transition metals in zno nanocrystals – magnetic …...i. kuryliszyn-kudelska et al. /science of...

Science of Sintering, 45 (2013) 31-48 ________________________________________________________________________

_____________________________

*) Corresponding author: [email protected]

doi: 10.2298/SOS1301031K UDK 621.315.59;676.017.5 Transition Metals in ZnO Nanocrystals – Magnetic and Structural Properties I. Kuryliszyn-Kudelska1,W. Dobrowolski1,M. Arciszewska1,N. Romčević2*), M. Romčević2,B. Hadžić2,D. Sibera3,U. Narkiewicz3,W. Lojkowski4

1Institute of Physics, Polish Academy of Sciences, Al. Lotników 32/46, 02-668 Warsaw, Poland 2 Institute of Physics, Belgrade University, Pregrevica 118, 11080 Belgrade, Serbia 3 West Pomeranian University of Technology, Institute of Chemical and Environment Engineering, Pulaskiego 10, 70-322 Szczecin, Poland 4Institute of High Pressure Physics, Polish Academy of Sciences, Sokołowska 29/37, 01-142 Warsaw, Poland Abstract:

Currently, wide-gap ZnO nanoparticles bear important potential application in electro-optical devices, transparent ultraviolet protection films, and spintronic devices. We have studied the magnetic properties of nanocrystals of ZnO(Fe, Co, Mn) prepared by two methods of synthesis. We have used the microwave assisted hydrothermal synthesis and traditional wet chemistry method followed by calcination. The detailed structural characterization was performed by means of X-ray diffraction and micro-Raman spectroscopy measurements. The morphology of the samples was studied by means of SEM and TEM microscopy. The results of systematic measurements of AC magnetic susceptibility as a function of temperature and frequency as well as SQUID magnetization are presented. The SQUID magnetization measurements revealed a clear bifurcation of the FC and ZFC plots. Such behavior suggested superparamagnetic behavior above the blocking temperature. The dynamic magnetic measurements were performed at small AC magnetic field with amplitude not exceeding 5 Oe and different frequency values (from 7 Hz to 9970 Hz). For ZnO(Fe) and ZnO(Mn), the AC susceptibility maxima has been found for in-phase susceptibility Re(χ) and for out of phase susceptibility Im(χ). We analyzed the observed frequency dependence of the peak temperature in the AC susceptibility curve using the empirical parameter Φ that is a quantitative measure of the frequency shift and is given by the relative shift of the peak temperature per decade shift in frequency, as well as Vogel-Fulcher law. We observed two different types of magnetic behavior, spin-glasslike behavior or superparamagnetic behavior, depending on the synthesis process. For ZnO(Co) nanocrystalline samples high temperature Curie-Weiss behavior in AC magnetic susceptibility was observed. We observed that the determined negative values of the Curie-Weiss temperature θ depend strongly on the nominal content of cobalt oxide. It was shown that for calcination method the values of θ increase with the increase of magnetic ion content indicating enhancement of predominance of antiferromagnetic interactions. For hydrothermal method the opposite effect was observed indicating the breakdown of predominance of aniferromagnetic coupling with the increase of nominal magnetic ion content.

I. Kuryliszyn-Kudelska et al. /Science of Sintering, 45 (2013) 31-48 ___________________________________________________________________________

32

This paper gives an in-depth discussion of the structural and magnetic properties of ZnO nanocrystals in addition to the technological issues such as different methods of wet chemical synthesis. Key words: Semiconductors, Magnetization, Magnetic measurements, Absorption and light scattering. 1. Introduction

Nanocrystalline materials derive their interesting and technologically useful properties with respect to bulk mainly from quantum confinement effects and the increasing role of surface atoms with the decrease in particle size. Magnetic nanocrystals are currently in the focus of activity of a wide scientific community: physics, chemistry, biology, medicine [1,2].

Superparamagnetic nanoparticles based on a core consisting iron oxides are very promising for in vivo applications (e.g. hypethermia and radiotherapy in vivo). In particular, there is an urgent need to increase the efficacy of cancer therapy by overcoming several limitations of conventional hyperthermia and chemotherapy. Composite nanoparticles based on magnetic oxides can be loaded with cytotoxic drugs, dispersed in a carrierfluid, injected into the bloodstream, and targeted to the tumor using an external magnetic field gradient. These particles release the drug following localization in the tumor, resulting in increased tumor drug concentration with minimal systemic toxicity ([3] and references therein). In addition, superparamagnetic nanoparticles are biocompatible and are in routine clinical use. Nanosized magnetic oxides are of particular interest due to the potential commercial applications [see e.g. 4-7]. There are several reasons why wide-gap oxides were chosen in our studies. The intense interest in dilute magnetic oxides is mainly driven by the search for new device applications in spin-based technologies. Unlike the diluted magnetic semiconductors based on III–V or II–VI group of elements showing ferromagnetism only at very low temperatures, these oxide-based DMS’s exhibit ferromagnetism at higher temperatures and even above RT. They are also optically transparent making them suitable for magneto–optoelectronic applications, particularly for electrodes in solar cells. However, the magnetic properties of these compounds are still under debate.

Transition metal (TM) doped ZnO received recently much attention [8]. However, the literature shows reports claiming for intrinsic ferromagnetism [9], paramagnetism [8], extrinsic ferromagnetism [9] or spin-glass behavior [10] in ZnO:TM . Recently, it was shown that the ferromagnetism in these materials can be induced by inclusions of nanoscale oxides of transitions metals [11] and/or nanoparticles containing a large concentration of magnetic ions [12]. Novel methods enabling a control of nanoassembling of magnetic nanocrystals in nonconducting matrices as well as functionalities specific to such a system were described [12]. Grain boundary defects have been also proposed to play role in facilitating magnetic ordering of oxide diluted magnetic semiconductors (see e.g. [13]). Still the reports to date have been contradictory. Despite of a number of available reports and experimentally observed evidences for ferromagnetism of ZnO:TM, there is still an open question about the origin of ferromagnetism in these compounds. That is due to the poor reproducibility by using different preparation methods or the poor resolutions of the commonly used structural characterization methods. Further intensive studies are necessary, particularly discriminating materials characterization could provide insight on the understanding of the magnetic properties in TM doped ZnO.

