Open Access

Tuning the dimensionality of ZnO nanowires through thermal treatment: An investigation of growth mechanism

Nanoscale Research Letters20127:354

DOI: 10.1186/1556-276X-7-354

Received: 3 May 2012

Accepted: 30 May 2012

Published: 28 June 2012


In this study, we synthesized various dimensionalities of ZnO nanowires using the Ti grid-assisted chemical vapor deposition process. Energy dispersive X-ray spectroscopic mapping technique accompanied with a lattice diffusion model was used to characterize the growth mechanism. A diffusion ratio γ, defined by short-circuit and lattice diffusion activation energies, was obtained to describe the growth mechanism of ZnO nanowires. The tunable dimensionalities of ZnO nanowires allow us to modify the morphology of ZnO nanocrystals by developing well-controlled potential applications.


nanocrystalline materials short-circuit diffusion lattice diffusion nanowires ZnO 61.46.Hk 61.82.Fk 62.23.Hj 66.30.Pa


Dense arrays of oriented, crystalline ZnO nanowires have attracted much attention for applications in nanoscale lasers[1], light-emitting diodes[2, 3], sensors[4], and solar cells[5]. Dimensionality and size are the two key factors that govern the properties of nanostructures affected by their high surface-to-volume ratio[6, 7]. The requirements for dimensional control, especially of size and morphology effects on nanoparticles, nanowires, and three-dimensional (3-D) nanowires, seem to still be a challenge. Some 3-D nanostructural materials have been synthesized. These works have primarily focused on the synthesis of inorganic[8], polymeric nanomaterials[9, 10], and dendrite nanowires[11, 12]. To date, ZnO has displayed a series of nanostructures with different morphologies or dimensionalities[13, 14]. It is believed that the discrepancy of the formation under different growing methods is vitally responsible for shape modifications of ZnO. Understanding the relation between growth mechanism and dimensionality is attracting more attention than ever before and is becoming urgently important for obtaining nanowires with a desired size, shape, and dimensionality. However, exact control for growing multidimensional-oriented arrays of ZnO still remains out of reach. The development of a simple, easily controllable method for growing one-dimensional (1-D) to 3-D ZnO nanowire arrays is of great significance[15, 16]. In this study, we have successfully synthesized well-separated one- to three-dimensional ZnO nanowire networks using a Ti grid-assisted thermal evaporation approach. Energy dispersive X-ray spectroscopic (EDS) mapping was used to investigate the scale structure of ZnO nanowire. The formation of various dimensional ZnO nanowires is attributed to short-circuit and lattice diffusion mechanisms.


A series of three-dimensional ZnO nanowire network was fabricated[17]. The samples were prepared by a process where a pure zinc ingot was mounted in a Ti hollow grid which was then placed in a ceramic boat inside a quartz tube. The tube was evacuated to about 10−3 Torr using a mechanical pump and heated in a tube furnace at about 200°C for 2 h to form a Zn film. The Ti grid samples with the resultant Zn film were then heated to various annealing temperatures TA ranging from 300°C to 700°C for 2 h in a mixed argon (100 sccm) and oxygen (10 sccm) atmosphere. The dimensionality and morphology of various ZnO nanocrystals are shown in Figure1a,b,c,d,e,f,g,h. Details of the shape and morphology of the prepared nanocrystals were characterized using field-emission scanning electron microscopy (FESEM; JEOL JSM-6500 F, JEOL Ltd., Akishima, Tokyo, Japan). Figure1a,b,c,d,e,f,g,h shows typical images of the SEM morphology on the as-synthesized products at various TA. It can be seen that the ZnO grew homogeneously in a large area of the Ti grid substrate to form a ZnO film at TA = 300°C, well-separated straight nanowires at TA = 400°C, three-dimensional ZnO nanowires between TA = 500°C to 650°C, and ZnO nanoflowers at TA = 700°C.
Figure 1

Annealing temperature dependence of SEM images and <d>. (a) Low- and (b) high-magnification SEM images of the ZnO film on Ti grid at TA = 300°C. SEM images of ZnO nanocrystals after various annealing temperatures with (c) 1-D ZnO nanowires at TA = 400°C; (d) to (g) 3-D ZnO nanowires at TA = 500°C, 550°C, 600°C, and 650°C, respectively; (h) ZnO nanoflowers at TA = 700°C; (i) the growth temperature TA dependence of the mean diameter <d> obtained from a portion of an SEM image of ZnO nanowires, where the solid curve is a guide for the eye.

