Ferromagnetism in Fe-doped ZnO bulk samples with additional Cu doping

Article abstract of DOI:10.1002/pssa.200925566

Room-temperature (RT) ferromagnetism (FM) has been observed in Fe-doped ZnO bulk samples fabricated by a coprecipitation method, and the effect of additional Cu doping on magnetic properties of the samples has been investigated. The results of XRD show that all doping samples are single-phase with a wurtzite structure. X-ray photoelectron spectroscopy reveals that Fe is in 3+ state substituting for Zn in the sample without Cu doping, however, Fe shows a mixed oxidation state of 2+ or 3+ in the sample with additional Cu doping, where Cu is 1+ valence. Magnetic measurements indicate that additional Cu doping enhances the FM of the system, and the magnetization increases gradually with the increasing Cu content x up to 1%. The observed RT FM was attributed to the ferromagnetic coupling between Fe³⁺ ions mediated by holes.

Full text of DOI:10.1002/pssa.200925566

Phys. Status Solidi A 207, No. 11, 2553–2557 (2010) / DOI 10.1002/pssa.200925566
a
p s s
www.pss-a.com
applications and materials science
Ferromagnetism in Fe-doped ZnO
bulk samples with additional Cu doping
1
1
2
2
1
1
2
*,2
Y. Q. Wang , L. Su , L. Liu , Z. M. Tian , T. Q. Chang , Z. Wang , S. Y. Yin , and S. L. Yuan
1
Department of Technology and Physics, Zhengzhou University of Light Industry, Zhengzhou 450002, People’s Republic of China
Department of Physics, Huazhong University of Science and Technology, Wuhan 430074, People’s Republic of China
2
Received 4 November 2009, revised 28 June 2010, accepted 7 July 2010
Published online 26 August 2010
Keywords doping, ferromagnetism, magnetic semiconductors, magnetization, ZnO
*
Corresponding author: e-mail yuansl@hust.edu.cn, Phone: þ86 27 87556580, Fax: þ86 27 87554525
Room-temperature (RT) ferromagnetism (FM) has been oxidation state of 2þ or 3þ in the sample with additional Cu
observed in Fe-doped ZnO bulk samples fabricated by a doping, where Cu is 1þ valence. Magnetic measurements
coprecipitation method, and the effect of additional Cu doping indicate that additional Cu doping enhances the FM of the
on magnetic properties of the samples has been investigated. system, and the magnetization increases gradually with the
The results of XRD show that all doping samples are single- increasing Cu content x up to 1%. The observed RT FM was
3þ
phase with a wurtzite structure. X-ray photoelectron spectro- attributed to the ferromagnetic coupling between Fe ions
scopy reveals that Fe is in 3þ state substituting for Zn in the mediated by holes.
sample without Cu doping, however, Fe shows a mixed
ß 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1
Introduction Numerous technological break- be responsible to the FM in Fe-implanted ZnO single-crystal
throughs are envisaged with the use of magnetic semicon- samples. A detailed structural, chemical bonding states, and
ductor materials that show a ferromagnetic ordering at or magnetic investigation with high-quality codoped samples is
above room temperature. Oxide diluted magnetic semicon- essential to resolve the controversy.
ductors (DMSs) have recently attracted a lot of research
Most recently, we have reported RT FM in Fe-doped
effort due to their potential application in the field of ZnO single-phase bulk samples prepared by a coprecipi-
spintronic devices [1–3] designed to process information by tation method, with the highest doping concentration of
manipulating the spin of electronic currents, such as spin x ¼ 2% [21]. In the present work, we prepared Fe-doped ZnO
light-emitting diodes (spin-LEDs) [4–7], spin-polarized bulk samples that are codoped with Cu under similar
solar cells [8], and magneto-optical switches [9].
preparation conditions and experimentally study the effect
Room-temperature (RT) ferromagnetism (FM) in ZnO of additional Cu doping on the magnetic properties of Fe-
doped with transition-metal elements (V, Cr, Mn, Fe, Co, and doped ZnO bulk samples, attempting to clarify the origin of
Ni) has been predicted by Dielt et al. [10] and Sato and FM in these DMS materials.