At present there are many studies reporting the preparation and experimental studies of TM-doped nanoparticulate ZnO compounds (see eg. [4, 8, 14-22]). The present review gives an in-depth discussion of the structural and magnetic properties of

D. Kuryliszyn-Kudelska et al./Science of Sintering, 45 (2013) 31-48 ___________________________________________________________________________

33

ZnO nanocrystals in addition to the technological issues such as different methods of wet chemical synthesis. Our aim was to examine the effect of ZnO nanocrystals transition metal (TM) doping on the magnetic properties of the resultant nanosized material. The systematic magnetic studies including AC magnetic susceptibility measurements and DC SQUID magnetometry were carried out. The micro-Raman spectroscopy measurements were performed. The samples were characterized by means of X-ray diffraction, scanning electron microscopy and transmission electron microscopy. 2. Experimental results and discussion 2.1. Structural characterization

The nanocrystalline samples of ZnO doped with Fe2O3, CoO and MnO were synthesized by use of two methods [23-26]. In the copreciptation-calcination method, a mixture of transition metal hydroxides and zinc hydroxides from an aqueous solution of nitrites were obtained. The obtained hydroxides were filtered, dried at the temperature of 700C and calcined at 3000C during 1 hour. In the hydrothermal synthesis the obtained hydroxides were put in the reactor with microwave emission. The microwave assisted synthesis was conducted under a pressure of 3.8 MPa during 15 min. The synthesized product was filtred and dried. These two methods allowed to obtain the series of nanosized samples with nominal concentration of Fe2O3 from 5 to 70 wt.%, CoO from 5 to 80 wt.% (for calcination method) and from 5wt.% to 50 wt.% (for hydrothermal method), MnO from 5 to 95 wt.% in calcination method.

The detailed characterization studies including X-ray diffraction (XRD) measurements, micro-Raman and low-frequency Raman spectroscopy investigations, scanning electron microscopy, transmission electron microscopy, specific surface area measurements were performed [23-28].

30 40 50 60 70 80

5 wt.% 10 wt.% 20 wt.% 30 wt.% 40 wt.% 50 wt.% 60 wt.% 70 wt.%

Inte

nsity

[arb

. uni

ts]

2θ [deg.]

-ZnO ZnO:Fe2O3calcinationmethod

a)

30 40 50 60 70 80

5 wt.% 10 wt.% 20 wt.% 30 wt.% 40 wt.% 50 wt.% 60 wt.% 70 wt.%

Inte

nsity

[arb

. un.

]

2θ [deg.]

-ZnO ZnO: Fe2O3hydrothermalmethod

b)

Fig.1. The selected XRD spectra for several nanocrystalline samples of ZnO doped with Fe2O3 prepared by calcination method (a) and hydrothermal synthesis (b). The characteristic

peaks attributed to ZnO (squares) are marked. The not marked peaks are attributed to ZnFe2O4.

For ZnO doped with Fe2O3 the detailed structural investigations revealed the presence

of crystalline phases of hexagonal ZnO and cubic ZnFe2O4 in both mehods. The results of XRD studies are shown in Fig. 1. The mean crystallite size of the detected phases was determined using Scherrer formula [29]. In particular, the mean crystallite size of ZnFe2O4 varied from 8 to 12 nm in both methods [23]. In the samples prepared using hydrothermal


34

synthesis the amount of spinel phase is lower comparing with calcination method. For example in the XRD spectrum of the sample containing 50 wt.% Fe2O3 in the hydrothermally prepared sample (Fig. 1b) the clear peaks attributed to ZnO phase are visible and in case of the same sample prepared by calcination (Fig. 1a) the peaks of ZnO phase are much less distinguishable.

0 400 800 1200 1600

647,

A1g

451,

F2g

(3)

355,

F2g

(2)

246,

Eg

221,

F2g

(1)

ZnFe2O4

ZnO1153

660

592,

E1(L

O)

577,

A1(L

O)

437,

E2(2

)41

0, E

1(TO

)37

9, A

1(TO

)33

0, 2

-ph

102,

E2(1

)

380

3251148

436

571665

Inte

nsity

[arb

. un.

]

Raman shift [cm-1]

ZnO:Fe2O35 wt.%calcination method

T=300 Kλ=514.5 m

Fig. 2. Raman spectrum for ZnO:Fe2O3 wit 5 wt.% prepared by calcination method.

Raman spectroscopy measurements allowed to determine the detailed structural

characterization [28, 30]. The advantage of micro-Raman technique is the very accurate chemical phases determination. Raman spectroscopy has already proven to be a unique tool for probing nanophases dispersed in a matrix (see eg. [31]). The micro-Raman spectra were taken in the backscattering configuration and analyzed using a Jobin Yvon T64000 spectrometer, equipped with nitrogen cooled charge-coupled-device detector. As an excitation source we used the 514.5 nm line of an Ar-ion laser. The exemplary Raman spectrum is shown in Fig. 2. For ZnO: Fe2O3 samples, the main characteristics of experimental Raman spectrum in 200 to 1600cm-1 spectral region are sharp peak at 436 cm-1 and broad two-phonon structure at ~1150 cm-1, typical for ZnO. Raman spectrum in region below 400 cm−1 is superposition of characteristic frequencies of ZnO phase and spectrum of ZnFe2O4 nanoparticles. Broad structure below 700 cm-1 corresponds to ZnFe2O4 nanopartilcles.

XRD analysis did not reveal the presence of ZnO phase in samples doped with high content of magnetic dopants. Despite the fact that XRD does not evident ZnO sharp peak at 436 cm-1 was clearly E2 (2) mode of ZnO. This peak is typical for undoped ZnO nanoparticles [32].

