NANO LETTERS
Controlled Fabrication of Epitaxial (Fe,Mn)3O4 Artificial Nanowire Structures and their Electric and Magnetic Properties
2009 Vol. 9, No. 5 1962-1966
Kazuya Goto, Hidekazu Tanaka,* and Tomoji Kawai The Institute of Scientific and Industrial Research, Osaka UniVersity, 8-1 Mihogaoka, Ibaraki, Osaka 567-0047, Japan Received January 16, 2009; Revised Manuscript Received March 6, 2009
ABSTRACT Epitaxial (Fe,Mn)3O4 (FMO) ferromagnetic oxide artificial nanowire (NW) structures were deposited across Pt/Cr bilayer electrodes with a high controllability on their shape and positioning by using atomic force microscope (AFM) lithography with molybdenum (Mo) lift-off in combination with a pulsed laser deposition technique. The oxide wire widths were systematically controlled from 5 µm down to 100 nm by controlling AFM lithography current for Mo-mask. The resistivity of FMO-NW structures was increased at below around 400 nm in width. Magnetic force microscope revealed that FMO-NW with 120 nm showed single line of aligned ferromagnetic domains that caused the resistivity increase. Our results offer well-defined epitaxial transition metal oxide nanostructures with a remarkable flexibility toward nanoscale oxide spintronics.
Transition metal oxides are important functional materials for novel electronics due to their unique properties such as perfect spin polarization,1,2 huge metal-insulator transition,3 ferroelectric,4 multiferroic,4 resistive switching effect,5 and so on. For example, magnetite (Fe3O4) and its derivatives of (Fe,Mn)3O4 (FMO) and (Fe,Zn)3O4 (FZO) are promising in spintronics application due to their high spin polarization with very high Curie temperature (TC). 1,2,6-8 Recently, low dimensional structures, typically a nanowire (NW) structure with one-dimension, are attracting much attention because of their unique properties as compared with bulk properties.9-16 Moreover, complex low dimensional nanostructures, which are connected with electrodes for realistic nanodevice applications such as a crossbar nanojunction or a Y-shape nanojunction, have demonstrated their high potentials for nanoelectronics and nanospintronics device applications by using semiconductors (Si, GaN, and soforth),12,15-17 magneticsemiconductors(Ge1-xMnx,(Ga,Mn)N, and so forth)18-21 and ferromagnetic metals (NiFe, CrO2, and so forth).22,23 Namely, constructions of “complex onedimensional (1D) functional oxide structures” with wellcontrolled shapes and positions are indispensable for not only creations of novel physical properties but also realistic applications of nano-oxide-device. NW fabrication can be divided into two approaches: namely top-down and bottom-up approaches. In the bottom-
up approach, NWs grown by self-assembly of materials (e.g., vapor-liquid-solid technique) are separated from the original substrate and dispersed onto target substrates with an uncontrollable high degree of freedom in various positions. Their accuracy is not enough to control the positions of NWs on the target substrates. For realizing a high density and high performance of nano-oxide device, it is important to artificially control NW structures with high accuracy of the shapes and positions. At this point of view, the top-down approach has some advantages on its perfect control over the shapes and positions of NW structures although the sizes of nanostructures produced using the top-down approach are generally larger than that using the bottom-up approach. Nevertheless, it is very difficult to control both their positions and shapes, and there are few reports on the complex NW structures of transition metal oxides artificially designed.
* To whom correspondence should be addressed. Tel: +81-6-6879-4280. Fax: +81-6-6879-4283. E-mail:
[email protected].