As shown in Figure1c, the as-grown 1-D ZnO nanowires grew homogeneously on the Ti grid substrate to form straight nanowires. Observation of the uniform nanowires (with lateral dimensions, on the order of nanometers, i.e., in the hundred to ten nanoscale) shows that they grew up to a few microns in length. By controlling the growth temperature TA = 500°C to 650°C, the 3-D hybrid ZnO nanostructure is seen in Figure1d,e,f,g. Also, the average diameter estimated from the main arms increases with TA increase. As a result, shown in Figure1h, the separated 3-D ZnO nanowire network finally grow into 2-D nanosheets through the high-temperature process, aggregating and leading to form nanosheet-assembled ZnO flowers. The mean ZnO nanowire diameters <d>, as determined from the SEM images (Figure1c,d,e,f,g,h) and described by the fit of the log-normal functions,a were approximately 32(3), 37(4), 38(4), 58(6), 131(13) and 31(3) nm, respectively. The diameter of a ZnONW ranges from 32 to 131 nm. The length of the ZnONWs was found to be dependent on the deposition time and annealing temperature. It is worth noting that the present size of ZnO nanoflowers is defined from the distribution of the mean width of each nanosheet.

The TA dependence of the mean diameters obtained from a portion of the SEM image is shown in Figure1i, where the solid curve shows the fit to the growth law. This can be well described as <d> = 169(19) − 0.9(8)TA + 0.0013(1)TA2. In this present study, the growth temperature of ZnO nanowires was confined to between 400°C to 650°C, which is 0.21 and 0.33 times the melting point of ZnO (1,975°C), following the Wagner's scaling theory[18]. This result is different with the previous report by Jeong group that claimed higher annealing temperature results in lower dimensionality in ZnO nanowire using MOCVD method[19]. In case of the MOCVD, the growth of ZnO nanowires could be explained by the island formation on compressively strained sites and surface diffusion of source materials at supersaturation level. In the present case of the CVD, the diffusion of Zn on the surface, it is necessary to consider the thermal energy and initial growth into wire-like structure that are kinetically favored. Thereby, detailed studies of size and morphology of ZnO nanowires may greatly contribute to the understanding of growth mechanism.

Results and discussion

Analysis of crystal structure by TEM

Transmission electron microscopic (TEM), selected area electron diffraction (SAED) pattern, and high-resolution transmission electron microscopic (HRTEM) images from a JEM-3010 transmission electron microscope (JEOL Ltd., Japan) were obtained to study the crystalline structure. As an example, we show the case of 3-D ZnO nanowires with crystalline structures. Figure2a shows the TEM morphology of a typical 3-D ZnO nanowire. This TEM image reveals that most of the nanowires are straight, and the diameter along the growth direction is uniform, with a mean diameter of 43(1) nm for the main arms (marked as S1) and 55(2) nm for the branches (marked as S2). The single crystalline nature of the sample studied is clearly revealed. The Bragg spots correspond to the zone axis [0 0 1] reflection of the wurtzite structure of the ZnO. Shown in Figure2b is the SAED result; the pattern of the main spots can easily be seen appearing as hexagonal cells with lattice parameters of a = 3.250(5) Å and c = 5.205(3) Å, which indicate a predominantly crystalline hexagonal wurtzite ZnO. The detailed structure of the controlled-growth 3-D ZnO nanowires was further investigated using HRTEM. Two high-magnification enlargements of a selected region of the HRTEM images in Figure2a at branched regions are shown. Figure2c,d shows an enlargement along two selected perpendicular section lines with the corresponding HRTEM images shown in the insets. Gray-level analysis of the image is employed to extract height information. The resulting height-position information is fitted using Gaussian functions to obtain the average atomic spacing. The fitted inter-planar distances of the two fingers are 0.33(1) (marked S1) and 0.33(3) nm (marked S2) corresponding to the [1 0 0] and [0 1 0] planes of the ZnO nanowire, respectively. The results are plotted schematically in Figure2e. It can be seen that for the growth direction of these two branches is indicated by [1 0 0] and [1 1 0], and they are inclined towards each other by 60° at the boundary regions, resembling in form the dendritic 3-D ZnO nanowires. The growth mechanism of ZnO nanocrystals was reported in the previous work by Fan group[12]; they explained and proposed that the growth of the main arms and branches is through a self-catalytic liquid–solid and vapor-solid processes, respectively. However, in the multidimensional ZnO nanowire system, the growth process is dependent on the surface diffusion of Zn, with respect to the annealing temperature and time. To verify the formation of ZnO nanowires, further SEM investigation was carried out.
Figure 2