A magnetic resonance study performed with an electron paramagnetic resonance spectrometer was carried out at room temperature for the calcinated samples [27]. A very intense magnetic resonance line with asymmetry depending on the iron oxide concentration was recorded for all samples. The resonance line is centred at g = 2.005(2) and its integrated intensity increases with the ferrite content, reaching a maximum for a sample with a nominal 70 wt.% content of iron oxide. The presence of iron in the form of ZnFe2O4 was confirmed in all the investigated in present paper calcinated samples [27].

The morphology was studied by use of SEM and HRTEM. The investigations allowed to distinguish two morphologies – spherical and hexagonal nanograins are observed. Figs. 3 and 4 show selected SEM and TEM images. We assume the hexagonal crystals correspond to the ZnO, while the small spheroidal crystals — to the ZnFe2O4 spinel. It was shown that the degree of agglomeration of nanocrystals depends on the amount of Fe2O3 and decreases with increasing nominal Fe2O3 content [23]. In addition SEM measurements


35

showed that for hydrothermal process the tendency to the formation of agglomerates is less in comparison to the calcination process. The formation of agglomerates is minimized because during the hydrothermal process elevated under pressure of 3.8 MPa the synthesis temperature is only 2500 C and reaction time is short: 15 min. Due to the agglomeration process the size of nanocrystallites determined from TEM is much smaller in comparison to the SEM studies. The size of nanocrystallites determined from TEM is in agreement with the results of XRD studies.

a) b)

c) d)

e) f)

Fig. 3. SEM images of ZnO nanocrystalline samples doped in calcination process: (a) 10 wt.% Fe2O3, (c) 30 wt.%, (e) 70 wt.% and in hydtothermal process: (b) 10 wt.% Fe2O3, (d) 30

wt.%, (f) 70 wt.%.

For ZnO doped wih CoO, the crystalline phases of hexagonal ZnO and cubic Co3O4 for calcinated samples, hexagonal ZnO and cubic ZnCo2O4 for hydrothermal method were determined. The detailed results of structural investigations will be presented elsewhere. The selected results of XRD are shown in Fig. 5. The mean crystallite size of Co3O4 varied from 55 nm to 21 nm with the increase of CoO content. For calcination process the mean


36

crystalline size varied from 39 nm to 49 nm with the increase of CoO content.

a)

b)

c)

d)

Fig. 4. HRTEM images images of ZnO nanocrystalline samples doped in calcination process: (a) 10 wt.% Fe2O3, (c) 70 wt.% and in hydrothermal process: (b) 10 wt.% Fe2O3, (d) 70 wt.%.

20 30 40 50 60 70

Inte

nsity

[arb

. un.

]

2θ [deg]

40 wt. % 60 wt. %

20 wt. % 5 wt. %

80 wt. %

ZnO:CoOcalcination method

a)

20 30 40 50 60 70

Inte

nsity

[arb

. un.

]

2θ [deg]

40 wt. % 60 wt. %

20 wt. % 5 wt. %

80 wt. %


b) Fig. 5. The selected XRD patterns of ZnO doped CoO in calcination and hydrothemal processes. Peaks attributed to ZnO are marked as squares. The not marked peaks are

attributed to Co3O4 in calcination process (a) and to ZnCo2O4 (b) in hydrothermal method.

Raman spectroscopy measurements confirmed XRD results. The detailed micro-Raman characterization will be shown elsewhere. Fig. 6 a and 6 b show micro-Raman spectra for selected samples. For all samples Raman peak at 436 cm-1 was observed. This peak is


37

clearly E2 (2) mode of ZnO nanoparticles. Modes observed at ~ 204, 330, 379, 410, 541, 577, 592, 660, 1153 are also typical for ZnO. For the samples of ZnO doped with CoO by calcinations method, the Raman spectrum showed a band at ~ 691 cm-1. The presence of this band is typical for Co3O4. We also observed bands at ~ 194, 482, 521 and 618 cm-1, which are also typical for Co3O4. For samples prepared by hydrothermal method bands at ~ 185, 475, 520, 610, 690 cm-1 typical for ZnCo2O4 were observed.

a) b) Fig. 6a. The Raman spectra of selected nanocrystalline ZnO:CoO samples prepared by (a)

calcination method, (b) hydrothermal process.

The morphology of ZnO:CoO nanocrystalline samples were studied using SEM. The selected results are shown in Fig. 7. Two morphologies can be distinguished — spherical and hexagonal nanograins are observed. We assume the hexagonal nanocrystals correspond to the ZnO, while the spheroidal nanocrystals — to the spinel phases (Co3O4 for calcinations process and ZnCo2O4 for hydrothermal method). The obtained nanocrystalites size dependences from XRD measurements are clearly visible in SEM images.

a) b)

Fig. 7. The selected SEM images of ZnO:CoO nanopowders (30 wt.%) prepared in

calcination process (a) and hydrothermal process (b). For ZnO doped with MnO, XRD measurements revealed the presence of the

crystalline phases of hexagonal ZnO, and/or cubic ZnMnO3, and/or cubic Mn3O4, and/or cubic ZnMn2O4.

400 800 1200 1600

T=300 Kλ=514.5 nm20 mW

ZnO:CoO5 wt.%hydrothermal method

Inte

nsity

[arb

. un.

]

Raman shift [cm-1]200 400 600 800 1000 1200 1400 1600

Inte

nsity

[arb

. un.

]

Raman shift [cm-1]

ZnO:CoO5 wt.%calcination method

T=300 Kλ=514.5 nm20 mW


38

20 30 40 50 60 70 80

3 0 3 5 4 0

6 0 w t .%

7 0 w t.%

8 0 w t .%

9 0 w t .%

2 θ [d e g ]

9 5 w t .%

50 wt.%

40 wt.%

20 wt.% Inte

nsity

[arb

. un.