The key point of this method is that NW structures are able to finally be constructed by “deposition” with a high
10.1021/nl900158t CCC: $40.75 Published on Web 04/16/2009
2009 American Chemical Society
In this letter, in order to overcome this problem, we have demonstrated fabrication of epitaxial room temperature ferromagnetic FMO-NW structures with artificially designed and deposited across Pt/Cr electrodes (namely, artificial NW), using atomic force microscope (AFM) lithography with molybdenum (Mo) lift-off in combination with pulsed laser deposition (PLD) technique. This lithography has a remarkable accuracy and flexibility in control of their shapes and positions.
accuracy across Pt/Cr electrodes initially prepared. In conventional “removal” top-down approach such as focused ion beam lithography24 and electron beam lithography,10,22 electrodes are deposited after NWs are fabricated, leading to low accuracy of position, and low degree of freedom on further integration such as heterostructuring. We prepared a nanoscale metal template that was applicable for a metal oxide epitaxial growth by using AFM lithography, which guarantees very high accuracy by controlling AFM lithography current. Figure 1 shows schematic illustrations of the fabrication procedure of FMO-NW structures across Pt/Cr electrodes. The height scale bars of AFM images (in the insets of Figure 1b,c,e) are 30 nm. First, Pt (20 nm)/Cr (1-2 nm) bilayer electrode arrays were fabricated on Al2O3 (0001) substrate by RF sputtering using Ar gas at 50W of RF power, 8 sccm of Ar flow rate and room temperature. The Pt/Cr electrode sizes were 50 µm × 50 µm and their separation distances were 50 µm (Figure 1a). Second, Mo film was deposited on the Pt/Cr electrode substrate by DC sputtering with Ar gas at room temperature. The film thickness of Mo was 15 nm. The surface imaging and lithography were performed by AFM microscope (Dimension 3100, Veeco, U.S.A.) with Veeco NCHV-10V AFM tips resonant frequency of 320 kHz and a spring constant of 42 N/m. During AFM lithography, humidity was set to between 50 and 60%. The tip velocity was set to 1 µm/s and the initial tip voltage was set to -5 V. The flowing current between the tip and the Mo surface was adjusted in order to control resulting FMO wire widths. After performing AFM lithography, the exposed Mo part becomes a MoO3 line through a local anodic oxidation (Figure 1b).25 This MoO3 line was selectively removed by H2O with ultrasonic treatment for 10 min. After removing the MoO3, a nanotrench across Pt/Cr electrodes was fabricated (Figure 1c). Then, ferromagnetic oxide semiconductor of Fe2.5Mn0.5O4 was deposited on the patterned substrate by PLD (ArF Excimer: λ ) 193nm, COMPex 102 FAR, Lambda-Physik, Germany) technique (Figure 1d). The substrate temperature was set to 320 °C and oxygen pressure was 1.0 × 10-4 Pa during the deposition. The thickness and length of FMO-NW structure were fixed as 10 nm and 50 µm, respectively. After removing Mo layer by 30% H2O2 solution with ultrasonic treatment, FMO-NW structure was finally fabricated across Pt/Cr electrodes (Figure 1e). Figure 2a shows an optical micrograph of first prepared Pt/Cr electrode arrays. FMO-NW structures were fabricated across Pt/Cr electrodes as a demonstration of position control, and the distances between each Pt/Cr electrodes were 50 µm. Therefore, the length of FMO-NW structure was consequently fixed to 50 µm. Figure 2b-d shows the optical micrographs of the position controlled FMO-NW structures with various widths fabricated across Pt/Cr electrodes by controlling the AFM lithography current for Mo-mask. The 460 nm, 1.2 µm, and 2.2 µm of FMO-NW width were fabricated at 100, 500, and 790 pA, respectively (Figure 2b-d). Figure 2e summarizes the wire width versus AFM lithography current. The wire width was proportionally decreased as a function of AFM lithography current from 5 Nano Lett., Vol. 9, No. 5, 2009
Figure 1. Schematic illustrations of fabrication procedure for epitaxial grown FMO-NW structures. (a) Pt/Cr electrode are deposited on Al2O3 (0001) substrate. (b) Sputtering Mo and performing AFM lithography. The exposed Mo part becomes a MoO3 line by local anodic oxidation. (c) The MoO3 line is removed by H2O with ultrasonic treatment. (d) FMO is deposited by PLD technique. (e) Mo mask layer is removed by 30% H2O2 solution with ultrasonic treatment. Finally, a FMO-NW structure is fabricated. The height scale bars of AFM images are 30 nm. 1963
Figure 2. (a) Optical micrograph of Pt/Cr electrode arrays. Optical micrographs of (b) 460 nm width of FMO-NW at 100 pA of AFM lithography current, (c) 1.2 µm at 500 pA and (d) 2.2 µm at 790 pA, respectively. (e) Plots of wire width as a function of AFM lithography current.
