Ferromagnetic nanoclusters formed by Mn implantation in GaAs O. D. D. Couto Jr., M. J. S. P. Brasil, F. Iikawa, C. Giles, C. Adriano, J. R. R. Bortoleto, M. A. A. Pudenzi, H. R. Gutierrez, and I. Danilov Citation: Applied Physics Letters 86, 071906 (2005); doi: 10.1063/1.1863436 View online: http://dx.doi.org/10.1063/1.1863436 View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/86/7?ver=pdfcov Published by the AIP Publishing This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 186.217.234.100 On: Wed, 05 Feb 2014 16:42:24 http://scitation.aip.org/content/aip/journal/apl?ver=pdfcov http://oasc12039.247realmedia.com/RealMedia/ads/click_lx.ads/www.aip.org/pt/adcenter/pdfcover_test/L-37/1159426268/x01/AIP-PT/APL_ArticldDL_012214/aipToCAlerts_Large.png/5532386d4f314a53757a6b4144615953?x http://scitation.aip.org/search?value1=O.+D.+D.+Couto+Jr.&option1=author http://scitation.aip.org/search?value1=M.+J.+S.+P.+Brasil&option1=author http://scitation.aip.org/search?value1=F.+Iikawa&option1=author http://scitation.aip.org/search?value1=C.+Giles&option1=author http://scitation.aip.org/search?value1=C.+Adriano&option1=author http://scitation.aip.org/search?value1=J.+R.+R.+Bortoleto&option1=author http://scitation.aip.org/search?value1=M.+A.+A.+Pudenzi&option1=author http://scitation.aip.org/search?value1=H.+R.+Gutierrez&option1=author http://scitation.aip.org/search?value1=H.+R.+Gutierrez&option1=author http://scitation.aip.org/search?value1=I.+Danilov&option1=author http://scitation.aip.org/content/aip/journal/apl?ver=pdfcov http://dx.doi.org/10.1063/1.1863436 http://scitation.aip.org/content/aip/journal/apl/86/7?ver=pdfcov http://scitation.aip.org/content/aip?ver=pdfcov Ferromagnetic nanoclusters formed by Mn implantation in GaAs O. D. D. Couto, Jr., M. J. S. P. Brasil, and F. Iikawa Instituto de Física “Gleb Wataghin,” UNICAMP, C-6167, 13083-970, Campinas-SP, Brazil C. Giles Instituto de Física “Gleb Wataghin,” UNICAMP, C-6167, 13083-970, Campinas-SP, Brazil and Laboratório Nacional de Luz Síncrotron, CP-6192, 13084-971, Campinas-SP, Brazil C. Adriano, J. R. R. Bortoleto,a! M. A. A. Pudenzi, and H. R. Gutierrezb! Instituto de Física “Gleb Wataghin,” UNICAMP, C-6167, 13083-970, Campinas-SP, Brazil I. Danilovc! Departamento de Física, UFRGS, CP-15051, 91501-970, Porto Alegre-RS, Brazil sReceived 24 September 2004; accepted 16 December 2004; published online 8 February 2005d Ferromagnetic clusters were incorporated into GaAs samples by Mn implantation and subsequent annealing. The composition and structural properties of the Mn-based nanoclusters formed at the surface and buried into the GaAs sample were analyzed by x-ray and microscopic techniques. Our measurements indicate the presence of buried MnAs nanoclusters with a structural phase transition around 40 °C, in accord with the first-order magneto-structural phase transition of bulk MnAs. We discuss the structural behavior of these nanoclusters during their formation and phase transition, which is an important point for technological applications. ©2005 American Institute of Physics. fDOI: 10.1063/1.1863436g The possibility of preparing submicron ferromagnets in semiconductors is of strong interest to the development of new spintronic devices.1,2 The formation of Mn-based nano- clusters in GaAs samples by ion implantation and subsequent annealing3–5 is a relatively simple technique compared to the growth of sGaAsdMn diluted material by epitaxial tech- niques, with the advantages of easy control of the Mn rela- tive density and the possibility of spatial confinement with mask patterning. Both GaMnsRefs. 3,4d and MnAs sRefs. 6,7d present room temperature ferromagnetism. MnAs exhib- its a simultaneous structural and magnetic first-order phase transition at ,40 °C, from hexagonal-ferromagnetic sa-phased to orthorhombic-paramagneticsb-phased.6,7 A re- markable aspect from the structural point of view is that the Mn nanostructures obtained by ion implantation present good crystal quality despite the structural difference between those structures and the GaAs matrix. Both MnAs sorthorhombic/hexagonald6,7 and GaMn sicosahedrald8 ex- hibit lattice structures that are markedly different from the cubic GaAs. The dynamics of formation of Mn nanoclusters remains mostly unknown. A detailed analysis of this system requires several