]

2θ [deg.]

5 wt.%

ZnO: MnOcalcinationmethod

Fig. 8. The selected spectra for several nanocrystalline samples of ZnO doped with MnO. The characteristic peaks attributed to ZnO (arrows) and ZnMnO3 (solid lines) are marked. The inset shows the characteristic peaks attributed to Mn3O4 (diamonds)

and ZnMn2O4 (dashed lines) for selected samples.

2 0 0 4 0 0 6 0 0 8 0 0 1 0 0 0 1 2 0 0 1 4 0 0 1 6 0 0

R a m a n sh ift [cm -1 ]

5 w t.%

Z

ZZ

Z

Z

M

Inte

nsity

[arb

. un.

]

3 0 w t.%

Z ZZ

Z

Z

S

T +P

7 0 w t.%M T P

Z ,T ,P

Z - Z n OM - M n OT - M n3O 4P - Z nM n2O 4S - Z n M n O 3

Z n O :M n Oca lc in a tio nm e th o d

Z ,T ,P

9 5 w t.%M

T +P

P P

Fig. 9. The Raman spectra of selected nanocrystalline ZnO:MnO samples prepared by

calcination method.

Fig. 8 presents the XRD patterns of ZnO doped with MnO demonstrating the presence of spinel, manganium oxide and zinc oxide phases in the system. In samples with low nominal content of MnO two phases: ZnO and ZnMnO3 are observed. With increasing MnO concentration the content of ZnO phase decreases and the peak attributed to this phase disappears in sample containing 60 wt.% of MnO. The peak of ZnO disappears, because all zinc is now bonded in the spinel phase. In samples containing 60.wt% of MnO and more, the


39

phases of ZnMn2O4 and Mn3O4 are found. The detailed results of XRD analysis are presented in [25]. In particular the deconvolution of a single XRD line for sample with high content of MnO content is shown. XRD method was applied to determine a mean crystallite size in prepared samples, using the Scherrer formula. The results of XRD analysis and micro-Raman spectroscopy measurements are compared in Table I. The micro-Raman measurements were performed for selected samples (see Fig. 9). The Raman peak at 660 cm-1 is observed. This peak is typical for spinel structure. In these samples, especially in the case of high doping level, bands from ZnMnO3 and Mn3O4 crystal structures were identified at about 315 and 525 cm-1. For samples with high content of magnetic dopant, the Raman measurements revealed the presence of MnO crystalline phase. It should be stressed that XRD analysis did not show the presence of this crystalline phase. Tab. I.The results of XRD and micro-Raman measurements for ZnO:MnO prepared by calcination method. The determined crystalline phases are denoted by “+”. The mean crystallite size d was determined by use of Scherrer’s formula for selected samples. The micro-Raman measurements were performed for selected samples (denoted as “+”). The XRD did not reveal the presence of ZnO phases for samples with high content of magnetic dopant.

ZnO phase

ZnMnO3 phase

ZnMn2O4 phase

Mn3O4 phase

MnO phase

ZnO:MnO

Nominal

Conc. XRD

d [nm]

Raman

XRDd

[nm]

Raman

XRDd

[nm]

Raman

XRDd

[nm]

Raman

XRD d

[nm]

Raman

5 wt.% + + 13 + ⎯ ⎯ ⎯ ⎯ ⎯ ⎯ 10 wt.% + 9 ⎯ ⎯ ⎯ 20 wt.% + + 10 + ⎯ ⎯ ⎯ ⎯ ⎯ ⎯ 30 wt.% 100 + 9 + ⎯ ⎯ ⎯ ⎯ ⎯ ⎯ 40 wt.% 100 + ⎯ ⎯ ⎯ 50 wt.% 100 + + ⎯ ⎯ 60 wt.% ⎯ ⎯ ⎯ ⎯ + + 24 + ⎯ + 70 wt.% ⎯ ⎯ ⎯ ⎯ ⎯ + 37 + ⎯ + 80 wt.% ⎯ ⎯ ⎯ ⎯ ⎯ + 45 + ⎯ + 90 wt.% ⎯ ⎯ ⎯ ⎯ ⎯ + 47 ⎯ + 95 wt.% ⎯ ⎯ ⎯ ⎯ ⎯ + 43 + ⎯ +

The magnetic resonance investigations have been carried out at room temperature

[25]. Slightly asymmetric, broad and intense magnetic resonance line is recorded for all studied samples. It was shown that the phases ZnO and ZnMnO3 dominate for samples with low content of nominal dopant, below 60 wt. %. For higher concentrations of the phases Mn3O4 and ZnMn2O4 are recorded. The EPR measurements of samples with nominal MnO content below 60 wt.% have shown that the resonance line is centered above geff= 2 and is attributed to ZnMnO3 phase. The integrated intensity increases with increasing MnO content. Above 60 wt. % the recorded EPR line is centered below geff = 2 and it could be attributed to ZnMn2O4 phase. The morphology of the samples was studied by use of TEM. The selected TEM images are show in Fig. 10. For samples with low content of MnO the hexagonal and


40

spherical nanograins were observed. We assume that hexagonal nanograins correspond to ZnO and spherical nanograins correspond to ZnMnO3. For samples with higher content of MnO mainly spherical nanograins were observed. These results corroborate with XRD and Raman spectroscopy studies.

a) b) Fig. 10. The selected TEM images of ZnO:MnO prepared by calcination method a) 5 wt.%, b)

40 wt.%. 2. 2. Magnetic properties 2.1.1. Magnetic properties of ZnO:Fe2O3

The dynamic magnetic properties were studied by means of AC susceptibility χ. The real, Re(χ), as well as imaginary, Im(χ) parts of magnetic susceptibility were measured by use of a mutual inductance method in an AC magnetic field of frequency (f), range 7-10 000 Hz and amplitude (HAC) not exceeding 5 Oe. The measurements were performed in the temperature from 4.5 to 160 K. Fig. 11 shows the temperature dependence of the real part of the AC susceptibility for ZnO:Fe2O3 samples prepared by calcination method (a) and hydrothermal process (b) in the wide range of magnetic dopant, measured at f=625 Hz .

