Applied Surface Science 247 (2005) 493–496 www.elsevier.com/locate/apsusc
Cobalt-doped ZnO – a room temperature dilute magnetic semiconductor C.B. Fitzgerald *, M. Venkatesan, J.G. Lunney, L.S. Dorneles, J.M.D. Coey Physics Department, Trinity College, Dublin 2, Ireland Available online 26 February 2005
Abstract Room temperature ferromagnetism is observed in thin films of ZnO doped with 1–25 at.% cobalt. Bulk samples were synthesized by a solid-state reaction technique and (1 1 0) textured thin films were prepared by pulsed laser deposition on R-cut sapphire substrates. Films are semiconducting and transparent. Resisitivity increases with increasing cobalt content. Room temperature magnetic moments of 0.5-5 mB per Co atom were measured in the thin films. Optical spectrometry indicates that cobalt enters the tetrahedral sites of the wurtzite structure as Co2+. # 2005 Elsevier B.V. All rights reserved. Keywords: Cobalt-doped ZnO; Magnetic semiconductor; Ferromagnetism
Dilute magnetic semiconductors (DMS) produced by doping transition metal ions into non-magnetic semiconductors have attracted a great deal of interest [1]. It has been reported that the wide gap semiconductor ZnO exhibits ferromagnetism with a Curie temperature above room temperature when the oxide is doped with a few atomic percent of cobalt [2–8] or another transition element (V, Mn, Ni) [9–11]. Theoretical calculations predict high-Tc and large magnetization values for Mn-doped p-type ZnO [12]. However, thin films of ZnO doped with 3d transition * Corresponding author. Tel.: +353 1 6082171; fax: +353 1 6711759. E-mail address:
[email protected] (C.B. Fitzgerald).
metal ions prepared by different techniques – pulsed laser deposition [2], molecular beam epitaxy [13] and magnetron sputtering [14] – are usually n-type. Other reports suggest that no high-temperature ferromagnetic state exists in these systems [15]. Doubts also exist as to whether these are homogeneous, singlephase materials, particularly since the well-accepted mechanisms for ferromagnetic exchange – via spinpolarized p-band holes as in Ga1xMnxAs [16], or via double exchange as in mixed valence manganites [17] – do not seem to apply in these oxides. A new exchange mechanism involving donor electrons in an impurity band has been proposed [18]. Here we examine the effects of varying cobalt concentration in Zn1xCoxO thin films.
0169-4332/$ – see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.apsusc.2005.01.043
494
C.B. Fitzgerald et al. / Applied Surface Science 247 (2005) 493–496
Ceramic targets of Zn1-xCoxO, with x = 0.01–0.15, were synthesized from stoichiometric amounts of high purity ZnO and CoO powders. The oxide powders were ground under isopropanol for 1 h, then pressed into pellets and sintered at 850 8C for 12 h in 0.5 bar of argon. The thin films were deposited on R-cut (1 1¯ 0 2) sapphire substrates maintained at 600 8C using a KrF excimer laser operating at 248 nm and 10 Hz. Laser fluence on the target was 1.8 J cm2. The targetsubstrate distance was 35 mm and the oxygen pressure was 104 mbar. Phase analysis was carried out using X-ray diffraction (XRD) and scanning electron microscopy (SEM), with energy dispersive X-ray analysis (EDAX) for elemental analysis and mapping. Magnetization measurements were made in a SQUID magnetometer at 300 K. Film thickness was monitored during deposition using optical reflectivity at 635 nm, and it was independently calibrated by low angle X-ray reflectivity (XRR) measurements. Optical transmission spectra were measured between 1.65 and 3.95 eV to determine the band-gap and the electronic state of cobalt ions. Electrical resistivity measurements were carried out by the van der Pauw method in the temperature range 300 K. Elemental maps of the targets obtained by energydispersive X-ray diffraction (EDAX) all show the nominal transition metal concentration, with a uniform cobalt distribution. X-ray diffraction confirms the wurtzite structure, with no impurity phases detected (Fig. 1). Targets exhibited weak ferromag-
Fig. 1. X-ray diffraction pattern of the Zn0.95Co0.05O ceramic target, and a corresponding thin film deposited by pulsed-laser deposition. The film is well oriented with the (1 1 0) texture: ‘‘S’’ indicates substrate peaks.
Fig. 2. Plot of cobalt concentration in films vs. target composition.
netism at room temperature, with moments of no more than 0.02 mB/Co atom. The thin films are transparent with a pale green tinge. The X-ray diffraction patterns (Fig. 1) show the single-phase hexagonal wurtzite structure with welloriented (1 1 0) texture. The lattice parameter ‘a’ stays almost constant with increasing cobalt concentration, as expected for Co2+. Film thickness was typically 65 nm. EDAX analysis indicates a homogeneous cobalt distribution within the film, but the dopant concentration is greater in thin films than in the targets (Fig. 2). All Co-doped ZnO films are ferromagnetic at room temperature, with coercivity of 2–3 mT. The variation of the magnetic moment with cobalt concentration is shown in Fig. 3. The largest measured moment is
Fig. 3. Magnetic moment of Zn1xCoxO films measured at room temperature, with the moment expressed as mB/Co.
C.B. Fitzgerald et al. / Applied Surface Science 247 (2005) 493–496
Fig. 4. Plot of resistivity vs. temperature for Zn1xCoxO (x = 0.07, 0.14).