complex techniques. We investigate the structure and composition of mag- netic nanoclusters created by Mn implantation on GaAs samples with special emphasis on MnAs clusters and their properties. Our results indicate the presence of MnAs nano- clusters embedded in the GaAs sample presenting a struc- tural phase transition typical of bulk MnAs with a substantial variation of their lattice constants despite severe three- dimensional constraints imposed by the GaAs matrix. Semi-insulating GaAss001d substrates were implanted with Mn+ ions with 200 keV and doses of 1–331016 cm−2. Subsequent rapid thermal annealingsRTAd was performed at 700–900 °C for 2–60 s in Ar atmosphere with the samples covered with a Si plate to minimize As evaporation. X-ray diffraction measurements were performed at the Brazilian National Synchrotron Light LaboratorysLNLSd using a 7600 eV energy beam with the sample temperature controlled by a Peltier thermoelectric device. Atomic force microscopy sAFMd images were used to spatially probe the topography of the samples surface. For structural and composition analy- sis, cross-sectional transmission electron microscopysTEMd and energy dispersive x-raysEDXd measurements were em- ployed with a JEM-3010 URP microscope operating at 300 kV. The depth distribution of Mn was investigated by sec- ondary ion mass spectrometrysSIMSd measurements. After annealing we observed a strong diffusion of Mn towards the sample surface resulting in the formation of sur- face clusters with a ferromagnetic character. The AFM image from a typical samplefFig. 1sadg shows nanoclusters with a mean diameter of,200 nm, mean height of,15 nm and average density of,2.53108 cm−2. TEM measurements re- vealed that the surface nanoclusters may assume various dis- tinct shapes, as shown by the cross-sectional images pre- sented in Figs. 1sbd and 1scd. Point 1 in Fig. 1sbd marks a shallow surface structures,20 nm heightd, whereas the structure presented on Fig. 1scd spoint 4d is much larger s,70 nm heightd and shows more defined facets. Further- more, we observe the presence of buried nanoclusters as a series of nearly circular structures with diameters of the or- der of ,50 nm along a line parallel to the sample surface fpoint 3 in Fig. 1sbdg. The buried nanoclusters are formed at a depth of,150 nm below the sample surface. This agrees with the mean projected range of the Mn ions simulated by the TRIM code and with the local maximum of Mn concen- tration observed by SIMS even for annealed samples. Similar results were previously reported.9 adPresent address: UNESP, U. D. Sorocaba/Iperó, Sorocaba 18087-180, SP - Brazil. bdPresent address: Penn State University, Department of Physics, Ins. Mat. Res., University Park, Pennsylvania 16802. cdPresent address: Nizhny Novgorod Sate University, 603950 Nizhny Novgorod, Russia. APPLIED PHYSICS LETTERS86, 071906s2005d 0003-6951/2005/86~7!/071906/3/$22.50 © 2005 American Institute of Physics86, 071906-1 This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 186.217.234.100 On: Wed, 05 Feb 2014 16:42:24 http://dx.doi.org/10.1063/1.1863436 http://dx.doi.org/10.1063/1.1863436 EDX spectra at the points marked by numbers on Figs. 1sbd and 1scd are presented in Fig. 1sdd. The composition analysis is semiqualitative due to the large uncertainty origi- nating from the signal integration over a relatively large vol- ume, defined by the area of the electron beam probesdiam- eter of,20 nmd and the sample thickness prepared for EDX measurementss,100 nmd. When this volume is larger than that of the nanocluster, the signal includes a non-negligible contribution from the surrounding GaAs matrix that must be taken into account. Emission from Mn atoms was observed for all nanoclusters, but no Mn emission was detected from regions adjacent to nanostructuresfsuch as point 2 in Fig. 1sbdg indicating that the Mn is strongly concentrated at the nanoclusters. We estimated the Ga/As ratio of the nanoclus- ters considering the ratio between the emission intensities from Ga and As for their surrounding regions as a reference of a pure GaAs material and taking into account their dimen- sions relative to the probe volume. The buried structures fpoint 3 in Fig. 1sadg present a relatively strong EDX signal from Mn atoms and a Ga/As composition ratio smaller than that of the GaAs