Katayama-Yoshida [11]. Following the original report by
Ueda et al. [12], many experiments have been carried out to
confirm the theory prediction and the initial result. Some ponent Zn0
2 Experimental Bulk samples with a nominal com-
Fe0.02Cu O (x ¼ 0, 0.25, 0.5, and 1%) were
.98ꢀx
results dealing with RT FM of Fe-doped ZnO samples have prepared by a coprecipitation method. Appropriate pro-
x
also been published [13–20]. However, there remain portions of Zn(NO ) ꢁ 6H O, Cu(NO ) ꢁ 3H O, and
3
2
2
3 2
2
some questions regarding the origin of the magnetic behavior Fe(NO ) ꢁ 9H O high-purity (99.9%) powders were
3
3
2
in these DMS materials. Additional Cu doping is deemed thoroughly mixed according to the desired stoichiometry,
essential to achieve RT FM in Fe-doped ZnO bulk samples the powders were dissolved in distilled water to get a
[
18]. Contrarily, Shim et al. [19] showed that RT FM in Fe- homogeneous solution. The mixture was stirred strongly
doped ZnO with additional Cu doping stems from nano- while a proper amount of Na(OH) aqueous solution was
crystalline ZnFe O with an inverted spinel structure. Zhou introduced, controlling the pH to 7 to deposit all cations
et al. [20] reported that the metallic a-Fe nanocrystals may completely. The obtained precipitate was thoroughly washed
2
4
ß 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

a
p s s
2
554
Y. Q. Wang et al.: Ferromagnetism in Fe-doped ZnO bulk samples
with distilled water and dried in air at 200 8C, and then
prefired at 400 8C for 8 h. The prepared powders were
ground, pelletized, and sintered at 600 8C for 12 h in an Ar
(
a) pure ZnO
(
b) Zn0.98Fe0.02
O
gas atmosphere. The crystal structure of Zn0
samples was characterized by X-ray diffraction (XRD,
Fe0.02Cu O
c) Zn0.97Fe0.02Cu0.01
( O
.98ꢀx
x
˚
X’Pert PRO) using Cu Ka radiation (l ¼ 1.5406 A).
Chemical bonding states and chemical compositions of the
samples were analyzed by X-ray photoelectron spectroscopy
(
c)
(b)
a)
(
(
XPS, VG Multilab 2000) with Al
hn ¼ 1486.6 eV). A physical properties measurements
K
radiation
(
system (PPMS, Quantum Design) was used to investigate
the magnetic properties of the doped samples.
20
30
40
50
θ (degree)
60
70
80
2
3
Results and discussion The crystal structure of the
Figure 1 Typical powder XRD patterns for pure ZnO,
Zn0.98Fe0.02O, and Zn0.97Fe0.02Cu0.01O.
samples was characterized by XRD using Cu Ka radiation.
Data werecollected using a step scan of 0.0178in 2u. Figure1
presents the typical powder XRD patterns for pure ZnO,
Zn0.98Fe0.02O, and Zn0.97Fe0.02Cu0.01O. No clear difference located at 1021.8 eV for Zn0.98Fe0.02O and 1021.9 eV for
was found in XRD patterns for all the samples, an additional Zn0.97Fe0.02Cu0.01O, respectively. The Fe 2p3/2 peak for
1
% Cu doping has induced no change in structure. All the Zn0.98Fe0.02O occurs at 710.7 eV as presented in Fig. 2c,
diffraction peaks can be indexed to a wurtzite structure as suggesting that Fe ions in the sample should have a valence
ZnO, no spurious phase was found within our detection limit of 3þ, consistent with the results of XRD. For
of less than 5%. This means that iron and copper have been Zn Fe Cu O, the main peak of Fe 2p at 711.2 eV
0
.97 0.02
0.01
3/2
incorporated into the lattice structure, forming a solid splits into two peaks with the binding energy of 710.5 and
solution instead of precipitates. In addition, the position of 710.9 eV, as shown in Fig. 2d, which corresponds to a mixed
the diffraction peak of Zn0.98Fe0.02O was found a shift to oxidation state of 2þ or 3þ. In addition, the peak of Cu 2p
3/2
1
higher angles compared with pure ZnO, suggesting the is located at 931.8 eV, corresponding to Cu , in agreement
þ
3
þ
2þ
possibility of Fe substitution for Zn because of its with another report [17]. According to the experimental
smaller ionic radius of 0.064 nm compared to 0.074 nm for results, we can exclude the possibility of the formation of Fe
2
þ
Zn
.