TEM image, electron diffraction pattern, height-position information, and schematic plot. (a) High-magnification bright-field TEM image revealing the dendritic ZnO structure. Many short branches extending from the side of the main arm can be observed. (b) Electron diffraction pattern of the selected area of the sample, revealing the [0 0 1] zone axis of the ZnO nanowires. (c) to (d) The resulting height-position information for two selected perpendicular section lines taken from HRTEM images of a high-magnification enlargement of the selected region marked in Figure2a. (e) Schematic plot of a single dendritic ZnO nanowire structure.

EDS mapping

An EDS (Oxford Instruments Inca x-sight model 7557, Abingdon, Oxfordshire, UK) mapping technique was used to measure the surface thickness of the ZnO film on the Ti grid. EDS mapping generates a two-dimensional image indicating the abundance of an element. The intensity of the image allows direct visualization of the spatial distribution of any element, such as zinc, titanium, or oxygen, on the surface of the Ti grid. The typical EDS elemental spectrum taken at the surface of the Ti grid is shown in Figure3a, where the peaks are associated with a series of elemental O, Zn, Si, and Ti, which can be assigned to O-Kα1, Zn-Lα, Si-Kα1 (Si substrate from mounting the sample) and Ti-Kα. Moreover, the Zn/O ratio is estimated to be 1.05(5), which is close to the stoichiometric composition of ZnO, indicating the high purity of the nanowires. Figure3b depicts an EDS mapping of a cross section of the Ti grid of the selected sample (at TA = 400°C), where the distribution of elements is presented using the lock-in energy of Ti-Kα (4.3 to 4.7 keV), O-Kα1 (0.4 to 0.6 keV), and Zn-Lα (0.8 to 1.2 eV), respectively. The formation of ZnO nanowires can be mapped by EDS observations through cross section of the Ti grid. It can be seen that the core is dominated by Zn that diffuses into the hollow Ti grid to form a Zn core (blue color), Ti shell (green color), and surface ZnO film (orange color), respectively. The line profile EDS analysis extracted from EDS mapping clearly shows the distribution of oxygen in Figure3c. As shown in Figure3c, the width of the ZnO thickness <s> can be used to estimate the length of diffusion of the zinc at various TA and time τ. Details related to the estimating length are listed in Table1. Thermal treatment of the Zn on the Ti grid is known to influence the rates of oxide growth during nucleation and nanowire formation. The diffusion length is also sensitive to the thermal treatment time. A diffusion model is employed to interpret the oxidation kinetics wherein zinc transport proceeds in ZnO both by short-circuit and lattice diffusions.
Figure 3

Typical EDS spectrum, two-dimensional EDS mapping images, and line profile EDS analysis. (a) Typical EDS spectrum revealing the elemental composition of the selected sample at TA = 400°C. (b) Two-dimensional EDS mapping images of the distribution of elements presented using the lock-in energy of Ti-Kα (4.3 to 4.7 keV), O-Kα1 (0.4 to 0.6 keV), and Zn-Lα (0.8 to 1.2 eV). (c) Line profile EDS analysis clearly showing the presence of O in the sample. The line width of the distribution can be used to define the mean diffusion length <s>.

Table 1

ZnO thickness < s > along with simulated results

TA (K)

< s> (μm)


QS = γQD (kJ/mol)






4.21 (5)











Results for the lattice and short-circuit diffusion theories over the complete range of annealing temperatures, where QS is the short-circuit activation energy. Lattice activation energy QD = 318 kJ/mol[20].

Short-circuit diffusion in ZnO nanowires

In case of ZnO, where the diffusion coefficient of zinc is higher than that of oxygen, lattice diffusion takes place, resulting in an increase in the forming of ZnO with annealing temperature. It is well known that annealing of the Zn at a temperature above 300°C results in parabolic growth of ZnO nanocrystals (seen in Figure1)i. The contribution from boundary diffusion at the grain boundaries decreases with increasing nanocrystal size[21, 22]. It is therefore important to consider the influence of the scale on zinc transport when metal facets are oxidized at temperatures less than one-half the melting point of ZnO. At these temperatures, recrystallization and grain growth proceed slowly, with polycrystalline oxide boundaries serving as effective short-circuit diffusion paths[23]. A simple diffusion model is employed to interpret the oxidation kinetics. In this model, Zn transport proceeds in ZnO both by short-circuit diffusion at the grain boundaries and by lattice diffusion at higher annealing temperatures. The theory of lattice diffusion fully explains the diffusion mechanism[20, 2426]. In order to clarify the contribution from short-circuit or lattice diffusion, we have defined a diffusion ratio γ taken as the percentage of lattice activation energy of Zn ions (where γ = 1/3 and 1 indicates short-circuit or lattice diffusion, respectively). The estimated thickness <s> is related to the diffusion length D Zn * · τ that can be obtained from the lattice diffusivity D Zn * cm 2 s - 1 utilizing the following formula:
D Zn * = β 2 α 2 v D exp γ Q D R T A ,
where β = 2 is the number of positions a Zn atom can jump along the [1 0 0] plane; α = 0.285 nm, the d-spacing of the [1 0 0] plane; vD 1.73 × 1011 s−1, the vibration frequency;bτ, the growth time (approximately 7,200 s); QD = 318 kJ/mol, the activation energy of Zn[27]; R (1.987 cal mol K−1), the gas constant; and TA (K), the annealing temperature. The resultant diffusion ratio γ dependence on the annealing temperature for ZnO nanowire, together with < s >, was shown in Figure4. Details related to the fitting parameters are listed in Table1. The γ value of approximately 0.26(6) is slightly lower than the short-circuit diffusion predicted value of approximately 1/3 at lower TA = 400°C, revealing that the growth of nanowires is influenced by grain boundaries, whether oxygen or zinc is the diffusion species along short-circuit path dislocations, resulting in the growth of 1-D ZnO nanowalls at the lower temperature regime. These characteristics agree with previously reported for a single CuO nanowire that occurs through the short-circuit diffusion mechanism obtained by Cheng et al.[25]. At higher TA, corresponding to our experimental data, the value of γ is close to 0.41(3), indicating that lattice diffusion is the dominant transport mechanism in the oxidation of Zn that forms a 3-D ZnO nanowire network.
Figure 4

Plots of diffusion ratio γ dependence on the annealing temperature for ZnO nanowire together with < s >.

In general, at high temperatures where the difference between lattice diffusivity and short-circuit diffusivity is relatively small, but at low temperatures, such at 400°C, the short-circuit diffusivity is many orders of magnitude greater than lattice diffusivity. Short-circuit diffusion makes the greatest contribution to the net flux (mass transport through a unit area) for the growth of nanowires at low temperatures. In our previous study[17], at lower TA (approximately 400°C), it reveals that it is likely that the nanowire formation proceeds through the nucleation of ZnO x and ZnO and that a boundary transition occurred during the growth process. The formation of a single ZnO nanowire is attributed to crystallization from the Zn/ZnO mixed phase to form the ZnO structure. As a result, such mixed phase processes can also explain why the ZnO nuclei grow into 1-D ZnO nanowires through short-circuit diffusion. As the annealing temperature increases, the thermal-enhanced surface diffusion occurs at the nodes of the ZnO nanowire, which are favored to form 3-D nanowires and nanoflowers at higher TA. This result is agreed with previous reported by Ng and co-authors for the growth of epitaxial ZnO nanowires[28].


In summary, a diffusion ratio was obtained to deeply explore dimensionality-controllable synthesis of ZnO nanowires. For a lower dimensionality of ZnO nanowires, the γ value closes to 0.26(6), revealing a short-circuit diffusion mechanism. However, this tends to have a higher value of 0.41(3) as a further increase of annealing temperature results in the formation of 3-D ZnO nanowire network through lattice diffusion mechanism. These findings help us to proceed the fabrication of other novel nanostructure materials at various dimensionalities and the application in energy storage or memory devices through this growth mechanism.


aThe log-normal distribution is defined as follows:
f ( d ) = 1 2 π d σ exp ( ln d ln < d > ) 2 2 σ 2 ,
Where <d> is the mean value, and σ is the standard deviation of the function. bThe vibrational frequency of a Zn atom is defined as follows:
v D = 1 2 Q D m α 2 1 2 ,

where α = 0.285 nm is the d-spacing of the [100] plane; m = 65.4 g/mol, the Zn molar weight; and QD = 318 kJ/mol, the activation energy of Zn.



This research was supported by a grant from the National Science Council of Taiwan, the Republic of China, under grant number NSC-100-2112-M-259-003-MY3.

Authors’ Affiliations

Department of Physics, National Dong Hwa University


  1. Huang MH, Mao S, Feick H, Yan H, Wu Y, Kind H, Weber E, Russo R, Yang P: Room-temperature ultraviolet nanowire nanolasers. Science 2001, 292: 1897–1899. 10.1126/science.1060367View ArticleGoogle Scholar
  2. Könenkamp R, Word RC, Schlegel C: Vertical nanowire light-emitting diode. Appl Phys Lett 2004, 85: 6004–6006. 10.1063/1.1836873View ArticleGoogle Scholar
  3. Nadarajah A, Word RC, Meiss J, Könenkamp R: Flexible inorganic nanowire light-emitting diode. Nano Lett 2008, 8: 534–537. 10.1021/nl072784lView ArticleGoogle Scholar
  4. Wang HT, Kang BS, Ren F, Tien LC, Sadik PW, Norton DP, Pearton SJ, Lin J: Hydrogen-selective sensing at room temperature with ZnO nanorods. Appl Phys Lett 2005, 86: 243503. 10.1063/1.1949707View ArticleGoogle Scholar
  5. Law M, Greene LE, Johnson JC, Saykally R, Yang P: Nanowire dye-sensitized solar cells. Nat Mater 2005, 4: 455–459. 10.1038/nmat1387View ArticleGoogle Scholar
  6. Alivisatos AP: Semiconductor clusters, nanocrystals, and quantum dots. Science 1996, 271: 933–937. 10.1126/science.271.5251.933View ArticleGoogle Scholar
  7. Banin U, Cao YW, Katz D, Millo O: Identification of atomic-like electronic states in indium arsenide nanocrystal quantum dots. Nature 1999, 400: 542–544. 10.1038/22979View ArticleGoogle Scholar
  8. Zhou J, Ding Y, Deng SZ, Gong L, Xu NS, Wang ZL: Three-dimensional tungsten oxide nanowire networks. Adv Mater 2005, 17: 2107–2110. 10.1002/adma.200500885View ArticleGoogle Scholar
  9. Srinivasarao M, Collings D, Philips A, Patel S: Three-dimensionally ordered array of air bubbles in a polymer film. Science 2001, 292: 79–83. 10.1126/science.1057887View ArticleGoogle Scholar
  10. Wang S, Zeng C, Lai S, Juang YJ, Yang Y, Lee LJ: Self-rolled polymer and composite polymer/metal micro- and nanotubes with patterned inner walls. Adv Mater 2005, 17: 1177–1182. 10.1002/adma.200401836View ArticleGoogle Scholar
  11. Wang RC, Liu CP, Huang JL, Chen SJ: ZnO symmetric nanosheets integrated with nanowalls. Appl Phys Lett 2005, 87: 053103. 10.1063/1.2005386View ArticleGoogle Scholar
  12. Fan HJ, Scholz R, Kolb FM, Zacharias M: Two-dimensional dendritic ZnO nanowires from oxidation of Zn microcrystals. Appl Phys Lett 2004, 85: 4142–4144. 10.1063/1.1811774View ArticleGoogle Scholar
  13. Xinzheng L, Yang J, Xinmei L, Wenjun W, Wang Binbin Wu, Di LC, Yugang Z, Honghai Z: Large-scale growth of a novel hierarchical ZnO three-dimensional nanostructure with preformed patterned substrate. Cryst Growth Des 2011, 11: 3837–3843. 10.1021/cg200384hView ArticleGoogle Scholar
  14. Lee WW, Yi J, Kim SB, Kim YH, Park HG, Park WI: Morphology-controlled three-dimensional nanoarchitectures produced by exploiting vertical and in-plane crystallographic orientations in hydrothermal ZnO crystals. Cryst Growth Des 2011, 11: 4927–4932. 10.1021/cg200806aView ArticleGoogle Scholar
  15. Gao PX, Lao CS, Hughes WL, Wang ZL: Three-dimensional interconnected nanowire networks of ZnO. Chem Phys Lett 2005, 408: 174–178. 10.1016/j.cplett.2005.04.024View ArticleGoogle Scholar
  16. Song HS, Zhang WJ, Cheng C, Tang YB, Luo LB, Chen X, Luan CY, Meng XM, Zapien JA, Wang N, Lee CS, Bello I, Lee ST: Controllable fabrication of three-dimensional radial ZnO nanowire/silicon microrod hybrid architectures. Cryst Growth Des 2011, 11: 147–153. 10.1021/cg101062eView ArticleGoogle Scholar
  17. Gandhi AC, Hung HJ, Shih PH, Cheng CL, Ma YR, Wu SY: In situ confocal Raman mapping study of a single Ti-assisted ZnO nanowire. Nanoscale Res Lett 2010, 5: 581–586. 10.1007/s11671-009-9509-1View ArticleGoogle Scholar
  18. Wagner C: Beitrag zur Theorie des Analufvorgangs. Z Phys Chem B 1933, 21: 25.Google Scholar
  19. Jeong MC, Oh BY, Lee W, Myoung JM: Optoelectronic properties of three-dimensional ZnO hybrid structure. Appl Phys Lett 2005, 86: 103105. 10.1063/1.1872209View ArticleGoogle Scholar
  20. Gandhi AC, Huang CY, Yang CC, Chan TS, Cheng CL, Ma YR, Wu SY: Growth mechanism and magnon excitation in NiO nanowalls. Nanoscale Res Lett 2011, 6: 485. 10.1186/1556-276X-6-485View ArticleGoogle Scholar
  21. Kashchiev D: Dependence of the growth rate of nanowires on the nanowire diameter. Cryst Growth Des 2006, 6: 1154–1156. 10.1021/cg050619iView ArticleGoogle Scholar
  22. Kofstad P: Nonstoichiometry, Diffusion and Electrical Conductivity in Binary Metal Oxides. John Wiley, New York; 1972.Google Scholar
  23. Herchl R, Khoi NN, Homma T, Smeltzer WW: Short-circuit diffusion in the growth of nickel oxide scales on nickel crystal faces. Oxid Met 1972, 4: 35–49. 10.1007/BF00612506View ArticleGoogle Scholar
  24. Shewmon P: Diffusion in Solids. 2nd edition. TMS, Warrendale; 1989.Google Scholar
  25. Cheng CL, Ma YR, Chou MH, Huang CY, Yeh V, Wu SY: Direct observation of short-circuit diffusion during the formation of a single cupric oxide Nanowire. Nanotechnology 2007, 18: 245604. 10.1088/0957-4484/18/24/245604View ArticleGoogle Scholar
  26. Wu SY: Design and controlling of in-plane CuO nanowires: an investigation of growth mechanism and phonon confinement effect. J Nanosci Lett 2012, 2: 5.Google Scholar
  27. Gao F, Chino N, Naik SP, Sasaki Y, Okubo T: Photoelectric properties of nano-ZnO fabricated in mesoporous silica film. Mater Lett 2007, 61: 3179–3184. 10.1016/j.matlet.2006.11.033View ArticleGoogle Scholar
  28. Ng HT, Li J, Smith MK, Nguyen P, Cassell A, Han J, Meyyappan M: Growth of epitaxial nanowires at the junction of nanowalls. Science 2003, 23: 1249.View ArticleGoogle Scholar


© Shih et al.; licensee Springer. 2012

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 cited.