50 100 1500.00

0.01

0.02

0.03

Re(

χ) [e

mu/

g]

Temperature [K]

5% 10% 30% 40% 50% 60% 70%

ZnO:Fe2O3 calcination method

HAC=2 Oef=625 Hz

50 100 1500,000

0,005

0,010

Re(

χ) [e

mu/

g]

Temperature [K]

5% 10% 20% 30% 40% 60% 70%

ZnO: Fe2O3 hydrothermal synthesis

HAC=2 Oef=625 Hz

a) b)

Fig. 11. The temperature dependence of Re(χ) for selected ZnO: Fe2O3 nanocrystalline samples a) (calcination process), b) (hydrothermal method).

The Re(χ) curves show pronounced maxima. The above described experimental

feature can be observed in both the superparamagnetic system and spin-glasslike systems. It is


41

clearly visible that the temperature maxima Tf shift towards higher temperatures with the nominal Fe2O3 concentration. It should be stressed that the mean crystalline size d for magnetic crystalline phase (ZnFe2O4) varied in the narrow range from 8 nm to 12 nm and should not be responsible for the observed distinct increase in Tf .

The most common method to distinguish between superparamagnetic and spin-glasslike systems is analysis of Tf(f) behavior. It was shown that useful criterion for classifying the observed freezing/blocking process is the empirical parameter Φ [33].

Φ = ΔTf / Tf Δlog10(f) (1), where ΔTf is the difference between the peak temperature measured in the Δlog10(f), and f is the AC magnetic field frequency.

We performed the measurements of AC susceptibility as a function of frequency f for selected samples synthesized by two methods. For all samples we observed that the positions of the maxima (Tf) shift towards higher values with increasing driving frequency (see Fig. 12). The observed frequency-dependent behavior in Re(χ) data may be attributed to the blocking process of superparamagnetism or freezing process in spin-glass systems. It was shown that for interacting nanoparticles, superparamagnetic systems exhibit values of Φ between 0.05 and 0.10, in spin-glass systems Φ < 0.05 [34-36].

0 10 20 30 40 50 60 70 80 900,000

0,001

0,002

0,003

0,004

0,005

0,006

0,007

Re(

χ) [e

mu/

g]

Temperature [K]

f=9970 Hz; HAC=1 Oe f=625 Hz; HAC=2 Oe f=80 Hz; HAC=5 Oe f=7 Hz; HAC=5 Oe

ZnO:Fe2O330 wt.%calcination process

0 10 20 30 40 50 60 70 80 900,000

0,001

0,002

0,003

Re(

χ) [e

mu/

g]

Temperature [K]

9970 Hz; 1 Oe 625 Hz; 5 Oe 80 Hz; 5 Oe 7 Hz; 10 Oe

ZnO: Fe2O3

30 wt.%hydrothermal process

a) b)

0 10 20 30 40 50 60 70 80 900,00

0,01

0,02 9970 Hz; 1 Oe 625 Hz; 2 Oe 80 Hz; 2 Oe 7 Hz; 5 Oe

Re(

χ) [e

mu/

g]

Temperature [K]

ZnO:Fe2O360 wt.%calcination process

0 10 20 30 40 50 60 70 80 900,000

0,002

0,004

0,006

0,008 9970 Hz; 1 Oe 625 Hz; 5 Oe 80 Hz; 5 Oe 7 Hz; 10 Oe

Re(

χ) [e

mu/

g]

Temperature [K]

60 wt. % of Fe2O3hydrothermal process

c) d)

Fig. 12. The frequency dependence of Re(χ) for nanocrystalline sample with low and high concentration of Fe2O3 a) x= 30 wt.% (calcination process); b) x= 30 wt.% (hydrothermal

process); c) x= 60 wt.% (calcination process); d) x= 60 wt.% (hydrothermal process).


42

For calcination process, only for the lowest concentration of Fe2O3 (5 wt. %), the value typical for superparamagnetic system of interacting nanoparticles was determined. For higher concentration of Fe2O3, the calculated values of Φ are in the regime typical for spin-glass systems (0.03-0.04). For hydrothermal method, the values of parameter Φ are in the range 0.06-0.07 for all measured samples (from 5 wt. % to 70 wt. % of Fe2O3). It was shown in our structural studies that hydrothermal method leads to lower agglomeration degree of nanocrystals. The dynamic magnetic properties of the system of nanoparticles strongly depend on the degree of agglomeration and particularly random agglomeration of magnetic nanoparticles can lead to the spin-glasslike behavior. Additionally we used the phenomenological Vogel-Fucher (VF) law to explain the dynamic magnetic behavior of investigated systems [37]. The VF law was introduced in the field of spin-glasses and disordered magnetism and it has been stated clearly that this law is phenomenological:

f = f0 exp[-Ea/kB(Tf – T0)] (2), where f is the driving frequency of HAC, f0 is the frequency factor for the relaxation process, Ea the height of the energy barrier due to magnetic anisotropies, and kB is a Boltzman’s constant, T0 is a physically unclear parameter but has been regarded as a phenomenological parameter which describes the interactions between particles.

The VF law is useful to compare the frequency sensitivity of Tf for different magnetic systems (spin-glasses, superparamagnetic-like systems) through the ratio (Tf -T0)/Tf , where T0 is the phenomenological parameter and can be only an estimate of the interaction strength [34-35]. For small coupled particles (superparamagnetic system) the values of criterium (Tf -T0)/Tf are in the range 0.25-1.0.