Chemical bonding states in Zn0.98Fe0.02
metal clusters, since, if Fe exists as a metal cluster itself in
and the samples, the binding energy of the Fe 2p3/2 state should
O
Zn0.97Fe0.02Cu0.01O bulk samples were characterized by be typical located at 706.9 eV [22].
XPS measurement. The XPS spectra have been charge The magnetic properties of Zn0
Fe0.02Cu O were
x
.98ꢀx
corrected to the adventitious C 1s peak at 284.6 eV and the studied by checking the temperature and magnetic field
core states of Zn 2p, Fe 2p peaks for Zn0.98Fe0.02O and Zn 2p, dependence of the magnetization. Figure 3 shows the
Fe 2p, Cu 2p peaks for Zn0.97Fe0.02Cu0.01O were presented in magnetic field dependence of magnetization (M–H) for
Fig. 2. The binding energy of the Zn 2p3/2 state is typical Zn0.98Fe0.02O bulk samples both without and with additional
(
a) Zn 2p
2p1/2
S
2
p3/2
(d) Fe 2p
p3/2
2
p1/2
2
*
(
b) Zn 2p
2p3/2
2
p1/2
1050
1040
1030
1020
730
720
710
700
Binding Energy (eV)
Binding Energy (eV)
2
p1/2
2p3/2
(
c) Fe 2p
S
(e) Cu 2p
p3/2
2
Figure 2 XPS spectra for Zn0.98Fe0.02O and
Zn0.97Fe0.02Cu0.01O bulk samples: (a) Zn 2p3/2
XPS spectrum for Zn0.98Fe0.02O; (b) Zn 2p3/2
XPS spectrum for Zn0.97Fe0.02Cu0.01O; (c) Fe
2
p1/2
S
*
2
2
p3/2 XPS spectrum for Zn0.98Fe0.02O; (d) Fe
3/2 XPS spectrum for Zn0.97Fe0.02Cu0.01O;
p
(
e) Cu 2p3/2 XPS spectrum for Zn0.97Fe0.02-
730
720
710
700
950
940
930
920
Cu0.01O.Thepeaksmarkedby‘‘S’’aresatellite
peaks.
Binding Energy (eV)
Binding Energy (eV)
ß 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.pss-a.com

Original
Paper
Phys. Status Solidi A 207, No. 11 (2010)
2555
0
0
0
.02
.01
.00
a)
b)
0.15
0.10
x=0
x= 0
x=0.25%
x=0.5%
x=1%
x=1%
H=1000Oe
T=300 K
-
0.01
0.02
0
.05
.00
-
0
0
0
.4
.2
.0
0
x=0
0
100
200
300
x=0.25%
x=0.5%
x=1%
T (K)
Figure 4 Magnetization as a function of temperature (M–T) for
Zn0.98ꢀxFe0.02Cu
1000 Oe from 10 to 300 K.
T=10 K
2500
x
O (x ¼ 0–1%) bulk samples in an applied field of
-0.2
-0.4
-5000
-2500
0
5000
O (x ¼ 0–1%) bulk
showing a paramagnetic-like behavior. However, it is clear
that both the x ¼ 0 and 1% samples are in magnetization
saturation at about 15 K. Furthermore, the nonzero magne-
tization values and the hysteresis loops (seen in Fig. 3) at
H (Oe)
Figure 3 M–H curves for Zn0.98ꢀxFe0.02Cu
samples at 300 K (a) and 10 K (b).