5.9 mB/Co, compared to the spin-only moment mspin = 3 mB Co2+ in the high spin d7 configuration e4t23. The moment drops as x increases, which can be understood in terms of a random distribution of cobalt ions over the cation sites in the wurtzite lattice. Isolated ions contribute the full moment, pairs and most groups of four are antiferromagnetically coupled and make no net contribution, triplets contribute mspin/ 3. Large antiferromagnetically coupled clusters of N atoms will make a contribution of mspin/N1/2. The model [18] reproduces the trend, but suggests a slightly greater tendency to form more cobalt neighbour pairs than a purely random distribution would provide. Films exhibit semiconducting behaviour, with resistivity increasing with increasing cobalt content (Fig. 4). Room temperature resistivity for the x = 0.07 sample is 200 V cm. They are n-type conductors, where the carriers are probably associated with oxygen vacancies. Optical transmission spectra for Zn1xCoxO (x = 0, 0.07, 0.14, 0.23) thin films are compared in Fig. 5. The band edge for the undoped sample appears at 3.27 eV, while the band edge of the doped samples shift to progressively lower energies with x. The red shift with increasing cobalt content could be attributed to the sp– d exchange interactions between the band electrons and the localized d electrons of the Co2+ ions [19]. Also, as the cobalt content increases, so does the absorption band intensity. Clear evidence that cobalt enters the tetrahedral sites of the wurtzite structure as Co2+ is seen as transitions in the spectra at 1.88, 2.04
495
Fig. 5. Optical transmission spectra of Zn1xCoxO thin films (x = 0, 0.07, 0.14, 0.23).
and 2.2 eV, assigned as d–d absorption levels of Co2+ ions in a tetragonal crystal field [20]: 4A2 ! 4T1(4P) and 4A2 ! 2E(2G) transitions. A debate as to whether these dilute oxide hightemperature ferromagnets are truly homogeneous and single-phase is ongoing. Here, we have shown that cobalt does enter the ZnO structure as Co2+ rather than forming metallic clusters, and that no secondary phases are observed by X-ray diffraction or scanning electron microscopy. The recent model proposed to explain ferromagnetism in dilute n-type magnetic semiconductors [18] suggests that the exchange is mediated by carriers in a spinsplit impurity band derived from extended donor orbitals. In conclusion, we have found that thin films of Zn1xCoxO are ferromagnetic and semiconducting at room temperature, with a moment that drops sharply with increasing cobalt concentration. Room temperature resistivity is 200 V cm. The cobalt-doped ZnO films appear to be homogeneous and single-phase materials, where cobalt enters the ZnO structure as Co2+ rather than forming metallic clusters. The large moments at low x are incompatible with any known impurity phase.
Acknowledgement This work was supported by Science Foundation Ireland.
496
C.B. Fitzgerald et al. / Applied Surface Science 247 (2005) 493–496
References [1] H. Ohno, Science 281 (1998) 951. [2] K. Ueda, H. Tabata, T. Kawai, Appl. Phys. Lett. 79 (2001) 988. [3] Y.M. Cho, W.K. Choo, H. Kim, D. Kim, Y.E. Ihm, Appl. Phys. Lett. 80 (2001) 3358. [4] H.J. Lee, S.Y. Jeong, C.R. Cho, C.H. Park, Appl. Phys. Lett. 81 (2002) 4020. [5] K. Rode, A. Anane, R. Mattana, J.P. Contour, et al. J. Appl. Phys. 93 (2003) 7676. [6] W. Prellier, A. Fouchet, B. Mercey, C. Simon, B. Raveau, Appl. Phys. Lett. 82 (2003) 3490. [7] D.P. Norton, M.E. Overberg, S.J. Pearton, K. Preusser, et al. Appl. Phys. Lett. 83 (2003) 5488. [8] K. Ando, H. Saito, Z-W. Jin, T. Kukumura, et al. Appl. Phys. Lett. 78 (2001) 2700. [9] H. Saeki, H. Tabata, T. Kawai, Solid-State Commun. 120 (2001) 439. [10] P. Sharma, A. Gupta, K.V. Rao, F.J. Owens, R. Sharma, R. Ahuja, J.M. Osorio Guillen, B. Johansson, G.A. Gehring, Nature Mater. 2 (2003) 673.
[11] P.V. Radovanovic, D.R. Gamelin, Phys. Rev. Lett. 91 (2003) 157202. [12] T. Dietl, H. Ohno, F. Matsukura, J. Cibe`rt, D. Ferrand, Science 287 (2000) 1019. [13] Z.-W. Jin, T. Fukumura, K. Hasegawa, et al. J. Cryst. Growth 548 (2002) 237–239. [14] S.-W. Lim, D.-K. Hwang, J.-M. Myoung, Solid-State Commun. 125 (2003) 23. [15] N. Jedrecy, H.J. von Bardeleben, Y. Zheng, J.L. Cantin, Phys. Rev. B 69 041308 (2004). [16] F. Matsukura, H. Ohno, A. Shen, Y. Sugawara, Phys. Rev. B 57 (1998) R2037. [17] J.M.D. Coey, M. Viret, S. von Molnar, Adv. Phys. 48 (1999) 167. [18] J.M.D. Coey, M Venkatesan and C.B. Fitzgerald, Nature Materials, in press. [19] J. Diouri, J.P. Lascaray, M. El Amrani, Phys. Rev. B 31 (1985) 7995. [20] P. Koidl, Phys. Rev. B 15 (1977) 2493.