10 15 20 25 30 35

0,030

0,035

0,040

0,045

calcination process

1/ln

(f 0/f)

Tf [K]

ZnO:Fe2O3

30 wt.%

hydrothermal synthesis

Fig. 13. The variation of Tf with the frequency f at the AC magnetic field for ZnO doped 30 wt.% of Fe2O3 in calcination and hydrothermal processes. The solid line represents the VF

law fitted to the experimental data. We fitted our experimental data with Eq. (2). The example results of the fitting

procedure for VF law are shown in Fig. 13. The determined values of the ratio (Tf -T0)/Tf are higher in the case of hydrothermal method. This is in agreement with calculated Φ criterion. 2.1.2. Magnetic properties of ZnO:CoO

The systematic magnetic measurements were performed. The AC magnetic


43

susceptibility χ studies in the temperature range 4.2 – 150 K using a mutual inductance method were carried out. The selected results of susceptibility measurements are shown in Fig. 14. In our experiment we have not observed any evidence of antiferromagnetic or superaparamagnetic behavior at low temperatures for all the investigated samples. The high temperature behavior of the inverse low-field susceptibility χ--1 was nearly linear for all samples prepared by calcination and hydrothermal method samples. All samples exhibited Curie-Weiss law behavior at high temperatures. The diamagnetic AC susceptibility of -0.33·10-6 emu/g for ZnO [38] was substracted from the measured magnetic susceptibility.

0 50 100 1500,0

1,0x10-5

2,0x10-5

3,0x10-5

Re(

χ) [e

mu/

g]

Temperature [K]

ZnO:CoO40 wt.%calcination method

0 50 100 1500,0

1,0x10-4

2,0x10-4

ZnO: CoO40 wt.%hydrothermal method

Re(

χ) [e

mu/

g]

Temperature [K]

a) b)

Fig. 14. The selected results of AC susceptibility measurements as a function of temperature (f=625 Hz, HAC=5 Oe) a) 40 wt.% (calcination process) b) 40 wt.% (hydrothermal method).

The solid lines represent the Curie-Weiss law fitted to the experimental data. The selected results of the fitting procedure to the Curie-Weiss law are shown in Fig.

15. For all samples the negative values of the Curie-Weiss temperature θ were determined. The obtained negative values of θ indicate that the antiferromagnetic interactions are predominant for all studied samples, calcinated as well as hydrothermal. For calcinated ZnO:CoO samples the Curie-Weiss temperature increases significantly with the content of magnetic dopant, indicating the enhancement of the predominant antiferromagnetic coupling. For the hydrothermal method, the opposite effect is observed. We observe the considerable decrease of Curie-Weiss temperature with the content of CoO indicating the breakdown of the predominant antiferromagnetic interaction. The magnetic parameters like Curie-Weiss temperature depend on the size of the nanoparticles. It is well known effect that the role of the surface atoms increases with the decrease in particle size. The structural measurements revealed that the ZnO phase is much less distinguishable for samples with higher content of magnetic dopant. Thus, the predominance of Co3O4 phase can be the reason of the increase of Curie-Weiss temperature values with the content of magnetic dopant for calcination method. In the case of hydrothermal method, the considerable amount of ZnCo2O4 phase for samples with higher content of magnetic dopant can cause the breakdown of predominant antiferromagnetic coupling and result in the observed decrease of θ values.


44

0 50 100 1500.0

2.0x105

4.0x105

6.0x105

8.0x105

10 wt. %, θ =- 4K

1/R

e(χ)

[g/e

mu]

Temperature [K]

θ

40 wt. %, θ = -18 K


a)

0 50 100 1500,0

2,0x104

4,0x104

6,0x104

8,0x104

50 wt. %, θ = -0.5 K

40 wt. %, θ = -8 K

1/R

e(χ)

[g/e

mu]

Temperature [K]

10 wt. %, θ = -22 KZnO:CoOhydrothermal method

b)

Fig.15. Inverse magnetic susceptibility for ZnO:CoO nanocrystalline samples, a) calcination method, b) hydrothermal process. The results of the fitting procedure - the values of

determined Curie-Weiss temperature are shown in figures. 2.1.3. Magnetic properties of ZnO:MnO

The results of magnetic AC susceptibility for selected samples of ZnO doped with MnO are shown in Fig. 16. For the samples with low content of MnO (up to 50 wt. %) the broad peak in the real part of magnetic susceptibility was observed. We did not observe the shift of the maximum of the peaks with the frequency for all the samples. However one should notice that observed peaks in AC susceptibility are very broad. This can be a result of a large nanocrystallite size distribution in the sample.

0 20 40 60 80 100 120 140 1600,0000

0,0005

0,0010

0,0015

0 20 40 60 80 100 120 140 1600,0

1,0x10-5

2,0x10-5

3,0x10-5

4,0x10-5

5,0x10-5

M (

emu)

T (K)

ZnO:MnO30 wt. %H = 10 Oe

10.4 K

FC

ZFC

Re(

χ) [e

mu/

g]

Temperature [K]

5 wt.% 10 wt. % 20 wt. %

30 wt. % 40 wt.%

50 wt. % 60 wt. %

70 wt. % 80 wt. % 95 wt. %

ZnO:MnOcalcination processHAC = 5 Oe; f = 625 Hz

Fig. 16. Magnetic AC susceptibility for ZnO:MnO nanocrystalline samples prepared by calcination method. The inset shows SQUID magnetization for sample with 30 wt. % of

MnO. For samples with higher content of magnetic dopant we observed peak at around 43

K, characteristic of a Mn3O4 ferrimagnet. This is consistent with the structural measurements (the presence of Mn3O4 crystalline phase for samples above 60 wt.% of MnO).


45

To better understand the magnetic properties of these systems we performed SQUID magnetization measurements as a function of temperature at small applied magnetic fields. The inset in Fig. 16 shows typical SQUID magnetization for sample with low content of magnetic dopant. A clear bifurcation of the FC and ZFC plots is visible. The furcation of ZFC and FC curves at a certain temperature is one of the characteristic features of superparamagnetic system. However, the coinciding broad maximum observed on the ZFC curve occurs at a slight lower temperature than temperature maximum observed for AC magnetic susceptibility. Such behavior usually signalizes a certain particle size distribution in nanocomposite. For ZnO:MnO samples we observed five different crystalline phases. One should realize that interpretation of magnetic results in such a case is quite a difficult task. However the results of magnetic measurements corroborate very well with the magnetic resonance studies [25]. 3. Conclusions

We have studied structural and magnetic properties of nanocrystalline ZnO doped with transition metal oxides. For ZnO:Fe2O3 the structural studies revealed that in the samples obtained using both methods the peaks belonging to ZnO and ZnFe2O4 can be found. We analyzed the observed frequency dependence of the peak temperature in the AC susceptibility curve using the empirical parameter Φ that is a quantitative measure of the frequency shift and is given by the relative shift of the peak temperature per decade shift in frequency as well as Vogel-Fulcher law with equally good fit. For nanocrystals synthesized by hydrothermal process the results of low-field AC susceptibility are satisfactorily explained by superparamagnetic model including inter-particle interactions. For samples synthesized by calcinations process with the increase of magnetic Fe2O3 content, the spin-glasslike behavior is observed.

For ZnO:CoO the structural studies revealed that in the samples obtained using calcination method the peaks belonging to ZnO and Co3O4 can be found. For hydrothermal method ZnO and ZnCo2O4 phases were recognized. For all samples high temperature Curie-Weiss behavior in AC magnetic susceptibility was observed. We observed that the determined negative values of the Curie-Weiss temperature θ depend strongly on the nominal content of cobalt oxide. It was shown that for calcination method the values of θ increase with the increase of magnetic dopant indicating enhancement of predominance of antiferromagnetic interactions. For hydrothermal method the opposite effect was observed indicating the breakdown of predominance of aniferromagnetic coupling with the increase of nominal magnetic dopant.

For ZnO doped with MnO (up to 50 wt. %) the peak in low temperature magnetic susceptibility is observed. The DC magnetic measurements suggested superparamagnetic behavior. Futher magnetic studies, including nonlinear magnetic susceptibility could be helpful in studying magnetic properties of ZnO:MnO. For samples with higher magnetic dopant, the presence of ferrimagnetic Mn3O4 is clearly visible. The structural measurements revealed the presence of Mn3O4 crystal structure for samples with MnO content above 50 wt. %. Acknowledgements This work was supported by the project financed from funds of National Science Center in Poland granted under decision No. DEC-2011/01/B/ST5/06602.

We acknowledge Piotr Dziawa for performing SQUID measurements and prof. E. Mijowska for TEM measurements.


46

4. References

1. M. Knobel, W. C. Nunes, L. M. Socolovsky, E. De Niasi, J. M. Vargas, and J. C. Denardin, J. Nanosci. Nanotechnol. 8 (2008) 2836 and references therein.

2. An-Hui Lu, E. L. Salabas, and Ferdi Schüth, Angew. Chem. Int. Ed. 46 (2007) 1222. 3. S. Purushotham and R. V. Ramanujan J. Appl. Phys. 107 (2010) 114701. 4. I. Djerdj, Z. Jagličić, D. Arčon, and M. Niederberger, Nanoscale, 2 (2010) 1096. 5. A. Sundaresan, R. Bhargavi, N. Rangarajan, U. Siddesh, and C. N. R. Rao, Phys. Rev.

B 74 (2006) 161306. 6. X. Jia, W. Yang, M. Qin, and J. Li, J. Magn. Magn. Mater. 321 (2009) 2354. 7. N. Obradovic, S. Stevanovic, V. Zeljkovic, M. M. Ristic, Powder Metallurgy and

Metal Ceramics, Vol. 48, Nos. 3-4 ( 2009) 182-185. 8. G. Lawes, A. S. Risbud, A. P. Ramirez, Ram Seshadri Phys. Rev. B. 71(2005)

045201 and references therein. 9. K. Ueda, H. Tabota, and T. Kamai, Appl. Phys. Lett. 79 (2001) 988. 10. S. Kolesnik,B. Dabrowski, and J. Mais, J. Appl. Phys. 95 (2004) 2582. 11. C. Sudakar, J. S. Thakur, G. Lawes, R. Naik, V. M. Naik, Phys. Rev. B 75 (2007)

054423. 12. T. Dietl, Acta Physica Polonica A, 111 (2007) 27 13. T. C. Kaspar, T. Droubay, S M. Heald, P. Nachimuthu, C.M. Wang, V

Schutthanandan, C. A. Johnson, D. R. Gamelin, S. A. Chambers. New J. Phys. 10 (2008) 055010.

14. X. G. Zhao and Z. Tang J. Appl. Phys. 111 (2012) 084321. 15. R. Vidya Sagar, Advanced Science, Engineering and Medicine, 4 (2012) 237. 16. T. Singh, T. J. Mountziaris, and D. Maroudas Appl. Phys. Lett. 97 (2010) 073120. 17. P. K. Sharma, R. K. Dutta, A. C. Pandey, S. Layek and H. C. Verma, Journal of

Magnetismand Magnetic Materials, 321 (2009) 2587. 18. D. Karmakar, S. K. Mandal, R. M. Kadam, P. L. Paulose, A. K. Rajarajan, T. K.

Nath, A. K. Das, I. Dasgupta and G. P. Das, Phys. Rev. B, 75 (2007) 14. 19. S. A. Chambers, D. A. Schwartz, W. K. Liu, K. R. Kittilstved, D. R. Gamelin

“Growth, Electronic and Magnetic Properties of Doped ZnO Epitaxial and Nanocrystalline Films” in Magnetism in New Semiconducting and Insulating Oxides, edited by N. H. Hong, 2007, Research Signpost Press

20. L. B. Duan, G. H. Rao, J. Yu, Y. C. Wang, W. G. Chu and L. N. Zhang, J. Appl. Phys., 102 (2007) 103907.

21. O. D. Jayakumar, H. G. Salunke, R. M. Kadam, M. Mohapatra, G. Yaswant and S. K. Kulshreshtha, Nanotechnology, 17 (2006) 1278.

22. A. R. Jamaludin, S. R. Kasim, Z. A. Ahmad, Science of Sintering, 42 (2010) 345-355. 23. D. Sibera, R. Jędrzejewski, J. Mizeracki, A. Presz, U. Narkiewicz, W. Łojkowski,

Acta Physica Polonica A, 116 (2009) S-133. 24. I. Kuryliszyn-Kudelska, W. D. Dobrowolski, Ł. Kilanski, B. Hadžić, N. Romčević,

D. Sibera, U. Narkiewicz and P. Dziawa, J. Phys.: Conf. Ser. 200 (2010) 072058. 25. N. Guskos, G. Zolnierkiewicz, J. Typek, D. Sibera, U. Narkiewicz and W. Łojkowski,

Acta Physica Polonica A, 120 (2011) 1074. 26. U. Narkiewicz, D. Sibera, I. Kuryliszyn-Kudelska, L. Kilanski, W. Dobrowolski, N.

Romcevic Acta Phys. Pol. A 113 (2008) 1695. 27. N. Guskos, G. Zolnierkiewicz, J. Typek, D. Sibera, U. Narkiewicz, Rev. Adv. Mater.

Sci. 23 (2010) 224. 28. N. Romčević, R. Kostić, B. Hadžić, M. Romčević, I. Kuryliszyn-Kudelska, W. D.

Dobrowolski, U. Narkiewicz, D. Sibera, Journal of Alloys and Compounds 507 (2010), 386

29. A. L. Patterson, Phys. Rev., 56 (1939) 978.


47

30. R. Kostić, N. Romčević, M. Romčević, B. Hadžić, R. Rudolf, I. Kuryliszyn-Kudelska, W. D. Dobrowolski, U. Narkiewicz, D. Sibera, Acta Physica Polonica A, 116 (2009) 65.

31. Ph. Colomban and G. Gouadec J. Raman Spectroscopy 38 (2007) 598. 32. R. Y. Sato-Berru, A. Vázquez-Olmos, A. L. Fernández-Osorio and S. Sortes-

Martinez, J. Raman Spectrosc 38 (2007) 1073. 33. A. Mydosh, Spin-glasses: An Experimental Introduction (London: Taylor and

Francis) (1993) 34. G. F. Goya, T. S. Berquo, F. C. Fonseca, M. P. Morales, Journal of Appl. Phys. 94

(2003) 3520. 35. L. Dorman, D. Fiorani, E. Tronc, Adv. Chem. Phys. 98 (1997) 283. 36. S. H. Masunaga and R. F. Jardim, P. F. P. Fichtner, J. Rivas, Phys. Rev. B 80 (2009)

184428. 37. I. Kuryliszyn-Kudelska, B. Hadžić, D. Sibera, M. Romčević, N. Romčević, U.

Narkiewicz, W. Dobrowolski, Journal of Alloys and Compounds, 509 (2011) 3756. 38. CRC Handbook of Chemistry and Physics, 80th edn, 1999-2000, p.4-136. (CRC Press

LLC, Boca Raton,London, New York, Washington, D. C. (1999). Садржај: Широко-зонске ZnO наночестице тренутно имају важну потенцијалну примену у електро-оптичким уређајима, транспарентним заштитним филмовима од ултраљубичастог зрачења и уређајима за примену у спинтроници. У раду смо проучавали магнетна својства нанокристала ZnO(Fe, Co, Mn) припремљених помоћу две методе синрезе. Користили смо микроталасну потпомогнуту хидротермалну синтезу и традиционални хемијски поступак уз каснију калцинацију. Детаљна структурна карактеризација извршена је помоћу рентгенске дифракције и микро-Раманове спектроскопије. Морфологија узорака проучавана је помоћу ЦЕМ и ТЕМ микроскопије. Приказани су резултати систематских мерења AC магнетне сусцептибилности у зависности од температуре и фреквенције, као и магнетитзације мерене помоћу SQUID магнетометра. Мерење магнетизације је показало јасно раздвајање између FC и ZFC кривих. Такво понашање сугерише суперпарамагнетно понашање изнад температуре блокирања. Динамичка магнетна мерења су изведена у малом AC магнетном пољу чија амплитуда није прелазила 5 Oe и различитим вредностима фреквенције (од 7 Hz до 9970 Hz). За ZnO(Fe) иi ZnO(Mn), максимум у AC sсусцептибилности је нађен у реалном Re(χ) и имагинарном Im(χ) делу AC сусцептибилности. Уочену зависност температуре максимума у AC кривама од фреквенције смо анализирали користећи емпиријски параметар Φ који је квантитативна мера фреквенцијског помераја, као и помоћу Вогел-Фулхер закона. Уочили смо два типа магнетног понашања, понашање налик спинском стаклу и понашање налик суперпарамагнетизму, зависно од начина синтезе. За ZnO(Co) нанокристалне узорке, уочено је Кири-Вајсово понашање у високотемпературном делу AC магнетне сусцептибилности. Уочили смо да негативне вредности Кири-вајсове температуре θ јако зависе од номиналног садржаја кобалт оксида. Показано је да за поступак калцинације вредности θ расту са порастом садржаја магнетних јона, указујући на јачање доминације антиферомагнетних интеракција. За хидротермалну методу уочен је супротан ефекат указујући на слабљење доминације антиферомагнетног спрезања при порасту номиналног садржаја магнетних јона.


48

Овај рад даје детаљну дискусију структурних и магнетних својстава ZnO нанокристала, поред технолошких разматрања као што су различити поступци хемијске синтезе. Кључне речи: полупроводници, магнетизација, магнетна мерења, апсорпција и расејање светлости

transition metals in zno nanocrystals – magnetic …...i. kuryliszyn-kudelska et al. /science of...

Documents