x
300 K indicate that both the samples are FM at 300 K.
So one can conclude that the Curie temperature of
Cu doping. Even though the coercive force (H ) and the Zn
Fe0.02Cu O (x ¼ 0–1%) series samples is higher
C
0.98ꢀx
remnant magnetization (M ) are modest, the hysteresis than 300 K.
x
R
loops that can be seen clearly from the M–H curves ensure
It is noted that M of the samples is quite small, for
S
all samples are FM at RT, as shown in Fig. 3a. A small example, the M is only 0.25m /Fe for the x ¼ 0.5% sample
S
B
3
þ
amount of additional Cu doping (x ¼ 0.25%) brought a slight at 10 K, however, the theory magnetic moment of Fe ions
increment in M, and M gradually increases with the increase is 3m /Fe, the experiment value is one order of magnitude
B
of Cu concentration up to x ¼ 1%. The saturation magnetiza- less than the theory value, meaning that the ferromagnetic
tion (M ) of the Zn
S
Fe Cu O series samples is exchange interaction is quite weak.
x
0.98ꢀx 0.02
B
estimated to be 0.009m /Fe, 0.011m /Fe, 0.011m /Fe, and
B
To further clarify this issue, M as a function of T for the
.012m /Fe for the x ¼ 0, 0.25, 0.5, and 1% samples, x ¼ 0.25 and 0.5% samples were measured in zero-field-
B
0
B
respectively.
Figure 3b presents the M–H curves of Zn0
cooling (ZFC) and field-cooling (FC) modes with an applied
Fe0.02 field of 1000 Oe, as shown in Fig. 5.
.98ꢀx
Cu O (x ¼ 0–1%) at 10 K. One can find that the H are quite
A clear discrepancy between the ZFC and FC curves can
small, and the M–H curves show superparamagnetic be observed in the whole temperature range, indicating
x
C
behavior, indicating the existence of small ferromagnetic clear superparamagnetic behavior with a blocking tempera-
clusters. The saturation magnetization (M ) of the series of ture at about 20 K. The results further show that there are
S
samples is estimated to be 0.15m /Fe, 0.21m /Fe, 0.25m / many small ferromagnetic clusters in our samples. Most of
B
B
B
Fe, and 0.28m /Fe for the x ¼ 0, 0.25, 0.5, and 1% samples, these clusters are isolated due to the long distance between
B
respectively.
them, fixed by the low doping concentration, therefore they
From the above, a RT ferromagnetic behavior has been have a small contribution to the long range ferromagnetic
observed in a Fe-doped ZnO bulk sample. One can find that order, so the M of the system is lower than the theory value. It
additional Cu doping is beneficial to the FM of the system, is noted that the M in the ZFC mode are even larger than the
but not very crucial as theories have predicted. This result is M in the FC mode in the temperature range from 20 to 300 K.
different from some previous publications [18, 23]. In those The anomaly may be ascribed to the existence of small
reports, additional Cu doping is quite important to achieve ferromagnetic clusters. In the FC mode, magnetic clusters
and enhance RT FM in Zn–Fe–O and Zn–Co–O system.
Further studies on magnetic behaviors of Zn0
Cu O were performed through measuring the magnetization field turned the magnetic moments of these small magnetic
only felt the applied magnetic field at high-temperature.
Fe0.02 With the decrease of temperature to T, the applied magnetic
.98ꢀx
x
as a function of temperature (M–T). Figure 4 shows the M–T clusters against the thermal fluctuation. In the ZFC mode,
curves typical for the x ¼ 0 and 1% samples in an applied all ferromagnetic clusters were turned by the applied
field of 1000 Oe.
magnetic field at low temperature. Every magnetic cluster
For the two samples, no clear magnetic transition is was in the applied magnetic field and the internal magnetic
found in the whole temperature range from 10 to 300 K, field stemmed from the near-neighbor ferromagnetic
www.pss-a.com
ß 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

a
p s s
2
556
Y. Q. Wang et al.: Ferromagnetism in Fe-doped ZnO bulk samples
3
þ
ions in the samples, thereby,
0.16
concentration of Fe
weakening the ferromagnetic coupling of the samples. For
this reason, the additional Cu doping increase the FM of
Zn0.98Fe0.02O, but not very crucial as theories have
predicted. Further research on this issue might need to be
done in the future to clarify the supposition.