reference. The Ga emission corresponds approximately to the signal expected from the surrounding GaAs region of those relatively small clusters. Therefore they must be composed of pure MnAs or asGa,AsdMn alloy with a residually low Ga concentration. On the other hand, the EDX spectrum for the larger surface structurefpoint 4 in Fig. 1scdg shows almost no emission from As atoms, indicat- ing that those structures are composed mainly of GaMn. The uncertainty is obviously larger for the smaller surface struc- turesfpoint 1 at Fig. 1sbdg for which EDX results indicate a sGa,AsdMn alloy with a As concentration slightly larger than Ga. The inset of Fig. 2 shows a x-ray diffraction spectrum from a Mn implanted GaAs sample measured at room tem- perature in a grazing-incidence diffractionsGIDd configura- tion. The dominant peak at 2u=48.2° corresponds to the GaAs s220d diffraction and the various additional peaks are attributed tosGa,AsdMn alloys with different compositions. The peak at 2u,52.2° coincides with thes112̄0d reflection of a-MnAs,6,7 giving evidence to the presence of pure MnAs nanoclusters. Based on EDX results, we believe that those MnAs nanoclusters are buried in the GaAs matrix. Figure 2 shows x-ray diffraction patterns for a heating cycle from 14 to 81 °C at this diffraction peak. At low temperatures the peak is at,52.15 °; between 40 °C and 50 °C it shifts to larger angles and for higher temperatures it is at,52.33 °. All other diffraction peaks of the sample remain basically constant with temperature, showing only a slight shift to smaller angles, i.e., larger lattice parameters, due to thermal expansion. The behavior is reversed in a cooling cycle and is fully reproducible for various cycles. The temperature dependence of the basal-lattice param- eter of the hexagonal structurea obtained from thes112̄0d MnAs diffraction peak is compared in Fig. 3 with the corre- sponding bulk MnAs values extracted from Ref. 6. The basal-lattice constant from the nanoclusters also presents a significant variation around the bulk phase transition tem- FIG. 1. sad AFM image showing the nanoclusters formed at the surface of a Mn implanted GaAs sample with sub- sequent annealing.sbd and scd Cross- sectional TEM image on Mn- implanted samples. The points indicated by numbers correspond to the positions analyzed by EDX.sdd EDX emission spectra for the points marked at the TEM images. FIG. 2. Thermal variation of the x-ray diffraction patterns of the MnAs s112̄0d plane from the nanoclusters. The inset shows a x-ray diffraction pattern obtained from the same sample at room temperatures,20 °Cd for a larger 2u range. 071906-2 Couto, Jr. et al. Appl. Phys. Lett. 86, 071906 ~2005! This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 186.217.234.100 On: Wed, 05 Feb 2014 16:42:24 perature. The magnitude of the lattice parameter variation is, however, very different for the two systems. As compared to the bulk values, the value ofa obtained for the nanoclusters is larger at theb-phase sT.50 °Cd and smaller at the a-phasesT,30 °Cd, so that its variation during the phase transitions,0.3%d is much smaller than the corresponding bulk variations,1%d. Furthermore, the nanoclusters lattice parameter presents a thermal hysteresis, i.e., the phase tran- sition occurs at higher temperatures during heating as com- pared to a cooling cycle, and is not very abrupt, which is attributed to the coexistence of the two phases. This coexist- ence can arise from a single nanostructure and/or a distribu- tion of purea- andb-phase nanostructures and its origin is still a subject of investigation. Both the coexistence and the hysteresis have been observed in previous works10–12on thin MnAs films grown on GaAs and were attributed to strain effects. We also point out that the thermal expansion/contraction of the MnAs nanoclusters when they are completely at a-phasesT,30 °Cd or b-phasesT.50 °Cd does not follow the corresponding bulk behavior, which is clearly observed in Fig. 3. Instead, the basal lattice parameter of the Mn nano- clusters shows a thermal variation similar to GaAs. The dashed line in Fig. 3 was plotted using the GaAs thermal expansion coefficient and fits quite well the thermal variation of the basal-lattice constant forb-phase nanoclusters. In or- der to understand this result, we must analyze the nanoclus- ter formation. MnAs nanoclusters are formed during the ther- mal annealing at high temperaturess700–900 °Cd, where all strain due to the distinct lattice structures between MnAs and GaAs must be released. As the sample is cooled, a strain should develop due to the large difference between the ther- mal expansion coefficients of MnAss,6310−5/Kd sRef. 6d and GaAss,7310−6/Kd.13 At relatively high temperatures, the strain can be efficiently accommodated due to the ther- mal energy of the system, but we expect that at a certain temperature the thermal energy becomes insufficient to re- lieve the strain. Below this temperature, the MnAs nanoclus- ter must follow the GaAs thermal expansion coefficient and strain will therefore accumulate at the nanocluster. Since the GaAs thermal expansion coefficient is smaller than that of MnAs, the basal-lattice constant from the MnAs nanostruc- tures cooled to,80 °C sb-phased should end up larger than that of bulk MnAs, as observed. The temperature from which strain starts to build up at the MnAs nanostructure can be estimated by the crossing point of the extrapolation of the bulk and the nanoclusters data, which gives,160 °C, a very feasible value. The fact that the first-order magneto-structural phase transition of MnAs persists for MnAs nanoclusters with di- mensions of the order of 50 nm embedded in a GaAs matrix is a novel result that may have very interesting conse- quences. The expansion/contraction of buried MnAs nano- clusters during phase transition should be strongly con- strained by the GaAs matrix. In fact, an ideal nanocluster surrounded by an infinitely larger GaAs matrix should not present a structural transition at all, being obliged to follow the thermal variation of the GaAs matrix. The weakened, but nonzero, basal-lattice constant variation displayed by the MnAs nanoclusters may be attributed to a residual degree of freedom from lattice defects formed around the nanocluster and a partial relief of strain to the surrounding GaAs mate- rial. Strain effects significantly affect the stability of the fer- romagnetica-phase MnAs nanoclusters and they must there- fore be considered in the development of device applications based on such nanostructures. We kindly acknowledge the financial support from FAPESP, CAPES, and CNPq. TEM measurements were made at the LME laboratory of the Brazilian National Syn- chrotron Light LaboratorysLNLSd, Campinas-SP, Brazil. 1G. A. Prinz, Science250, 1092s1990d. 2G. A. Prinz, Phys. Today48, 58 s1995d. 3J. Shi, J. M. Kikkawa, R. Proksch, T. Schaeffer, D. D. Awschalom, G. Medeiros-Ribeiro, and P. M. Petroff, NaturesLondond 377, 707 s1995d. 4J. Shi, J. M. Kikkawa, D. D. Awschalom, G. Medeiros-Ribeiro, P. M. Petroff, and K. Babcock, J. Appl. Phys.79, 5296s1996d. 5C. Chen, M. Cai, X. Wang, S. Xu, M. Zhang, X. Ding, and Y. Sun, J. Appl. Phys. 87, 9 s2000d; 87, 5636s2000d. 6B. T. M. Willis and H. P. Rooksby, Proc. Phys. Soc. London, Sect. B67, 290 s1954d. 7R. H. Wilson and J. S. Kasper, Acta Crystallogr.17, 95 s1964d. 8J. P. Zhang, A. K. Cheetham, K. Sun, J. S. Wu, K. H. Kuo, J. Shi, and D. D. Awschalom, Appl. Phys. Lett.71, 143 s1997d. 9A. Serres, G. Benassayag, M. Respaud, C. Armand, J.-C. Pesant, A. Mari, Z. Liliental-Weber, and A. Claverie, Mater. Sci. Eng., B101, 119 s2003d; A. Serres, M. Respaud, G. Benassayag, C. Armand, J.-C. Pesant, A. Mari, Z. Liliental-Weber, and A. Claverie, Physica EsAmsterdamd 17, 371 s2003d. 10V. M. Kaganer, B. Jenichen, F. Schippan, W. Braun, L. Däweritz, and K. H. Ploog, Phys. Rev. Lett.85, 341 s2000d. 11V. M. Kaganer, B. Jenichen, F. Schippan, W. Braun, L. Däweritz, and K. H. Ploog, Phys. Rev. B66, 045305s2002d. 12F. Iikawa, P. V. Santos, M. Kästner, F. Schippan, and L. Däweritz, Phys. Rev. B 65, 205328s2002d. 13Landoldt-Börnstein, New Series, III/17a, 234. FIG. 3. Temperature dependence of the basal lattice constanta obtained for the MnAs nanoclusterssopen circlesd and for bulk MnAssdashed lined sRef. 6d. The continuous line is a linear fit for theb-phase lattice parameter from the nanoclusters using an angular coefficient corresponding to the GaAs thermal expansion coefficients,7310−6/Kd sRef. 13d. 071906-3 Couto, Jr. et al. Appl. Phys. Lett. 86, 071906 ~2005! This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 186.217.234.100 On: Wed, 05 Feb 2014 16:42:24