EPR line is observed in the whole temperature range. Figure 7 shows

3000 100

200

300

200

300 400 T (K)

400

500

500

600

600

T (K)

Figure 7. The temperature dependence of the EPR resonance field (HR) and the line width (DHpp) for the 2% MnO2 sample. The curves are guides to the eye.

the temperature dependence of the resonance field (HR) and the linewidth (DHpp) for the 2% MnO2 sample. Above ∼350 K the HR value is approximately constant (g ∼ 2.09), as expected for a paramagnetic material. Below ∼350 K a decrease of HR is observed. At the same time, the line width is approximately constant as the temperature is reduced to ∼450 K, below which it shows a increase. The HR and DHpp behaviour between 100 and ∼400 K is typical of ferromagnetic materials, and a Tc at around 400 K could be estimated in this case. For the 10% MnO2 sample similar EPR results are obtained. This confirms the appearance of a ferromagnetic phase with Tc above RT when starting with MnO2 powders. For the 10% MnO sample [13] the Tc value is below RT, and a g = 2.00 in the paramagnetic phase is obtained. A different set of samples prepared with starting powders having smaller grain size for the ZnO (average particles size of 0.5 µm) have been also processed by the same procedure and then measured. The magnetic features were similar, but the relative presence of the FM phase increased with decreasing grain size of the starting MnO2 powders, supporting the idea of Zn diffusion in the MnO2 grains as the origin (decreasing the grain size increases the relative surface area) for this phase. 217

J L Costa-Kr¨amer et al 0,18 0,20 0,15

0,16

0,02

Loop at 5K

0,14

m (e m u )

0,12

Loop at 300K

0,01

0,05

m (emu)

m (emu)

0,10

0,00 -0,05

0,00 -0,01

-0,10 -0,02

-0,15

0,10

-10000 -5000

0

5000

10000 -10000 -5000

H (G)

0

5000

10000

H (G)

0,08 0,06 0,04 0,02 0,00 0

50

100

150

200

250

300

T (K )

Figure 8. The thermal dependence of the magnetization (after ZFC) under an applied field of 10 kOe for (ZnO)1−x (MnO)x , with x = 10%. Notice the different behaviour compared to the same sample but fabricated with MnO2 instead—(ZnO)1−x (MnO2 )x with x = 10% shown in figure 3. The insets show the ±1 T loop at 5 K (left) and the ±1 T loop at 300 K (right).

Furthermore, the larger fraction of FM phase in the 2% MnO2 with respect to that of 10% could be associated with a larger relative area in the former, as the MnO2 grains will be more disperse in the ZnO matrix. From the loops at 300 K it is found that the sample with 10% MnO2 (that shows a lower relative FM signal than the 2% MnO2 sample) exhibits a larger paramagnetic susceptibility than that with 2%, indicating that the appearance of the FM phase is due to the partial transformation of MnO2 because of Zn diffusion into the grain. However, some variations in the susceptibility due to the partial transformation MnO2 → Mn2 O3 , evidenced by XRD, cannot be discarded. Regarding the increase of the magnetization at low temperatures, from the loops measured at different T , we found that this is due to an enhancement of both the FM phase and the PM one. The FM phase at 5 K is between one and two orders of magnitude larger than that at 300 K, but remains almost constant for T above 100 K (as shown in figure 6). This behaviour could be ascribed to an ordering temperature of the FM phase very dependent on the detailed Zn x Mn y Oz phase. As a matter of fact the ferromagnetic ordering temperatures reported in [1] are 1298 K (x = 1, y = 2, z = 4), and for low oxygen content they are very sensitive to composition: 45 K (x = 0.7, y = 0.3, z = 1) and 30 K (x = 0.9, y = 0.1, z = 1). The composition fluctuations expected from a diffusion process would produce a distribution of ordering temperatures that gives rise to a decrease of the ferromagnetic phase (and the magnetization) as the temperature increases. The ferromagnetic phase with Tc about 450 K should be related to higher oxygen content grains. This is also supported by the low Tc results obtained using MnO as a starting powder mentioned above. The magnetization versus T dependence shown in figure 8 shows a clear ferromagnetic to paramagnetic transition at about 45 K. Below that temperature a hysteresis loop is observed, while above that temperature, a linear reversible dependence of the magnetization versus field is obtained (see the insets of figure 8). This precursor has less oxygen than MnO2 , and thus should favour the formation of a low oxygen content spinel with low Tc . This hypothesis would also explain why at 5 K the susceptibility of the 2% 218

View publication stats

MnO2 sample is somewhat larger than that of the 10% MnO2 one, whereas at 300 K they are almost identical; this cannot be understood assuming a unique paramagnetic phase. Therefore, we suggest that the ferromagnetic phase observed in the samples is ascribed to an Mn–Zn–O FM phase formed at MnO2 grains by diffusion of Zn during the annealing process. The nanometric grains of this phase exhibit a broad stoichiometry spectrum, and consequently a distribution of ferromagnetic phases with different Tc .

4. Conclusions Ferromagnetism below Tc ∼ 450 K has been observed in samples prepared by conventional ceramic routes, using ZnO and MnO2 as starting powders. Our structural analysis points to a non-homogeneous material, where the ferromagnetic phase seems to be originated by diffusion at the nanoscale of Zn into the MnO2 grains.

References [1] Pearton S J, Hep W H, Ivill M, Norton D P and Steiner T 2004 Semicond. Sci. Technol. 19 R59–74 [2] Pearton S J, Abernathy C R, Norton D P, Hebard A F, Park Y D, Boatner L A and Budai J D 2003 Mater. Sci. Eng. R 40 137–68 [3] Coey J M D and Sanvito S 2004 J. Phys. D: Appl. Phys. 37 988–93 [4] Dietl T, Ohno H, Matsukura F, Cibert J and Ferrand D 2000 Science 287 1019–22 [5] Sharma P, Gupta A, Rao K V, Owens F J, Sharma R, Ahuja R, Guill´en O, Johansson B and Gehring A 2003 Nat. Mater. 21 673–7 [6] Theodoropoulou N A et al 2003 Solid State Electron. 47 2231–5 [7] Rode K et al 2003 J. Appl. Phys. 93 7676 [8] Kim J Y et al 2003 Phys. Rev. Lett. 90 017401 [9] Shinde S R et al 2003 Phys. Rev. B 67 115211 [10] Sonoda S et al 2002 J. Cryst. Growth 237–239 1358 [11] Theodoropoulus N et al 2002 Phys. Rev. Lett. 89 107203 [12] Tebble R and Craik D J 1969 Magnetic Materials (London: Wiley) [13] D´ıaz M et al 2004 unpublished