0
0
0
0
.12
.08
.04
.00
x=0.25% ZFC
x=0.25% FC
x=0.5% ZFC
x=0.5% FC
4
Conclusion In conclusion, RT FM is observed in a
single-phase Fe-doped ZnO bulk sample prepared by a
coprecipitation method. The effect of additional Cu doping
on the magnetic properties of an Fe-doped ZnO sample was
investigated. Based on the results of XRD and XPS, the
observed FM is attributed to the ferromagnetic coupling
0
50
100
150
T (K)
200
250
300
3
þ
between Fe
ions mediated by holes formed during
mF i og du er se. 5 M–Tcurvesforx ¼ 0.25and0.5%samplesinZFCandFC
preparation process. Additional Cu doping is beneficial to
the FM of the system to some extent, but as marked as
theories have predicted.
clusters. To break the ferromagnetic ordering of these
small magnetic clusters, the thermal fluctuation must
overcome the applied magnetic field and the internal
magnetic field, which need a higher temperature than T.
So, one can find that the contribution of these small
ferromagnetic clusters to the M of the system in the ZFC
mode is larger than is the case in the FC mode in a specific
temperature range.
Acknowledgements This work was supported by the
National 973 Project under grant no. 2006CB921606, by the
NationalScience Foundationof Chinaundergrant no. 10574049, by
the Foundation from the Ministry of the National Education under
Grant No. 20060487011, and by the Natural Science Foundation of
Henan Province under Grant No. 2009A140009.
Although the precise mechanism for ferromagnetic
coupling between magnetic ions is still being discussed
and debated in the literature, there is wide consensus that
carrier doping has an important effect on magnetic properties
of ZnO-based DMSs. Especially, as a p-type dopant in
semiconductors, additional Cu doping is deemed essential to
achieve RT FM in Fe-and Co-doped ZnO bulk samples
References
[
[
[
[
1] E. M. Wong and P. C. Searson, Appl. Phys. Lett. 74, 2939–
941 (1999).
2
2] D. C. Look, J. W. Hemsky, and J. R. Sizelove, Phys. Rev.
Lett. 82, 2552–2555 (1999).
3] H. J. Ko, Y. F. Chen, Z. Zhu, T. Yao, I. Kobayashi, and
H. Uchki, Appl. Phys. Lett. 76, 1905–1907 (2000).
4] B. T. Jonker, Y. D. Park, B. R. Bennett, H. D. Cheong,
G. Kioseoglou, and A. Petrou, Phys. Rev. B 62, 8180–8183
(2000).
[
18, 23]. In this research, however, RT FM is observed in an
Fe-doped ZnO sample without additional doping, additional
Cu doping has a positive effect on the FM of Zn0.98Fe0.02O.
The possibility of the formation of Fe metal clusters and
Fe O clusters can be excluded because Fe ions show 3þ
[5] Y. Ohno, D. K. Young, B. Beschoten, F. Matsukura, and
D. D. Awschalom, Nature (London) 402, 790–792 (1999).
3
4
[
[
6] J. M. Kikkawa and D. D. Awschalom, Nature (London) 397,
39–141 (1999).
7] R. Fiederling, M. Keim, G. Reuscher, W. Ossau, G. Schmidt,
valence according to the results of XPS. The only possible
Fe-based oxide is Fe O , however, Fe O is a typical anti-
ferromagneic oxide with a T of 963 K. Obviously, the FM
1
2
3
2 3
N
A. Waag, and L. W. Molenkamp, Nature (London) 402, 787–
observed in our samples should be intrinsic, and additional
Cu doping enhances the intrinsic ferromagnetic interaction
to a certain extent. Considering that Cu is a typical p-type
7
90 (1999).
8] I. Zutic, J. Fabian, and S. Das Sarama, Phys. Rev. B 64,
21201–121204 (2001).