µm to 100 nm. This result was consistent with the mechanism of a local anodic oxidation method for a metal nanotemplate.26-28 The widths of the NWs we fabricated were 1 order of magnitude larger than that of NWs that were produced using the self-assembly processes (either by catalytic growth or by chemical methods). Although taking into account that our method could not be presented as an alternative to the bottom-up approach (self-assembly) of the NW formation, we constructed an FMO-NW structure between initially prepared Pt/Cr electrodes with systematic control of their width by top-down AFM lithography technique in combination with bottom-up thin film growth technique (PLD). Figure 3a shows current-voltage (I-V) characteristics of the FMO-NW structures between Pt/Cr electrodes. The wire widths were 2.2 µm (blue line), 1.2 µm (red line), 460 nm (black line), 200 nm (gray line), and 100 nm (green line). Slopes (resistances) were decreased (increased) with decreasing widths. The inset of Figure 3a shows a 3D AFM image of a 100 nm width FMO-NW structure with flat surface without any grain structure. In addition, we have already confirmed that FMO nanostructures were epitaxially grown on Al2O3 (0001) substrate by X-ray diffraction (pole figure) measurement14 and exhibited a fine electronic structure, which was the same as bulk film after doing Mo lift-off process.29 Therefore, epitaxial FMO-NW structures grew with the same high quality as epitaxial thin film in structural and electronic properties. Figure 3b shows FMO-NW resistivity versus wire width. The FMO wire resistivity from 5 µm to around 400 nm was of similar value as the bulk film resistivity (dash line: 3.2 × 10-1 Ω·cm). However, it starts to narrowly increase around 400 nm. Figure 4a,b shows magnetic force microscope (MFM) images of 120 and 630 nm width FMO-NW. The 120 nm width FMO-NW was found to have “Pea-Pod”-like magnetic contrast, which indicated the existence of a chain structure 1964
Figure 3. (a) Current-Voltage (I-V) characteristics of FMO-NW structures. The inset shows a 3D AFM image of epitaxial FMONW structure with 100 nm width. (b) Resistivity of FMO wire as a function of wire width.
Figure 4. MFM images (a) 120 nm and (b) 630 nm width of FMONW. Conduction models of spin-polarized electron scattering (c) 120 nm width and (d) 630 nm width of FMO-NW.
of single ferromagnetic domains along wire direction, whereas the 630 nm width FMO-NW showed randomly assembled structure of many ferromagnetic domains. Figure 4c,d shows conduction models of spin scattering at magnetic domain wall in each samples. The red and blue regions are magnetic domains with different magnetization directions from each other, and domain walls are in their boundaries. Nano Lett., Vol. 9, No. 5, 2009
In conclusion, we have fabricated epitaxial artificial FMONW structures across Pt/Cr bilayer electrodes by using AFM lithography with Mo lift-off technique and PLD method. The wire widths were controllable from 5 µm to 100 nm by adjusting AFM lithography current. The resistivity of FMONW was increased below around 400 nm width. A MFM measurement revealed that the 120 nm width FMO-NW structure alternatively trapped the single ferromagnetic domains along the confined wire direction. Using this method, the two step widths of epitaxial FMO-NW structure connected to metal electrodes with a high position accuracy can be also fabricated. These results indicate that we are able to completely control the shape and position of epitaxial oxide nanostructure as designed. This method will open up a new field of well-defined functional oxide nanoelectronics and nanospintronics devices taking advantage of its high controllability of shape and position. Figure 5. 3D AFM image of the two step FMO-NW structure connected to metal electrodes.
Domain walls scatter spin polarized electrons, which was leading to an increase of resistivity. In the case of the 120 nm width FMO-NW, small ferromagnetic domains about 100 nm in diameter were alternatively aligned and connected as a one-dimension. Therefore, spin-polarized electrons should pass all domain walls in 1D conduction pass, which causes increasing resistivity (Figure 4c). In the case of the 630 nm width FMO-NW, magnetic domains were randomly arranged as two-dimensions; larger domains easily appear, so that spin polarized electrons are less scattered (i.e., carrier electrons can escape domain walls as much as possible in 2D conduction pass; this situation is the same as large scale thin film), and the resistivity was the same value as the thin film (Figure 4d). Therefore, we attribute the reason of increasing resistivity below 400 nm widths to the domain wall scattering effect in one-dimensionally confined FMO-NWs. As remarked above, we alternatively aligned single ferromagnetic domains in the confined 1D ferromagnetic oxide NW structure as shown in Figure 4a. This phenomenon had been observed in ferromagnetic metal materials30-32 but had not been observed in high TC ferromagnetic oxide material yet. This result will open up to develop a novel lowdimensional spin device such as a NW domain wall device or nanodot spin device with transition metal oxides which exhibited high TC ferromagnetism and a huge response for an external field that originated from a strongly correlated electron system.33 Finally, the artificial “complex” oxide NW structure was also fabricated as a demonstration toward further complicated oxide devices. Figure 5 shows a 3D AFM image of two step widths of epitaxial FMO-NW structure. The wide (2.2 µm) and the narrow (200 nm) parts were fabricated by applying 220 and 20 pA of AFM lithography current, respectively. Both edge parts were accurately positioned and connected to metal electrodes. Thus, we have artificially controlled the position where the epitaxial NW structure was grown as well as its shape with a very high accuracy and flexibility. Nano Lett., Vol. 9, No. 5, 2009
References (1) Dedkov, Y. S.; Ru¨diger, U.; Gu¨ntherodt, G. Phys. ReV. B 2002, 65, 064417. (2) Reisinger, D.; Majewski, P.; Opel, M.; Alff, L.; Gross, R. Appl. Phys. Lett. 2004, 85, 4980–4982. (3) Lee, Y. W.; Kim, B. -J.; Lim, J. -W.; Yun, S. J.; Choi, S.; Chae, B. -G.; Kim, G.; Kim, H. -T. Appl. Phys. Lett. 2008, 92, 162903. (4) Wang, J.; Neaton, J. B.; Zheng, H.; Nagarajan, V.; Ogale, S. B.; Liu, B.; Viehland, D.; Vaithyanathan, V.; Schlom, D. G.; Waghmare, U. V.; Spaldin, N. A.; Rabe, K. M.; Wuttig, M.; Ramesh, R. Science 2003, 299, 1719–1722. (5) Lee, M. -J.; Park, Y.; Suh, D. -S.; Lee, E. -H.; Seo, S.; Kim, D. -C.; Jung, R.; Kang, B. -S.; Ahn, S. -E.; Lee, C. B.; Seo, D. H.; Cha, Y. -K.; Yoo, I. -K.; Kim, J. -S.; Park, B. H. AdV. Mater. 2007, 19, 3919– 3923. (6) Ishikawa, M.; Tanaka, H.; Kawai, T. Appl. Phys. Lett. 2005, 86, 222504. (7) Takaobushi, J.; Tanaka, H.; Kawai, T.; Ueda, S.; Kim, J.-J.; Kobata, M.; Ikenaga, E.; Yabashi, M.; Kobayashi, K.; Nishino, Y.; Miwa, D.; Tamasaku, K.; Ishikawa, T. Appl. Phys. Lett. 2006, 89, 242507. (8) Takaobushi, J.; Ishikawa, M.; Ueda, S.; Ikenaga, E.; Kim, J. -J.; Kobata, M.; Takeda, Y.; Saitoh, Y.; Yabashi, M.; Nishino, Y.; Miwa, D.; Tamasaku, K.; Ishikawa, T.; Satoh, I.; Tanaka, H.; Kobayashi, K.; Kawai, T. Phys. ReV. B. 2007, 76, 205108. (9) Hochbaum, A. I.; Chen, R.; Delgado, R. D.; Liang, W.; Garnett, E. C.; Najarian, M.; Najarian, M.; Majumdar, A.; Yang, P. Nature (London) 2008, 451, 163–168. (10) Im, Y.; Lee, C.; Vasquez, R. P.; Bangar, M. A.; Myung, N. V.; Menke, E. J.; Penner, R. M.; Yun, M. Small 2006, 3, 356–358. (11) Dai, Z. R.; Pan, Z. W.; Wang, Z. L. AdV. Funct. Mater. 2003, 13, 9–24. (12) Gunawan, O.; Sekaric, L.; Majumdar, A.; Rooks, M.; Appenzeller, J.; Sleight, J. W.; Guha, S.; Haensch, W. Nano Lett. 2008, 8, 1566– 1571. (13) Ra, H. -W.; Choi, K. -S.; Kim, J. -H.; Hahn, Y. -B.; Im, Y. -H. Small 2008, 4, 1105–1109. (14) Suzuki, N.; Tanaka, H.; Yamanaka, S.; Kanai, M.; Lee, B. K.; Lee, H. Y.; Kawai, T. Small 2008, 4, 1661–1665. (15) Yang, Y.; Kung, S. C.; Taggart, D. K.; Xiang, C.; Yang, F.; Brown, M. A.; Gu¨ell, A. G.; Kruse, T. J.; Hemminger, J. C.; Penner, R. M. Nano Lett. 2008, 8, 2447–2451. (16) Huang, Y.; Duan, X.; Cui, Y.; Lieber, C. M. Nano Lett. 2002, 2, 101– 104. (17) Lu, W.; Lieber, C. M. Nat. Mater. 2007, 6, 841–850. (18) Meulen, M. I.; Petkov, N.; Morris, M. A.; Kazakova, O.; Han, X.; Wang, K. L.; Jacob, A. P.; Holmes, J. D. Nano Lett. 2009, 9, 50–56. (19) Choi, H. -J.; Seong, H. -K.; Chang, J.; Lee, K. -I.; Park, Y. -J.; Kim, J. -J.; Lee, S. -K.; He, R.; Kuykendall, T.; Yang, P. AdV. Mater. 2005, 17, 1351–1356. (20) Liang, W.; Yuhas, B. D.; Yang, P. Nano Lett. 2009, 9, 892–896. (21) Seong, H. -K.; Kim, J. -Y.; Kim, J. -J.; Lee, S. -C.; Kim, S. -R.; Kim, U.; Park, T. -E.; Choi, H. -J. Nano Lett. 2007, 7, 3366–3371. (22) Nam, C.; Jang, Y.; Lee, K. -S.; Cho, B. K. Appl. Phys. Lett. 2008, 92, 202501. 1965
(23) Song, Y.; Schmitt, A. L.; Jin, S. Nano Lett. 2008, 8, 2356–2361. (24) Ce´spedes, O.; Watts, S. M.; Coey, J. M. D.; Do¨rr, K.; Ziese, M. Appl. Phys. Lett. 2005, 87, 083102. (25) Rolandi, M; Quate, C. F.; Dai, H. AdV. Mater. 2002, 14, 191–194. (26) Cuesta, I. F.; Borrise´, X.; Murano, F. P. J. Vac. Sci. Technol., B 2006, 24 (6), 2988–2992. (27) Pellegrino, L.; Bellingeri, E.; Siri, A. S.; Marre´, D. Appl. Phys. Lett. 2005, 87, 064102. (28) Johannes, M. S.; Cole, D. G.; Clark, R. L. Appl. Phys. Lett. 2007, 91, 123111.
1966
(29) Goto, K.; Tanaka, H.; Kawai, T. J. Appl. Phys., in press. (30) Ru¨diger, U.; Yu, J.; Kent, A. D.; Parkin, S. S. P. Appl. Phys. Lett. 1998, 73, 1298–1300. (31) Leven, B.; Dumpich, G. Phys. ReV. B 2005, 71, 064411. (32) Aziz, A.; Bending, S. J.; Roberts, H. G.; Crampin, S.; Heard, P. J.; Marrows, C. H. Phys. ReV. Lett. 2006, 97, 206602. (33) Ahn, C. H.; Triscone, J. -M.; Mannhart, J. Nature (London) 2003, 424, 1015–1018.
NL900158T
Nano Lett., Vol. 9, No. 5, 2009
