Open Access

Growth of 3-D flower/grass-like metal oxide nanoarchitectures based on catalyst-assisted oxidation method

Nanoscale Research Letters20149:116

DOI: 10.1186/1556-276X-9-116

Received: 6 January 2014

Accepted: 24 February 2014

Published: 13 March 2014


Cu2O grass-like and ZnO flower-like nanoarchitectures were fabricated directly on Cu powders and Zn powders using a novel thermal oxidation stress-induced (TOS) method based on catalyst assistance at a low temperature of 150°C under moderate humid atmosphere. The experiments of Al powder were also carried out based on TOS method. Overlapping migration (OLM) of Cu and Zn atoms and toothpaste squeezing migration (TSM) of Al atoms caused by different atom densities in metal oxide materials were studied.


81. Materials science; 81.07.-b Nanoscale materials and structures: fabrication and characterization; 81.16.Hc Catalytic methods


Flower/grass-like Nanoarchitecture Thermal oxidation Stress gradient Nickel catalyst


Cuprous oxide (Cu2O) is a p-type semiconductor metal oxide with a direct band gap of approximately 2.17 eV [1], which has been used as a prospective candidate for low-cost solar energy conversion [2], photocatalysis [3], and sensors [4]. On the other hand, zinc oxide (ZnO) is an n-type semiconductor with a direct band gap of 3.37 eV [5]. Due to its unique optical, electrical, and magnetic properties, it has attracted a great attention and has been widely applied in solar cells and sensors [68]. Recently, great efforts have been devoted to fabricate different Cu2O and ZnO nanoarchitectures [912] because architectures, including geometry, morphology, and hierarchical structures, were found to have a crucial effect on the fundamental properties of micro/nanostructure semiconductors [1316]. However, until now, all the fabrication methods of Cu2O flower-like nanoarchitectures belong to chemical solution methods, which are complex and expensive [1719]. Moreover, ZnO flower-like nanoarchitectures were fabricated by chemical solution routes and hydrothermal method, which either need highly accurate quantitative chemical materials or need complex experimental steps [20, 21].

Recently, we have proposed a novel method using thermal oxidation with participation of catalyst and humidity to fabricate the three-dimensional Cu2O flower/grass-like nanoarchitectures (FGLNAs), and the morphology of the Cu2O FGLNAs can be controlled by the heating temperature [22]. Although the growth mechanism has been proposed, experimental investigation is still necessary in order to further verify it. Therefore, in this study, Cu powders were used to replace the Cu foil and Cu film for the fabrication of Cu2O FGLNAs based on thermal oxidation stress-induced (TOS) method. In addition, to verify the generality of the TOS method, Zn powders were also used to fabricate ZnO FGLNAs. Moreover, to investigate the growth mechanism affected by the atom density of metal oxide materials, Al powders were applied to the TOS method. Based on the experimental results, an overlapping migration (OLM) of Cu and Zn atoms and toothpaste squeezing migration (TSM) of Al atoms due to different atom densities of metal oxide materials were proposed in this study. Compared with the fabrication based on the thermal expansion stress-induced method [23, 24], the heating temperature of the approach used here is only 150°C which is at least 200°C lower than that of CuO and 250°C lower than that of ZnO nanowire growth [25]. The experimental fact that metal powders can replace metal foil and film for the growth of FGLNAs further proved that the growth mechanism is based on oxidation extension rather than thermal expansion induced stress.


Commercial silicon wafer with a thickness of 0.50 mm was cut into square samples which were 10 × 10 mm2 in size. The substrates were then ultrasonically cleaned (Bransonic 1510, Branson Ultrasonics Corp., Danbury, CN, USA) with acetone and washed with ethanol and de-ionized water sequentially to dissolve the contaminations. Afterwards, Ni catalyst was manually daubed on the surface of samples with a cuboid shape having dimensions of 10 × 3 × 2 mm3 approximately. The nickel catalyst in this experiment was used as a high temperature resistance electrically conductive coating material (service temperature 538°C Pyro-Duct™ 598-C, Aremco, Inc., Valley Cottage, NY, USA). Cu, Zn, and Al powders were dispersed around Ni on the substrate, respectively. Cu, Zn, and Al powder samples were then heated by a ceramic heater in air atmosphere under 55% to 75% humidity at the temperature of 150°C for 2, 7, and 10 days, respectively.

After the heating process, morphologies of FGLNAs were characterized by scanning electron microscopy (SEM; JSM-7000FK, JEOL Ltd., Akishima, Tokyo, Japan), energy-dispersive X-ray (EDX), and X-ray diffraction (XRD).

Results and discussion

Figure 1 shows SEM images of (a) Cu2O grass-like nanoarchitectures, (b) ZnO flower-like nanoarchitectures, and (c) Al nanowires grown on Cu, Zn, and Al powder samples at 150°C under 55% to 75% humidity for 2, 7, and 10 days, respectively. The size of Cu2O grass-like nanoarchitectures is 10 to 15 μm, and the width of their petals is 350 to 900 nm. The size of ZnO flower-like nanoarchitectures is 9 to 17 μm, and the width of their petals is 450 to 950 nm. The length of Al nanowires is approximately 120 μm, and the mean diameter is 10 μm approximately. It has been confirmed experimentally that there was no FGLNA growth when the experimental conditions were changed to vacuum environment, without catalyst or under the humidity lower than 55% or higher than 75%, respectively. Therefore, it is thought that besides temperature, oxygen atmosphere, catalyst, and humidity were three essential conditions for the growth of FGLNAs.
Figure 1

SEM images. (a) Cu2O grass-like nanoarchitectures, (b) ZnO flower-like nanoarchitectures, and (c) Al nanowires grown on Cu, Zn, and Al powders at 150°C under 55% to 75% humidity for 2, 7, and 10 days, respectively.

Figure 2 shows the EDX results of Cu2O grass-like nanoarchitectures. It indicates that the grass-like nanoarchitectures are mainly composed of Cu element (30.95%) and oxygen element (68.30%). We also obtained similar EDX results for the other samples. As shown in the XRD spectrum in Figure 3, orientations 111, 200, 311, etc. of Cu2O indicate that the FGLNAs are composed of Cu2O. As shown in the XRD spectrum, Ni was not oxidized. The reason is that oxidation temperature of Ni is above 400°C. Figure 4 shows EDX analysis of the flower-like nanoarchitectures grown on the Zn powder sample heated at 150°C for 10 days. It indicates that the architectures are mainly composed of Zn element (58.26%) and oxygen element (41.74%). As shown in the XRD spectrum in Figure 5, orientations 002, 101, 201, etc. of ZnO indicate that the FGLNAs are composed of ZnO. As shown in the XRD spectrum, Ni was not oxidized. We also obtained the EDX results of Al nanowires grown on the Al powder sample, as shown in Figure 6. The EDX result indicates that the nanowires grown on the Al powder sample are Al nanowires instead of Al2O3 nanowires.
Figure 2

EDX spectra of architectures on Cu powders heated at 150°C under 55% to 75% humidity for 2 days.
Figure 3

XRD spectra of architectures on Cu powders heated at 150°C under 55% to 75% humidity for 2 days.
Figure 4

EDX spectra of architectures on Zn powders heated at 150°C under 55% to 75% humidity for 7 days.
Figure 5

XRD spectra of architectures on Zn powders heated at 150°C under 55% to 75% humidity for 7 days.
Figure 6

EDX spectra of architectures on Al powders heated at 150°C under 55% to 75% humidity for 10 days.

The growth mechanism is similar with that described in our previous study of Cu foil/film samples [22]. An oxide layer formed on the surface of the metal powders when the samples were heated in air. Taking the Cu powder sample as an example, as shown in Figure 7, stress occurred in the oxide shell due to the oxide volume extension. Stress in the oxide shell can be determined by the Pilling-Bedworth ratio (PBR), which is defined as [26]
PBR = Volume of oxide Volume of metal
Figure 7

Illustration of grow mechanism.

Here, the volume of metal expresses the consumed metal volume during the formation of oxide shell in the oxidation process. When PBR >1, a compressive stress develops in the oxide shell, while a tensile stress develops in the shell with PBR <1. The larger the difference of PBR from 1, the larger the stress will be. In this experiment, due to volume extension of the Cu2O and ZnO oxide shell, the PBR of Cu2O and ZnO is larger than 1. Therefore, the Cu2O shell suffers tangential compressive stress (TCS) from the core, as illustrated in Figure 7. The same situation happens in the ZnO shell.

Meanwhile, tangential tensile stress (TTS), caused by the reactive force of TCS in the oxide shell, occurred in the inner part of Cu powders at the interface of Cu2O (shell)/Cu (core), which leads to the generation of radial stress gradient (RSG) in the thickness direction of the powders. RSG serves as a driving force for the migration of Cu atoms from the center of Cu powders to the interface between the oxide shell and Cu core. At the beginning, the relatively low temperature (150°C) causes relatively lower surface oxidation speed. The Cu2O shell which formed on the Cu powder sample is very thin, and the RSG is not large enough. Therefore, the diffused Cu atoms cannot penetrate the oxidation shell. However, sufficient bivalent oxygen ions with two chemical bonds (BOICBs) were generated from the water vapors during the process of hydrogen absorption of nickel catalyst, as indicated in Equation 2.
H O H Ni Δ O + H 2

Thus, the low-temperature oxidation was enhanced, and the thickness of the Cu2O shell became larger and larger. Therefore, the TCS in the Cu2O shell caused by oxide volume extension will be larger than the results without participation of catalyst and humidity, thereby creating larger RSG. On the other hand, the TCS in the oxide shell also made it difficult for Cu atoms to penetrate through the oxide layer from the weak spots on the surface. Consequently, Cu atoms kept accumulating under the oxide shell until there were enough Cu atoms to break the balance, and finally, a large number of Cu atoms suddenly penetrated the oxide shell through the weak spots in a flash. Fewer weak spots appear due to the relatively thicker oxide shell, and a relatively large penetration force is required. A large number of Cu atoms accumulate and penetrate the Cu2O shell through the same weak spots. Cu atoms burst out and get more easily oxidized. Moreover, during the formation of FGLNAs, the BOICBs served as a nuclear site. In conclusion, more migrations of Cu atoms caused by larger RSG and larger TCS caused by sufficient BOICBs are two key factors for the growth of FGLNAs.

The mechanism of RSG created in metal powder samples here is different from that in Cu film on Si substrate [10, 23, 24] in which the gradient stress was generated due to the thermal expansion mismatch of the materials. In this study, RSG was generated due to the TCS and reaction TTS caused by oxide volume extension. However, the experiments in this study proved that oxide volume extension degree has a decisive effect on the growth type of nanostructures. That is the reason that Cu2O and ZnO FGLNA growth under a relatively low temperature was realized, instead of that CuO and ZnO nanowires grow under a relatively high temperature.

To further study the growth mechanism of nanoarchitectures, we focus on the different migration mechanisms of Cu/Zn atoms and Al atoms. It is believed that the migration of Cu/Zn atoms is OLM and the migration of Al atoms is TSM [24]. For the OLM of Cu/Zn atoms, during the growth, the old atoms were covered by the new atoms, which seem to be overlapping. As shown in Figure 8a, using Cu as an example, before penetration through the oxide shell, Cu atoms were ordered by the colors yellow, blue, and green from top to bottom. After the penetration, the sequence of Cu atoms changes reversely. From top to bottom, the Cu atoms are green, blue, and yellow, respectively. On the contrary, the migration of the Al atom is TSM. As shown in Figure 8b, before penetration through the oxide shell, Al atoms were ordered by the colors yellow, blue, and green from top to bottom. After the penetration, the sequence of Al atoms stays the same. In other words, during the penetration, the Al atoms keep the same sequence and the first Al atom that penetrated the oxide shell will migrate to the top of the Al nanowire, which seems like toothpaste squeezing.
Figure 8

Illustration of metal atom migration mechanism. (a) Overlapping migration (OLM) of Cu atoms. (b) Toothpaste squeezing migration (TSM) of Al atoms.

The reason is believed to include two aspects. The first is the different atom densities of Cu2O, ZnO, and Al2O3.
ρ C u 2 O = m V = 6.00 g / c m 3
ρ C u 2 O molecular = m V M = 0.0416 mol / c m 3
ρ C u 2 O atoms = n m V M = 3 ρ C u 2 O molecular = 0.125 mol / c m 3

Here, m is the mass of Cu2O, M is the molar mass of Cu2O, V is the volume of Cu2O, n is the number of atoms in a single molecule (for Cu2O, it is 3), ρ C u 2 O is the density of Cu2O, ρ C u 2 O molecular is the Cu2O molecular density, and ρ C u 2 O atoms is the Cu2O atom density. As we know, the density of the Cu2O material is 6.00 g/cm3, as shown in Equation 3. Cu2O molecular density can be calculated to be 0.0416 mol/cm3, as shown in Equation 4. Cu2O atom density was calculated to be 0.125 mol/cm3, as shown in Equation 5. Using the same method, the ZnO and Al2O3 atom densities are calculated to be 0.138 and 0.194 mol/cm3, respectively.

ZnO and Cu2O atoms have much lower atom density than Al2O3 atoms. Taking Cu2O as an example, the lower atom density of the Cu2O oxide surface layer on FGLNA leaves makes Cu atoms easily penetrate the oxide surface layer and get oxidized. Afterwards, a new oxide layer forms on the top surface layer of FGLNA leaves. As shown in Figure 8a, the yellow Cu atoms are the first Cu atoms to penetrate the oxide surface layer, and after oxidation, the Cu2O atoms generated by the yellow Cu atoms would be laying on the bottom layer of Cu2O FGLNAs. Due to the sparse Cu2O FGLNA oxide surface layer, new blue Cu atoms penetrated the surface layer of FGLNA leaves and get oxidized. As shown in Figure 8a, the yellow dotted line arrow indicates the direction of blue Cu atoms migrating and penetrating the oxide surface layer of FGLNA leaves formed by yellow Cu atoms. Green Cu atoms keep this penetration and oxidation cycle. At last, the layer generated by green Cu atoms lie above the one generated by the blue Cu atoms. Due to this cycle, the leaves of Cu2O FGLNAs grow bigger and bigger. For the Al powder case, when the Al powder sample was heated in the air, dense thin oxide layers formed on the surface of Al powders, which prevent atoms from getting further oxidation. As has been calculated above, the Al2O3 atom density is much higher than Cu2O and ZnO atom densities, and this dense oxide shell on the surface of the Al powder makes it difficult for the Al atoms to penetrate through it. Thus, Al atoms keep the same sequence and migrate in a straight line during migration. Therefore, Al nanowires were generated on the surface of the oxide shell. Afterwards, due to the high density of the surface oxide layer on Al nanowires, Al atoms migrate straight and cannot penetrate the surface layer of Al nanowires. Therefore, the present TOS method is unavailable to generate Al2O3 FGLNAs due to the unique oxidation properties and higher atom density of Al2O3.

Secondly, according to the previous study, the PBR of Al2O3 is 1.28 which is much smaller than those of Cu2O and ZnO [27]. Higher PBR of Cu2O and ZnO means bigger oxide volume extension during oxidation. Therefore, higher TCS and TTS were generated, which results in higher RSG. Higher driving force of RSG leads to more Cu and Zn atoms migrating from the metal core to the interface of the oxide shell. More Cu and Zn atoms accumulate and erupt from the weak spots on the surface of metal powder to form FGLNAs.

On the other hand, the heating time for the first appearance of Cu2O, ZnO FGLNAs, and Al nanowires was also observed for the samples of Cu, Zn, and Al powders. As shown in Figure 9, the heating time for the samples of Cu, Zn, and Al powders is 2, 7, and 10 days, respectively. Ranking of PBR from big to small is Cu2O, ZnO, and Al2O3, respectively. Higher PBR leads to higher RSG. Higher RSG promotes the diffusion of Cu atoms, thereby speeding up the growth of FGLNAs. In addition, it is believed that during oxidation, the BOICBs serve as a bridge to connect metal atoms. In the unit cell of Cu2O, ZnO, and Al2O3, BOICBs connect with two Cu atoms, four Zn atoms, and three Al atoms, respectively. Due to the two chemical bonds of BOICBs, it is believed that the combination with two Cu atoms is the easiest. Next is that with four Zn atoms, and the most difficult is that with three Al atoms. Thus, the time required for the first appearance of Cu2O, ZnO, and Al nanoarchitectures increases orderly. Moreover, with the same length and width of Ni cuboid, the thickness of the Ni catalyst can also affect the growth time of Cu2O FGLNAs, but not their morphology and size. Thinner thickness of the Ni catalyst would lead to longer time for the growth of FGLNAs.
Figure 9

Time required for the first appearance of nanoarchitectures.


Cu2O and ZnO FGLNAs were successfully fabricated using the TOS method based on catalyst assistance under moderate humid atmosphere. Higher atom density and large oxide volume extension of Cu2O and ZnO lead to OLM of Cu and Zn atoms, respectively. On the contrary, smaller atom density and oxide volume extension of Al2O3 result in TSM of Al atoms. Higher PBR promotes the diffusion of metal atoms, thereby speeding up the growth of FGLNAs. Compared with other methods to fabricate FGLNAs, the TOS method featured remarkable simplicity and cheapness.



bivalent oxygen ions with two chemical bonds


flower/grass-like nanoarchitectures


overlapping migration


radial stress gradient


tangential compressive stress


thermal oxidation stress-induced


toothpaste squeezing migration


tangential tensile stress.



This work was supported by the Japan Society for the Promotion of Science under a Grant-in-Aid for Scientific Research (A) 23246024.

Authors’ Affiliations

Department of Mechanical Science and Engineering, Nagoya University, Furo-cho


  1. Xiong YJ, Li ZQ, Zhang R, Xie Y, Yang J, Wu CZ: From complex chains to 1D metal oxides: a novel strategy to Cu2O nanowires. J Phys Chem B 2003, 107: 3697–3702. 10.1021/jp027072sView ArticleGoogle Scholar
  2. Briskman RN: A study of electrodeposited cuprous oxide photovoltaic cells. Sol Energy Mater Sol Cells 1992, 27: 361. 10.1016/0927-0248(92)90097-9View ArticleGoogle Scholar
  3. Ho JY, Huang MH: Synthesis of submicrometer-sized Cu2O crystals with morphological evolution from cubic to hexapod structures and their comparative photocatalytic activity. J Phys Chem C 2009, 113: 14159. 10.1021/jp903928pView ArticleGoogle Scholar
  4. Poizot P, Laruelle S, Grugeon S, Dupont L, Tarascon JM: Nano-sized transition-metal oxides as negative-electrode materials for lithium-ion batteries. Nature 2000, 407: 496. 10.1038/35035045View ArticleGoogle Scholar
  5. Wei HM, Gong HB, Wang YZ, Hu XL, Chen L, Xu HY, Liu P, Cao BQ: Three kinds of Cu2O/ZnO heterostructure solar cells fabricated with electrochemical deposition and their structure-related photovoltaic properties. Crystengcomm 2011, 13: 6065–6070. 10.1039/c1ce05540bView ArticleGoogle Scholar
  6. Izaki M, Shinagawa T, Mizuno KT, Ida Y, Inaba M, Tasaka A: Electrochemically constructed p-Cu2O/n-ZnO heterojunction diode for photovoltaic device. J Phys D Appl Phys 2007, 40: 3326–3329. 10.1088/0022-3727/40/11/010View ArticleGoogle Scholar
  7. Fay S, Steinhauser J, Nicolay S, Ballif C: Polycrystalline ZnO: B grown by LPCVD as TCO for thin film silicon solar cells. Thin Solid Films 2010, 518: 2961–2966. 10.1016/j.tsf.2009.09.189View ArticleGoogle Scholar
  8. Ozturk S, Kilinc N, Ozturk ZZ: Fabrication of ZnO nanorods for NO2 sensor applications: effect of dimensions and electrode position. J Alloy Compd 2013, 581: 196–201.View ArticleGoogle Scholar
  9. Yu Y, Du FP, Ma LL, Li JL, Zhuang YY, Qi XH: Synthesis of Cu2O nanowhiskers with CTAB as a template in water-in-oil microemulsion system. RARE METALS 2005, 24: 283.Google Scholar
  10. Luo YS, Li SQ, Ren QF, Liu JP, Xing LL, Wang Y, Yu Y, Jia ZJ, Li JL: Facile synthesis of flowerlike Cu2O nanoarchitectures by a solution phase route. Cryst Growth Des 2007, 7: 87–92. 10.1021/cg060491kView ArticleGoogle Scholar
  11. Wu YJ, Liu YS, Hsieh CY, Lee PM, Wei YS, Chang YH, Lai KY, Liu CY: Light extraction enhancement of vertical LED by growing ZnO nano-rods on tips of pyramids. IEEE Photonic Tech L 2013, 25: 1774–1777.View ArticleGoogle Scholar