[
[
1
1
þ
2þ
dopant in DMS, the substitution of Cu ions for Zn ions
will increase the hole concentration, one can conclude
9] Y. S. Didosyan, H. Hauser, G. A. Reider, and W. Toriser,
J. Appl. Phys. 95, 7339–7341 (2004).
the FM of Zn–Fe–O system must be dependent on the hole [10] T. Dielt, H. Ohno, F. Matsukura, J. Cibert, and D. Ferrand,
Science 287, 1019–1022 (2000).
11] K. Sato and H. Katayama-Yoshida, Jpn. J. Appl. Phys. Part 2:
(39), L555–L558 (2000).
12] K. Ueda, H. Tabata, and T. Kawai, Appl. Phys. Lett. 79, 988–
990 (2001).
13] M. Venkatesan, C. B. Fitzgerald, J. G. Lunney, and J. M. D.
Coey, Phys. Rev. Lett. 93, 177206 (2004).
14] A. Y. Polyakov, A. V. Govorkov, N. B. Smirnov, N. V.
Pashkova, S. J. Pearton, K. Ip, R. M. Frazier, C. R. Abernathy,
D. P. Norton, J. M. Zavada, and R. G. Wilson, Mater. Sci.
Semicond. Process. 7, 77–81 (2004).
concentration. So, we suppose that the ferromagnetic
behavior observed in this experiment should stem from
the ferromagnetic coupling between Fe ions mediated
by the holes formed during the preparation process. The
[
[
[
[
3
þ
1
þ
2þ
Cu ions substitution for Zn increases the hole concen-
tration of the samples and enhance the ferromagnetic
3
coupling between Fe ions, therefore, the FM is enhanced.
þ
1
On the other hand, the introduction of Cu ions induce an
þ
enhanced compensation effect that results in the onset of
Fe2
R
in the samples, accordingly decreasing the
ß 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.pss-a.com

Original
Paper
Phys. Status Solidi A 207, No. 11 (2010)
2557
[
[
[
[
15] K. Potzger, S. Zhou, H. Reuther, A. M u¨ cklich, F. Eichhorn, [19] J. H. Shim, T. Hwang, S. Lee, J. H. Park, S.-J. Han, and Y. H.
N. Schell, W. Skorupa, M. Helm, and J. Fassbender, Appl.
Phys. Lett. 88, 052508 (2006).
16] P. Wu, G. Saraf, Y. Lu, D. H. Hill, R. Gateau, L. Wielunski,
R. A. Bartynski, D. A. Arena, and J. Dvorcak, Appl. Phys.
Lett. 89, 012508 (2006).
Jeong, Appl. Phys. Lett. 86, 082503 (2005).
[20] S. Q. Zhou, K. Potzger, Q. Y. Xu, G. Talut, M. Lorenz, W.
Skorupa, M. Helm, J. Fassbender, M. Grundmann, and H.
Schmidt, Vacuum 83, S13–S19 (2009).
[21] Y. Q. Wang, D. S. Zhang, L. Liu, P. Li, X. X. Lan, Z. M. Tian,
J. H. He, S. Y. Yin, and S. L. Yuan, J. Magn. Magn. Mater.
320, 1423–1426 (2008).
17] Q. J. Feng, D. Z. Shen, J. Y. Zhang, B. H. Li, Z. Z. Zhang,
Y. M. Lu, and X. W. Fan, Mater. Chem. Phys. 112, 1106–
1109 (2008).
[22] D. Brion, Appl. Surf. Sci. 5, 133–135 (1980).
18] S.-J. Han, J. W. Song, C.-H. Yang, S. H. Park, J.-H. Park, [23] H.-T. Lin, T.-S. Chin, J.-C. Shih, S. H. Lin, T. M. Hong, R. T.
and Y. H. Jeong, Appl. Phys. Lett. 81, 4212–4214
2002).
Huang, F. R. Chen, and J. J. Kai, Appl. Phys. Lett. 85, 621–
623 (2004).
(
www.pss-a.com
ß 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim