Epitaxial Electrodeposition of High-Aspect-Ratio Cu2O(110

Epitaxial Electrodeposition of High-Aspect-Ratio Cu2O(110) Nanostructures on InP(111). Run Liu .... The Journal of Physical Chemistry C 0 (proofing), ...
1 downloads 0 Views 493KB Size
Chem. Mater. 2005, 17, 725-729

725

Epitaxial Electrodeposition of High-Aspect-Ratio Cu2O(110) Nanostructures on InP(111) Run Liu,† Elizabeth A. Kulp,† Fumiyasu Oba,‡ Eric W. Bohannan,† Frank Ernst,‡ and Jay A. Switzer*,† Department of Chemistry and Graduate Center for Materials Research, UniVersity of MissourisRolla, Rolla, Missouri 65409-1170, Department of Materials Science and Engineering, Case Western ReserVe UniVersity, 10900 Euclid AVenue, CleVeland, Ohio 44106-7204 ReceiVed September 28, 2004. ReVised Manuscript ReceiVed December 7, 2004

Epitaxial cuprous oxide nanostructures with high aspect ratio were deposited electrochemically on n-InP(111) from aqueous solution at room temperature. High-resolution X-ray diffraction shows that the Cu2O and InP form three equivalent epitaxial orientation relationships that are rotated 120° relative to each other: Cu2O(110)[100] || InP(111)[11h0], Cu2O(110)[100] || InP(111)[1h01], and Cu2O(110)[100] || InP(111)[011h]. The size and aspect ratio of the Cu2O nanostructures depend on the applied deposition current density. At a deposition current density of 0.125 mA/cm2, uniformly sized nanostructures 30 nm wide and 1000 nm long were obtained. Transmission electron microscopy reveals an amorphous, oxygenrich interlayer and a crystalline Cu3P layer between the Cu2O and InP.

Introduction Epitaxial growth of ordered semiconductor nanocrystals on semiconductor substrates is of interest for a variety of potential applications, such as optical electronic devices, quantum computing, and information storage.1 One approach to obtain such structures is to produce self-assembled islands by strain relief during epitaxial growth, as in the vapor deposition of (InGa)As on GaAs2 or Ge on Si.3-5 Electrodeposition has also been shown to be well suited to grow nanostructures.6 A unique feature of electrodeposition is the ability to tune the orientation and morphology of electrodeposited films and to control the size and shape of the deposited nanocrystals by solution conditions (such as pH and additives) and the applied electrode overpotential.7 These degrees of freedom are not available, for example, in vapor deposition. We have previously used electrodeposition in aqueous solution to form epitaxial films of δ-Bi2O3,8 Cu2O,7 ZnO,9 and Fe3O410,11 on gold single crystals, and epitaxial * To whom correspondence should be addressed. E-mail: [email protected]. † University of MissourisRolla. ‡ Case Western Reserve University.

(1) Petroff, P. M.; Lorke, A.; Imamoglu, A. Phys. Today 2001, 54, 46. (2) Leonard, D.; Krishnamurthy, M.; Reaves, C. M.; Denbaars, S. P.; Petroff, P. M. Appl. Phys. Lett. 1993, 63, 3203. (3) Eaglesham, D. J.; Cerullo, M. Phys. ReV. Lett. 1990, 64, 1943. (4) Mo, Y.-W.; Savage, D. E.; Swartzentruber, B. S.; Lagally, M. G. Phys. ReV. Lett. 1990, 65, 1020. (5) Medeiros-Ribeiro, G.; Bratkovski, A. M.; Kamins, T. I.; Ohlberg, D. A. A.; Williams, R. S. Science 1998, 279, 353. (6) Hodes, G. Electrochemistry of Nanomaterials; Wiley: New York, 2001. (7) Switzer, J. A.; Kothari, H. M.; Bohannan, E. W. J. Phys. Chem. B 2002, 106, 4027. (8) Switzer, J. A.; Shumsky, M. G.; Bohannan, E. W. Science 1999, 284, 293. (9) Liu, R.; Vertegel, A. A.; Bohannan, E. W.; Sorenson, T. A.; Switzer, J. A. Chem. Mater. 2001, 13, 508. (10) Nikiforov, M. P.; Vertegel, A. A.; Shumsky, M. G.; Switzer, J. A. AdV. Mater. 2000, 12, 1351.

Cu2O12-14 films with tunable morphology on silicon and InP(100) single crystals. Lincot et al. have electrodeposited epitaxial films of CdTe on InP(111)15 and ZnO on GaN(0002).16 Recently, we have shown that it is possible to electrodeposit epitaxial chiral CuO films onto Au(100) and Cu(111) single crystals.17,18 Here, we show that epitaxial nanostructures of Cu2O(110) can be electrodeposited onto InP(111). The orientation of the film is controlled by, but different from, the substrate orientation, and the size of the nanostructures is tuned by the applied current density. Cuprous oxide (Cu2O) is a p-type semiconductor that has potential application in solar energy conversion and catalysis.19 There is also evidence that Bose-Einstein condensation of excitons can occur when Cu2O is irradiated with highintensity light.20,21 Single-crystalline Cu2O with nanometerscale dimensions would be expected to spatially confine excitons and effectively increase their concentration for a given light intensity. Also, the epitaxial growth of Cu2O on InP is an unexpected result because the lattice parameter of Cu2O is 27.2% smaller than the lattice parameter of InP, (11) Sorenson, T. A.; Morton, S. A.; Waddill, G. D.; Switzer, J. A. J. Am. Chem. Soc. 2002, 124, 7604. (12) Switzer, J. A.; Liu, R.; Bohannan, E. W.; Ernst, F. J. Phys. Chem. B 2002, 106, 12369. (13) Liu, R.; Oba, F.; Bohannan, E. W.; Ernst, F.; Switzer, J. A. Appl. Phys. Lett. 2003, 83, 1944. (14) Liu, R.; Oba, F.; Bohannan, E. W.; Ernst, F.; Switzer, J. A. Chem. Mater. 2003, 15, 4882. (15) Lincot, D.; Kampmann, A.; Mokili, B.; Vedel, J.; Cortes, R.; Froment, M. Appl. Phys. Lett. 1995, 67, 2355. (16) Pauporte, Th.; Lincot, D. Appl. Phys. Lett. 1999, 75, 3817. (17) Switzer, J. A.; Kothari, H. M.; Poizot, P.; Nakanishi, S.; Bohannan, E. W. Nature 2003, 425, 490. (18) Bohannan, E. W.; Kothari, H. M.; Nicic, I. M.; Switzer, J. A. J. Am. Chem. Soc. 2004, 126, 488. (19) De Jongh, P. E.; Vanmaelkelbergh, D.; Kelly, J. J. Chem. Commun. 1999, 12, 1069. (20) Snoke, D. Science 1996, 273, 1351. (21) Johnsen, K.; Kavoulakis, G. M. Phys. ReV. Lett. 2001, 86, 858.

10.1021/cm048296l CCC: $30.25 © 2005 American Chemical Society Published on Web 01/28/2005

726 Chem. Mater., Vol. 17, No. 4, 2005

Liu et al.

Figure 1. X-ray diffraction Bragg-Brentano θ-2θ scan results for Cu2O(110) nanostructures grown from a pH 12 solution on n-InP(111) using an applied current density of 0.025 mA/cm2. Only the (110) and (220) peaks for Cu2O are observed.

Figure 3. (111) pole figures for (A) Cu2O and (B) InP. The pole figures were obtained by setting 2θ equal to the angle of (111) diffraction (2θ ) 36.417° and 26.266° for the Cu2O and InP, respectively) and performing azimuthal scans at tilt angles, χ, from 0 to 90°. The radial grid lines on the pole figure correspond to 30° increments of the tilt angle. The six spots at χ ) 36° in (A) and three spots at χ ) 71° in (B) correspond to the angles between the (111) and (110), and the (111) and (1h11) planes, respectively. Figure 2. X-ray rocking curves for Cu2O(220) (A) and InP(111) (B). The values of full width at half-maximum (fwhm) for Cu2O and InP are 3.1° and 0.0067°, respectively.

and because of the high reactivity of the InP surface in aqueous solution. Experimental Section Electrochemical experiments were carried out using an EG&G Princeton Applied Research (PAR) model 273A potentiostat/ galvanostat. The deposition was performed in a three-electrode cell with a platinum wire as counter electrode and a saturated calomel electrode (SCE) as reference electrode. The deposition solutions were 0.4 M CuSO4 and 3 M lactic acid with the pH adjusted to 12 with concentrated NaOH. The deposition temperature was 25 °C. The applied deposition current densities were 0.025, 0.05, and 0.125 mA/cm2. The thickness of all of the films was estimated to be 2 µm based on the applied charge density of 2 C/cm2 and assuming 100% current efficiency. The n-InP(111) wafers were supplied by Wafer Technology Ltd., doped with sulfur to a resistivity of approximately 1.6 ×10-3 Ωcm, corresponding to a carrier concentration of 3 × 1018 cm-3. The wafers were degreased in ethanol and acetone and then rinsed with HPLC water before etching. The etch was carried out in a solution of 3 M H2SO4 for 3 min to remove the native oxide, followed by a thorough washing with HPLC water. Ohmic contacts were made using Ga-In eutectic reaction. X-ray diffraction measurements were done with a high-resolution Philips X’Pert MRD diffractometer. For the Bragg-Brentano scan, the primary optics module was a combination Go¨bel mirror and a two-crystal Ge(220) two-bounce hybrid monochromator which

Figure 4. Schematic illustration of the epitaxial relationship. The values of the lattice mismatches are 2.9% in the Cu2O[100] direction and 5.0% in the Cu2O[01h1] direction. Indium atoms in InP are green and copper atoms in Cu2O are brown.

produces pure CuKR1 radiation (λ ) 0.1540562 nm) with a divergence of 25 arc sec. A 0.18° parallel plate collimator served as the secondary optics. Pole figures were obtained in point-focus mode using a crossed-slit collimator as the primary optics, and a 0.27° parallel plate collimator and flat graphite monochromator as the secondary optics. SEM micrographs were obtained with a Hitachi model S4700 cold field-emission scanning electron microscope. The high-resolution cross-sectional transmission electron microscopy (TEM) image was acquired with a Tecnai F30 ST field-

Epitaxial Electrodeposition of Cu2O(110) Nanostructures

Figure 5. Scanning electron microscope (SEM) images of the Cu2O films on InP(111) deposited at different current densities: (A) 0.025 mA/cm2; (B) 0.05 mA/cm2; and (C) 0.125 mA/cm2.

emission gun instrument (FEI) operated at 300 kV. Conventional TEM bright-field images and diffraction patterns were obtained with a Philips CM20ST microscope operated at 200 kV.

Results and Discussion Bragg-Brentano X-ray 2θ scans were used to probe the out-of-plane orientation of the films. The XRD pattern for a Cu2O film grown at an applied current density 0.025 mA/ cm2 is shown in Figure 1. The film grows with a [110] orientation on the InP(111) substrate. Only the (110) and (220) peaks of Cu2O are observed. The lattice parameter of 0.4269 nm determined for the Cu2O film is within experimental error of the literature value for that of bulk Cu2O (a ) 0.4270 nm),22 suggesting that the majority of the material is completely relaxed. Figure 2 shows the rocking curves of Cu2O (220) (Figure 2A) and InP(111) (Figure 2B). The full

Chem. Mater., Vol. 17, No. 4, 2005 727

Figure 6. Cross-sectional transmission electron microscopy (TEM) images of the interface between the Cu2O and the InP substrate. (A) Conventional TEM bright-field image, showing the columnar structure of the 2-µm-thick film. (B) The diffraction patterns shown in the insets indicate that the viewing direction corresponds to [100] in the Cu2O and [110] in the InP, and that the [111] normal of the InP substrate corresponds to a [110] direction in the Cu2O. (C) High-resolution TEM image of the Cu2O/InP interface. The image confirms the orientation relationship indicated by the X-ray and electron diffraction patterns. The image also shows amorphous and crystalline interlayers between the Cu2O and InP.

width at half-maximum (fwhm) of Cu2O(220) is 3.1° and the fwhm of InP(111) is 0.0067°. These results indicate that the Cu2O film has a [110] out-of-plane orientation with a 3.1° mosaic spread. X-ray pole figures were used to determine the in-plane orientation of the films. Pole figures are acquired by choosing (22) Golden, T. D.; Shumsky, M. G.; Zhou, Y.; Vanderwerf, R. A.; Van Leeuwen, R. A.; Switzer, J. A. Chem. Mater. 1996, 8, 2499.

728 Chem. Mater., Vol. 17, No. 4, 2005

Liu et al.

Figure 7. (A) High-angle annular dark field image with an X-ray energy dispersive spectroscopy (XEDS) scan line. (B) Normalized typical chemical composition profile along the XEDS line. (C) and (D) XEDS line scans with a finer interval (∼0.85 nm), and two typical normalized chemical composition profiles along the XEDS lines.

a specific plane to probe while measuring the diffracted intensity as a function of tilt (χ) and rotation (φ). The (111) pole figure for Cu2O in Figure 3A shows six peaks at a tilt angle, χ, of 36°. This tilt angle corresponds closely with the interplanar angle (35.5°) between the {111} and {110} planes in a cubic crystal system. Because the Cu2O shows a [110] out-of-plane orientation, one would expect to observe 2-fold symmetry in the X-ray pole figure. The six peaks observed are due to three distinct Cu2O domains rotated 120° relative to each other on the 3-fold surface of InP(111). Figure 3B shows the (111) pole figure for the InP substrate. As expected for a [111] orientation, there is a sharp spot in the center of the pole figure at χ ) 0°. There are also three peaks at a tilt angle, χ, of 71°, which agrees well with the 70° angle between {111} planes in a cubic system. The three equivalent epitaxial relationships that are determined from the pole figures are Cu2O(110)[100] || InP(111)[11h0], Cu2O(110)[100] || InP(111)[1h01], and Cu2O(110)[100] || InP(111)[011h]. Although there is a very large mismatch of 27.2% between the lattice parameters of Cu2O and InP, this mismatch is lowered by depositing a [110] orientation of Cu2O on the InP(111) substrate. A schematic illustration of the epitaxial relationships determined from X-ray diffraction is shown in Figure 4. The values of the lattice mismatches are 2.9% in the Cu2O[100] direction and 5.0% in the Cu2O[01h1] direction. The schematic in Figure 4 is meant only to represent

the orientation of the film relative to the substrate, and is not intended to serve as a model of the interface. The size and aspect ratio of the Cu2O nanostructures is a function of the current density used to deposit the films. Scanning electron microscope images of films deposited at current densities of 0.25, 0.05, and 0.125 mA/cm2 to a charge density of 2 C/cm2 are shown in Figure 5. In all three films, the Cu2O(110) deposits as high-aspect-ratio nanostructures arranged in triangular patterns. This is consistent with the X-ray diffraction results that show that the films have three equivalent in-plane orientations and a [110] out-of-plane orientation. The width of the Cu2O nanostructures decreases as the applied current density increases, consistent with higher nucleation densities at higher overpotentials.23 The Cu2O crystallites are about 250 nm wide by 2000 nm long at 0.025 mA/cm2 (Figure 5A), 60 nm wide by 800 nm long at 0.05 mA/cm2 (Figure 5B), and 30 nm wide by 1000 nm long at 0.125 mA/cm2 (Figure 5C). The Cu2O/InP interface was also studied by cross-sectional transmission electron microscopy (TEM). Figure 6A shows a conventional TEM bright-field image, obtained with a Philips CM20ST operated at 200 kV. The TEM image shows that the films have a columnar microstructure with a (23) Budevski, E.; Staikov, G.; Lorenz, W. J. Electrochemical Phase Formation and Growth; VCH: New York, 1996.

Epitaxial Electrodeposition of Cu2O(110) Nanostructures

thickness of about 2 µm. Selected-area diffraction patterns are shown in the insets of Figure 6B. The top left diffraction pattern was taken from the Cu2O film and the bottom right diffraction pattern was taken from the InP substrate. The viewing direction corresponds to [100] in Cu2O and [11h0] in InP. The [111] normal of the InP corresponds to the [110] direction in Cu2O. The epitaxial relationship of the film and substrate obtained from the electron diffraction patterns is Cu2O(110)[100]||InP(111)[11h0], which agrees with the epitaxial relationship determined by XRD. Figure 6C shows a high-resolution TEM (HRTEM) image of the interface, obtained with a Tecnai F30 ST field-emission gun instrument (FEI) operated at 300 kV. The image confirms the epitaxial relationship indicated by the X-ray and electron diffraction patterns. The image also shows two interlayers: one amorphous and the other a 2-3-nm-thick crystalline layer between Cu2O and InP. On the basis of the lattice spacings and composition analysis (shown in Figure 7), the crystalline layer is identified to be hexagonal Cu3P. The epitaxial relationship between Cu3P and InP is Cu3P(0001)[101h0]||InP(111)[110]. The precipitation of Cu3P was also previously observed in an annealed Cu2O/InP(100) sample,24 and an amorphous layer (determined to be InPO4) was also observed previously in an as-deposited epitaxial film of Cu2O on InP(100).14 Composition profiles of the interlayers were obtained by X-ray energy-dispersive spectrometry (XEDS). The locations of the line scans are shown in the high-angle annular dark field TEM image in Figure 7A. A typical low-resolution line scan is shown in Figure 7B. To determine fine details in the variation of local composition across the interface, we also obtained line scans with a finer raster (∼0.85 nm) of spot measurements. The results of those measurements are shown in Figure 7C and D. Both profiles exhibit a Cu-rich layer at approximately x ) 20, which has a composition of about 75% Cu and 25% P. This is consistent with the above finding by electron diffraction that the crystalline interlayer consists of Cu3P. For the amorphous interlayer, however, the results (24) Oba, F.; Liu, R.; Bohannan, E. W.; Switzer, J. A.; Ernst, F. Unpublished results.

Chem. Mater., Vol. 17, No. 4, 2005 729

are less conclusive. According to Figure 7C, this layer, centered at x ) 12, is rich in O and poor in Cu. This agrees with the results of a recent study on the Cu2O/InP(100) system, where we have identified the amorphous layer to be InPO4.24 In fact, InP is known to form crystalline indium phosphate hydrate (InPO4‚xH2O) after H2SO4 etching.25 Figure 7D, on the other hand, indicates a Cu:O ratio close to 1:1 in this region. A possible reason for this discrepancy is that the interlayer thickness or composition are not very uniform in this specimen. More work needs to be done to determine the identity and origin of this amorphous layer, and to understand how the Cu2O can maintain registry with the substrate despite the amorphous interlayer. One possible explanation is that Cu2O initially grows in direct contact with the InP(111), thus epitaxially, until a solid-state reaction occurs that produces the crystalline Cu3P and amorphous InPO4 interlayers. While at first sight the solid-state reaction at room temperature does not appear to be very likely, it is often observed that interfaces of dissimilar materials are regions of enhanced reactivity. In agreement with the hypothesis of an interface reaction, Spicer et al.26,27 have observed that elemental Cu reacts with InP to produce Cu-containing phophides and metallic In. Acknowledgment. This work was supported by National Science Foundation grants CHE-0243424, CHE-0437346, DMR0071365, and DMR-0076338, and Department of Energy grant DE-FC07-03ID14509. The Case School of Engineering and the Case Alumni Society are gratefully acknowledged for financial support of this work. F. Oba is supported by a Grant-in-Aid for JSPS Fellows from the Ministry of Education, Culture, Sports, Science and Technology of Japan. CM048296L (25) Liu, H. C.; Tsai, S. H.; Hsu, J. W.; Shih, H. C. J. Electrochem. Soc. 1999, 146, 3510. (26) Kendelewicz, T.; Petro, W. G.; Lindau, I.; Spicer, W. E. J. Vac. Sci. Technol. B 1984, 2, 453. (27) Cao, R.; Miyano, K.; Kendelewicz, T.; Lindau, I.; Spicer, W. E. Appl. Phys. Lett. 1988, 53, 210.