  12. Fang B, Zhang CH, Zhang W, Wang GF: A novel hydrazine electrochemical sensor based on a carbon nanotube-wired ZnO nanoflower-modified electrode. Electrochim Acta 2009, 55: 178–182. 10.1016/j.electacta.2009.08.036View ArticleGoogle Scholar
  13. Prabakar K, Son M, Kim WY, Kim H: TiO2 thin film encapsulated ZnO nanorod and nanoflower dye sensitized solar cells. Mater Chem Phys 2011, 125: 12–14. 10.1016/j.matchemphys.2010.09.028View ArticleGoogle Scholar
  14. Yue YM, Chen MJ, Ju Y, Zhang L: Stress-induced growth of well-aligned Cu2O nanowire arrays and their photovoltaic effect. Scripta Mater 2012, 66: 81–84. 10.1016/j.scriptamat.2011.09.041View ArticleGoogle Scholar
  15. Tan YW, Xue XY, Peng Q, Zhao H, Wang TH, Li YD: Controllable fabrication and electrical performance of single crystalline Cu2O nanowires with high aspect ratios. Nano Lett 2007, 7: 3723. 10.1021/nl0721259View ArticleGoogle Scholar
  16. Zhao X, Saka M: Growth of copper microflowers assisted by electromigration under a low current density and subsequent stress-induced migration. Mater Lett 2012, 79: 270–272.View ArticleGoogle Scholar
  17. Gu YE, Su X, Du YL, Wang CM: Preparation of flower-like Cu2O nanoparticles by pulse electrodeposition and their electrocatalytic application. Appl Surf Sci 2010, 256: 5862–5866. 10.1016/j.apsusc.2010.03.065View ArticleGoogle Scholar
  18. Ma LL, Li JL, Sun HZ, Qiu MQ, Wang JB, Chen JY, Yu Y: Self-assembled Cu2O flowerlike architecture: polyol synthesis, photocatalytic activity and stability under simulated solar light. Mater Res Bull 2010, 45: 961–968. 10.1016/j.materresbull.2010.04.009View ArticleGoogle Scholar
  19. Liang ZH, Zhu YJ: Synthesis of uniformly sized Cu2O crystals with star-like and flower-like morphologies. Mater Lett 2005, 59: 2423–2425. 10.1016/j.matlet.2005.02.086View ArticleGoogle Scholar
  20. Kim YJ, Yoo J, Kwon BH, Hong YJ, Lee CH, Yi GC: Position-controlled ZnO nanoflower arrays grown on glass substrates for electron emitter application. Nanotechnology 2008, 19: 315202. 10.1088/0957-4484/19/31/315202View ArticleGoogle Scholar
  21. Jiang CY, Sun XW, Lo GQ, Kwong DL, Wang JX: Improved dye-sensitized solar cells with a ZnO-nanoflower photoanode. Appl Phys Lett 2007, 90: 263501. 10.1063/1.2751588View ArticleGoogle Scholar
  22. Hu LJ, Ju Y, Hosoi A, Tang YP: The surface condition effect of Cu2O flower/grass-like nanoarchitectures grown on Cu foil and Cu film. Nanoscale Res Lett 2013, 8: 445. 10.1186/1556-276X-8-445View ArticleGoogle Scholar
  23. Saka M, Yamaya F, Tohmyoh H: Rapid and mass growth of stress-induced nanowhiskers on the surfaces of evaporated polycrystalline Cu films. Scripta Mater 2007, 56: 1031–1034. 10.1016/j.scriptamat.2007.02.036View ArticleGoogle Scholar
  24. Chen MJ, Yue YM, Ju Y: Growth of metal and metal oxide nanowires driven by the stress-induced migration. J Appl Phys 2012, 111: 104305. 10.1063/1.4718436View ArticleGoogle Scholar
  25. Chao LC, Tsai SY, Lin CN, Liau CC, Ye CC: Vertically aligned ZnO nanowires prepared by thermal oxidation of RF magnetron sputtered metallic zinc films. Mat Sci Semicon Proc 2013, 16: 1316–1320. 10.1016/j.mssp.2013.05.011View ArticleGoogle Scholar
  26. Pilling NB, Bedworth RE: The oxidation of metals at high temperatures. J Clim Appl Meteorol 1923, 29: 529–582.Google Scholar
  27. Yuan L, Wang YQ, Mema R, Zhou GW: Driving force and growth mechanism for spontaneous oxide nanowire formation during the thermal oxidation of metals. Acta Mater 2011, 59: 2491–2500. 10.1016/j.actamat.2010.12.052View ArticleGoogle Scholar


© Hu et al.; licensee Springer. 2014

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited.