Investigations of a Statistical and Analytical Method to Find the Relationship between the Morphological and Optical Properties of ZnO Nanoflower Arrays.

In this study, a sapphire substrate with a patterned concave structure was used to prepare ZnO film/A-B glue, and the ZnO film/A-B glue with a patterned convex matrix was transferred onto a silicon wafer using the lift-off technology as the seed layer. Then, the hydrothermal method with different Zn(CH3COO)2 and C6H12N4 concentrations as precursors was used to synthesize ZnO nanoflower arrays on the patterned convex ZnO seed layer. XRD pattern, FESEM, FIB, and photoluminescence (PL) spectrometry were employed to observe and analyze the properties of the synthesized ZnO nanoflower arrays. When Zn(CH3COO)2 and C6H12N4 concentrations were 0.01, 0.02, 0.03, and 0.04 M, the average heights of the ZnO nanorods in the ZnO nanoflower arrays were 993, 1500, 1550, and 1650 nm, the average diameters of the ZnO nanorods were 50, 90, 105, and 225 nm, and the aspect ratios (H/D) of the ZnO nanorods were 19.9, 16.7, 14.8, and 7.33, respectively. A simple statistical and analytical method was investigated to estimate the densities (number of nanorods) of the ZnO nanoflower arrays in one 1 μm × 1 μm area. The total surface area (S) of the ZnO nanoflower arrays first increased from 5.05 × 106 and then reached a maximum value of 1.20 × 107 nm2 as Zn(CH3COO)2 and C6H12N4 concentrations increased from 0.01 to 0.02 M. For the systhesized ZnO nanoflower arrays, as the Zn(CH3COO)2 and C6H12N4 concentrations increased from 0.01 to 0.04 M, their total volume (V) increased from the 6.23 × 107 to 5.90 × 108 nm3 and the S/V ratio decreased from 8.10 × 10-2 to 1.84 × 10-2. We found that ZnO nanoflower arrays with Zn(CH3COO)2 and C6H12N4 concentrations of 0.2 M presented the maximum PL emission intensities. The calculated S/V ratios and X-ray photoelectron spectroscopy analyses are used to discuss the reasons for these results.

Introduction

In the past, many methods were utilized to synthesize ZnO-based nanomaterials. For example, Huang et al. used the vapor–liquid–solid method to synthesize ZnO nanowires, and they could control the diameters of ZnO nanowires by varying the thickness of the Au catalyst film.[1] Katiyar et al. used the ultrasonic-assisted hydrothermal method to synthesize ZnO nanoflowers at 95 °C, and the nanoflowers were constructed by hexagonal-shaped nanorods to form the petal-like arrangement.[2] Ai et al. used a reactive vapor deposition method to synthesize ZnO nanoflowers with the outward appearance of flake petals.[3] The hydrothermal method is a low-temperature process to synthesize ZnO nanomaterials with different structures. For example, Chen et al. found that as SiO2/Si substrates faced different directions or ZnO-based materials grew at different times, ZnO nanomaterials had different structures, including chrysanthemum flower-like clusters, nonuniform nanorods with different lengths and diameters, large hexagonal bars, and nanoflowers.[4] When the hydrothermal method is used to synthesize ZnO nanomaterials, the synthesis parameters, including the solution concentration, solution type, synthesis temperature, and synthesis time, will affect their outward appearance and morphology.[5−7] However, as the synthesis parameters are well controlled, uniformed ZnO nanorods can successfully grow perpendicular to the substrate in a large area.[8]

When the hydrothermal method is used to synthesize ZnO nanomaterials, the degree of supersaturation can be divided into three different types.[9] The first is the low level of supersaturation, in which the heterogeneous nucleation will dominate the crystallization of the ZnO nanomaterials, and under these conditions, the typically produced ZnO nanomaterials are polyhedral crystals. Second is the intermediate supersaturation range, in which ZnO nanomaterials occur by a mixture of two-dimensional nucleation of clusters and slow spiral growth. In this region, the synthesized ZnO nanomaterials will yield a hollowed rod morphology. The third is characterized by high supersaturation, in which the homogeneous nucleation of ZnO nanomaterials becomes more important and dominates the crystallization. In this type, the crystal faces of the synthesized ZnO nanomaterials become rough and they synthesize into a dendritic morphology, and the crystal is columnar with a high aspect ratio. Usually, the concentrations of reactants can dominate the synthesis density, height, diameter, and aspect ratio of one-dimensional ZnO nanorods. In this study, we used the concentrations of zinc acetate dihydrate (Zn(CH3COO)2·2H2O, abbreviated as Zn(CH3COO)2), and hexamethylenetetramine (C6H12N4, HMT), abbreviated as C6H12N4, as the main factors in the investigation of the properties of synthesized ZnO nanoflower arrays.

When ZnO nanorods are used as the optoelectronic devices or gas sensors, their top surfaces will contact the optical and detection gas first.[10] As ZnO nanorods grow like radiating needles to form the nanoflowers or nanoflower arrays, the top areas of the first contact reaction sites can be increased greatly, which can enhance the reaction sensitivity of ZnO-based nanomaterials. When nanoflower arrays are used as optoelectronic devices or gas sensors, their reaction speeds and sensing responses are quicker because they have a larger side area.[11] Therefore, many researchers have investigated the synthesis of nanoflower arrays with different technologies. For example, Tang et al. used the grooves as a template, which were created by ZnO nanowalls and nanorods, to control the synthesis of ZnO-based nanoflowers along the grooves. They also found that the number of ZnO nanoflowers depended on the concentration of the solution.[12] Wang et al. prepared the monolayer polystyrene spheres (PS) using a self-assembled method, and through modifying the diameters of the polystyrene spheres, they could control the morphologies and dimensions of ZnO nanoflowers.[13] Katiyar et al. used the ultrasonic-assisted hydrothermal method to synthesize ZnO nanoflowers at 95 °C, but apparently, ZnO nanoflowers grew in a versatile direction.[2] However, the problems of these processes are that people have difficulty duplicating the fabrication processes to manufacture the templates and to manufacture ZnO nanoflowers with large areas. Therefore, we investigated a new technology in which the details were described clearly in the paragraph of experimental procedures to synthesize ZnO nanoflowers with an easy process and a large area, because the template could be easily fabricated.

In the past, many researchers have investigated the effects of different Zn(CH3COO)2) and C6H12N4 concentrations on the properties of ZnO nanorods, but almost no researcher has used different Zn(CH3COO)2) and C6H12N4 concentrations as the parameters to investigate the variations in the morphological and optical properties of ZnO nanoflower arrays. Because we found that, as the same synthesis time was used, the Zn(CH3COO)2) and C6H12N4 concentrations had apparent effects on the morphology of synthesized ZnO nanoflower arrays. Therefore, another novel aspect of this study is that we investigated the relationships and found the reasons for the variations between precursor concentrations and the heights, diameters, aspect ratios, and photoluminescence (PL) properties of ZnO nanoflower arrays. A series of experiments were performed by varying Zn(CH3COO)2 and C6H12N4 concentrations and keeping the parameters of synthesis time and temperature constant. After these analyses, a simple statistical and analytical method was investigated to estimate the densities of the ZnO nanoflower arrays in one 1 μm × 1 μm. The total volumes (V), the total surface areas (S), and the S/V ratios were calculated, and a relationship was found between the morphological and optical properties of the synthesized ZnO nanoflower arrays. Finally, the X-ray photoelectron spectroscopy (XPS) technique was used to find the defect variations in ZnO nanoflower arrays as the Zn(CH3COO)2 and C6H12N4 concentrations were changed, which were used to find the reason to cause the variations of optical properties of the ZnO nanoflower arrays.

Results and Discussion

The crystal properties of ZnO nanoflower arrays were analyzed and their XRD patterns are displayed in Figure as a function of Zn(CH3COO)2 and C6H12N4 concentrations. However, when the results in Figure are analyzed, there are large differences between the diffraction results of the ZnO nanoflower arrays. Figure depicts that the diffraction peaks of the (100), (002), (101), and (102) planes (JCPDS card with No. 36-1451) were really observed, which correspond to the peaks located at 2θ values 31.9°, 34.6°, 36.48°, and 47.64°. These diifraction results also prove that ZnO nanoflower arrays presented a Wurtzite structure with a hexagonally closed package. Figure also demonstrates that the diffraction intensities of the (100), (002), (101), and (102) planes increased with Zn(CH3COO)2 and C6H12N4 concentrations. The X-ray patterns of the ZnO nanoflower arrays indicate that the diffraction intensities of all the (100), (002), (101), and (102) planes increased as Zn(CH3COO)2 and C6H12N4 concentrations increased from 0.1 to 0.3 M. However, as Zn(CH3COO)2 and C6H12N4 concentrations increased from 0.3 to 0.4 M, the diffraction intensity of the (002) plane increased, but the diffraction intensities of the (100), (101), and (102) planes decreased, which demonstrates that the c-axis preferred orientation property of the ZnO nanoflower arrays increased with the increase in Zn(CH3COO)2 and C6H12N4 concentrations. In the next section we will demonstrate that the increase in the average diameter of ZnO nanoflower arrays with the increases of Zn(CH3COO)2 and C6H12N4 concentrations is the reason for this result.

Figure 1

XRD patterns of ZnO nanorflower arrays as a function of Zn(CH3COO)2 and C6H12N4 concentrations.

The surface morphologies and the cross-sectional observations of ZnO nanoflower arrays are displayed in Figures and 3, respectively, as a function of Zn(CH3COO)2 and C6H12N4 concentrations. As Figures and 3 show, because the ZnO seed layer has a convex structure and ZnO nanorods are grown in the radial direction and vertical to the substrate surface, therefore a ZnO nanoflower is really formed. Because the ZnO seed layer is in a patterned array structure, ZnO nanorflower arrays are formed. When Zn(CH3COO)2 and C6H12N4 concentrations were 0.1 M, as Figures a and 3a show, ZnO nanorods were too fine to cause the weak diffraction intensities of the (100), (002), (101), and (102) planes. When Zn(CH3COO)2 and C6H12N4 concentrations increased from 0.1 to 0.3 M, as Figure a–c depicts, most ZnO nanorods were grown in a radial direction and the diameter increased with Zn(CH3COO)2 and C6H12N4 concentrations; therefore, the diffraction intensities of the (100), (002), (101), and (102) planes increased. As Zn(CH3COO)2 and C6H12N4 concentrations were 0.4 M, the diameter of ZnO nanorods increased apparently, the pushing effect during synthesis process was really observed, as Figure d shows. Therefore, the probability of upward growth increases and then the diffraction intensity of the (002) plane increases and then the diffraction intensities of the (100), (101), and (102) planes decrease. Figures and 3 also demonstrate that the density of ZnO nonorods significantly reduced in ZnO nanoflower arrays, which suggests that the crystal property of ZnO nanoflower arrays can be enhanced with the increases of Zn(CH3COO)2 and C6H12N4 concentrations. These results prove that the advantage of the prepared ZnO seed layer is that it enables a versatile growth, in that the vertical bottom growth of ZnO nanorods can be controlled in a radial direction to form ZnO nanoflowers rather than in random directions. Because we prepare the ZnO seed layer in the format of a matrix pattern, the synthesized ZnO nanoflowers do not have a random distribution, but rather they are synthesized according to the structure of convex ZnO seed layer to form ZnO nanoflower arrays.

Figure 2

Surface morphologies of ZnO nanoflower arrays as a function of Zn(CH3COO)2 and C6H12N4 concentrations: (a) 0.01, (b) 0.02, (c) 0.03, and (d) 0.04 M.

Figure 3

Cross-sectional observations of ZnO nanoflower arrays as a function of Zn(CH3COO)2 and C6H12N4 concentrations: (a) 0.01, (b) 0.02, (c) 0.03, and (d) 0.04 M.

The results of XRD patterns in Figure also indicate that the c-axis preferred orientation (the mainly crystalline peak is (002) plane) is not the main growth direction of ZnO nanoflower arrays, the growth direction of ZnO nanoflower arrays is perpendicular to substrate and spreads out like a flower, as Figures and 3 display. Additionally, the two figures show that as the synthesis time and temperature are unchanged, the height, diameter, and aspect ratio of ZnO nanoflower arrays are closely related to Zn(CH3COO)2 and C6H12N4 concentrations. The average heights, average diameters, and aspect ratios of ZnO nanorflower arrays are estimated and compared in Table as a function of Zn(CH3COO)2 and C6H12N4 concentrations. When Zn(CH3COO)2 and C6H12N4 concentrations were 0.01, 0.02, 0.03, and 0.04 M, the average heights of ZnO nanorods in ZnO nanoflower arrays were 993 (the diameters were in the range of 972–1009 nm), 1500 (1478–1529 nm), 1550 (1533–1576 nm), and 1650 nm (1627–1682 nm), and the average diameters of ZnO nanorods were 50 (the diameters were in the range of 44–55 nm), 90 (83–96 nm), 105 (99–113 nm), and 225 nm (218–224 nm).

Table 1

Relative Results of ZnO Nanoflower Arrays as a Function of Zn(CH3COO)2 and C6H12N4 Concentrations

concentration (M)	0.01	0.02	0.03	0.04
avg diameter (D, nm)	50	90	105	225
avg height (H, nm)	993	1500	1550	1650
aspect ratio (H/D)	19.9	16.7	14.8	7.33
total volume (V, nm3)	6.23 × 107	2.67 × 108	2.95 × 108	5.90 × 108
total surface area (S, nm2)	5.05 × 106	1.20 × 107	1.14 × 107	1.08 × 107
density (No. per μm2)	32 ± 4	28 ± 4	22 ± 3	9 ± 2
S/V ratio	8.10 × 10–2	4.51 × 10–2	3.87 × 10–2	1.84 × 10–2

The results in Figures and 3 prove again that as the same synthesis times and temperatures are used, Zn(CH3COO)2 and C6H12N4 concentrations are an important factor to affect the appearance and morphology of ZnO nanoflower arrays, and the average height (H) and diameter (D) of ZnO nanorods in ZnO nanoflower arrays also increase with Zn(CH3COO)2 and C6H12N4 concentrations. Correspondingly, the aspect ratio (H/D) of ZnO nanorods in ZnO nanoflower arrays decreases from 19.9 to 7.33 when Zn(CH3COO)2 and C6H12N4 concentrations increase from 0.01 to 0.04 M, as Table demonstrates. When Zn(CH3COO)2 and C6H12N4 are used as sources to synthesize ZnO nanomaterials, the Zn2+ ions would react with OH– and Zn(OH)2.[8] Because the mixed solution is heated, Zn(OH)2 will decompose into H2O and ZnO, and the precipitated ZnO nuclei are synthesized to form the ZnO nanomaterials on the convex ZnO film/A-B glue seed layer. When Zn(CH3COO)2 and C6H12N4 concentrations increase, the relative concentration of Zn(OH)2 (or ZnO nuclei) produced from the decomposition of the precursors increases correspondingly, and then during the synthesis process, the synthesis speeds in the radial-axis and vertical-axis directions also increase. Because the Zn(CH3COO)2 and C6H12N4 concentrations affect the relative synthesis rates of ZnO nanorods in the radial-axis and vertical-axis directions, they affect the aspect ratio of ZnO nanorods in ZnO nanoflower arrays. Therefore, it can be used to control the diameter, length, and aspect ratio of ZnO nanorods in ZnO nanoflower arrays by adjusting Zn(CH3COO)2 and C6H12N4 concentrations.

In order to estimate the relationships between the diameter, height, aspect ratio, total surface area (S, nm2), total volume (V, nm3), and S/V ratio of ZnO nanoflower arrays, a simple model is represented in Figure to estimate their differences with the variations of Zn(CH3COO)2 and C6H12N4 concentrations. For the synthesized ZnO nanoflower arrays, because ZnO nanorods have the structure of hexagonal cylinder rather than that of column.[14] Therefore, the volumes in Figure are and as the side length is a and the height is decreased from a to a/2. As the side lengths are a/2, a/4, and a/6, and the height is a, the total surface areas are , , and ; As the side lengths are a/2, a/4, and a/6, and the height is a/2, the total surface areas are , , and , respectively. Therefore, as the side lengths are a/2, a/4, and a/6, and the height is a, the S/V ratios are , , and ; As the side lengths are a/2, a/4, and a/6 and the height is a/2, the S/V ratios are , , and , respectively.

Figure 4

Schematic diagrams of the fabrication of ZnO nanoflower arrays with different side lengths.

The aspect ratio of ZnO nanoflower arrays is also determined by the relative growth rates of the radial-axis and vertical-axis directions. In order to estimate the density of ZnO nanorods in a unit area (1 μm2 defined here), we also segmented the surface area of ZnO nanoflower arrays into the square each with area of 1 μm2, as Figure c depicts. The average densities of ZnO nanoflower arrays also decreased from 32 to 9 μm–2 as Zn(CH3COO)2 and C6H12N4 concentrations increased from 0.01 to 0.04 M, as Table illustrates. Next, the total volume (V, nm3), the total surface area (S, nm2), and the S/V ratio of ZnO nanorflower arrays were also estimated, and the results are compared in Table as a function of Zn(CH3COO)2 and C6H12N4 concentrations.

As Zn(CH3COO)2 and C6H12N4 concentrations increased from 0.01 to 0.04 M, the total volume of ZnO nanoflower arrays increased from the 6.23 × 107 to 5.90 × 108 nm3. The total surface area first increased from 5.05 × 106 and reached a maximum value of 1.20 × 107 nm2 as Zn(CH3COO)2 and C6H12N4 concentrations increased from 0.01 to 0.02 M, and the total surface area decreased to 1.14 × 107 and 1.08 × 107 as Zn(CH3COO)2 and C6H12N4 concentrations further increased to 0.03 and 0.04 M. The S/V ratio of ZnO nanoflower arrays decreased from 8.10 × 10–2 to 1.84 × 10–2 as Zn(CH3COO)2 and C6H12N4 concentrations increased from 0.01 to 0.04 M. Therefore, the diameter, length, and aspect ratio of ZnO nanorods in ZnO nanoflower arrays can also be controlled by adjusting Zn(CH3COO)2 and C6H12N4 concentrations. Table depicts that the average diameter and average length of ZnO nanorods in ZnO nanoflower arrays increased with Zn(CH3COO)2 and C6H12N4 concentrations. Therefore, the reason to cause the increases of total surface area and total volume is that the diameter and length of ZnO nanorods increase with Zn(CH3COO)2 and C6H12N4 concentrations, even the average density of ZnO nanorods in ZnO nanoflower arrays decreases with Zn(CH3COO)2 and C6H12N4 concentrations. However, we believe that Zn(CH3COO)2 and C6H12N4 concentrations have an apparent effect on the total surface area of ZnO nanoflower arrays and then further on the PL properties of ZnO nanoflower arrays. However, as Table shows, as Zn(CH3COO)2 and C6H12N4 concentrations increased from 0.01 to 0.04 M, the average height, the average diameter, and the total volume increase, the aspect ratio, the density decrease, and the S/V ratio decrease, but the total surface area had a maximum value at 0.02 M.

Figure shows the PL properties of ZnO nanoflower arrays as a function of Zn(CH3COO)2 and C6H12N4 concentrations. The values of full width at half-maximum (fwhm) for the IUV peaks (the strong emission peaks located at about 379.8–378.2 nm) of the PL spectra were 23.9 nm (373.2–397.1 nm), 17.2 nm (371.5–388.7 nm), 18.8 nm (371.4–390.2 nm), and 22.6 nm (369.9–392.5 nm), respectively, as Zn(CH3COO)2 and C6H12N4 concentrations were 0.01, 0.02, 0.03, and 0.04 M. Apparently, as the emission peak is stronger, the fwhm value is smaller. As Figure illustrates, for the PL properties of ZnO nanoflower arrays, the emission intensities of IUV and the IG did not have positive correlation or negative correlation with Zn(CH3COO)2 and C6H12N4 concentrations. IUV value increased and IG (the weak emission peaks located in the range of visible light and centered at about green light) value increased slightly and achieved a maximum at 0.02 M, and then IUV and IG values decreased as Zn(CH3COO)2 and C6H12N4 concentrations was further increased to 0.03 and 0.04 M. As Table demonstrates, even ZnO nanoflower arrays with 0.01 M Zn(CH3COO)2 and C6H12N4 concentrations have higher aspect and S/V ratios, they have the least total surface area, and their structure is too slender and scattered; therefore, they cannot have an effective luminescence.

Figure 5

PL spectra of ZnO nanorflower arrays as a function of Zn(CH3COO)2 and C6H12N4 concentrations.

Apparently, these results again prove that the variation in the UV emission intensity of ZnO nanoflower arrays has no apparent relation to the total volume and aspect ratio, but it has the positive correlation with total surface area, because more the total surface area is, more area will absorb exciting light and emit UV light. However, the IUV value of ZnO nanoflower arrays also exhibits a positive correlation with total surface area but not with Zn(CH3COO)2 and C6H12N4 concentrations. The reason for this is that the crystal properties of ZnO nanoflower arrays are enhanced with Zn(CH3COO)2 and C6H12N4 concentrations (Figure ) and the c-axis preferred orientation is really observed in ZnO nanoflower arrays. As Zn(CH3COO)2 and C6H12N4 concentrations are 0.2 M, ZnO nanoflower arrays have the maximum surface area, and then the UV light and green light have the maximum emission intensities. As Zn(CH3COO)2 and C6H12N4 concentrations are more than 0.02 M, the total surface areas of ZnO nanoflower arrays to emit UV light and green light decrease; therefore, the IUV and IG values also decreases.

The UV emission in the PL spectra of ZnO nanoflower arrays is recognized as the near-band-edge emission. Correspondingly, as Zn(CH3COO)2 and C6H12N4 concentrations increased, the UV emission intensity of ZnO nanoflower arrays first increased and achieved a maximum at 0.02 M, and then the UV emission intensity decreased as Zn(CH3COO)2 and C6H12N4 concentrations were further increased to 0.03 and 0.04 M. When Zn(CH3COO)2 and C6H12N4 concentrations increase, even the height and diameter growth rates of ZnO nanoflower arrays increase, the total surface area decrease as Zn(CH3COO)2 and C6H12N4 concentrations are more than 0.02 M. Apparently, the variation in the UV emission intensity of ZnO nanoflower arrays has no apparent relation to the total volume and aspect ratio, but it has a positive correlation with total surface area, and the increase of aspect ratio is mainly due to the increase of total surface area. These results prove again that the emission intensity of UV light is positively correlated with a total surface area of ZnO nanoflower arrays. When Zn(CH3COO)2 and C6H12N4 concentrations are more than 0.02 M, the total surface area of ZnO nanoflower arrays absorb exciting light and emit UV light decreases and the emission intensity decreases.

Optical properties of ZnO-based nanomaterials are an important index for their applications in many different technologies. ZnO-based materials have many different luminescence mechanisms; for example, ultraviolet light emission caused by near-band-edge emission (at 3.18 eV or 390 nm), violet light emission by zinc vacancies (VZn at 3.06 eV or 405 nm), violet light emission by interstitial zinc (Zni at 2.99 eV or 428 nm), green light by antisite defect (OZn, at 2.38 eV or 521 nm), green light by interstitial oxygen (Oi, at 2.28 eV or 544 nm), and near-infrared by oxygen vacancies (VO at 1.62 eV or 765 nm).[15] The PL spectra of ZnO seed layer and ZnO nanoflower arrays synthesized with different Zn(CH3COO)2 and C6H12N4 concentrations are compared in Figure ; all had a sharp emission peak and one broad emission band. The sharp peak was located at 379.8 for the ZnO seed layer, and they were located at 379.8, 378.4, 378.4, and 378.2 nm as Zn(CH3COO)2 and C6H12N4 concentrations were increased from 0.01, 0.02, 0.03, and 0.04 M. Wu et al. found that as the ZnO film was annealed in a N2 atmosphere at a higher temperature, the near-band-edge emission peak shifted to lower wavelength.[16] Therefore, the wavelength shift of the sharp peak may originate from the variations of internal defects in the ZnO nanorods itself.[17]

The emission intensities of all ZnO nanoflower arrays in the visible emission band were smaller than that of ZnO seed layer, and they had the same trend with the variations of the UV emission band. These results indicate that the defects in ZnO nanoflower arrays are less than in ZnO seed layer. The visible emission band in the PL spectra was usually observed in most reported studies for ZnO nanomaterials. Yousefi and Kamaluddin also found that there are two bands in the PL spectra of ZnO nanorods; the first is related to the deep level emission at 508–522 nm and the other is related to the UV emission at 378–382 nm, which is the near-band-edge emission.[18,19] They found that the green band emission corresponds to the singly ionized oxygen vacancy in ZnO, which results from the recombination of a photogenerated hole with the singly ionized charge state of this defect. Yousefi also found that the weak green band of the undoped ZnO nanorods in the PL spectrum is that ZnO nanorods have very low concentrations of oxygen vacancies.[19] These results suggest that the oxygen vacancies are low in our synthesized ZnO nanoflower arrays. Many studies also assumed that the green emission is caused by intrinsic defects in ZnO nanomaterials, such as interstitial zinc (Zni), zinc vacancies (VZn), and oxygen vacancies (VO), or their complexes.[20,21] For example, Vanheusden et al. demonstrated that the free-carrier depletion existed on the surfaces of ZnO particles, and it had an apparent effect on the ionization state of the oxygen vacancies that had strongly impacted the intensity of green emission.[21] However, they also proved that as the defect of oxygen vacancies decreased, the emission intensity of green light also decreased. Therefore, the crystal quality of ZnO nanomaterials is the most important factor affecting their PL properties.

The broad emission band of ZnO seed layer ranged from 415 to 590 nm and centered at 492 nm, and the broad emission bands of all ZnO nanoflower arrays ranged from 435 to 640 nm and centered at 545 nm. Therefore, the defects in the ZnO seed layer are believed to be closely related to the defects induced by the complexes of interstitial zinc defect (428 nm), antisite defect (521 nm), and interstitial oxygen (544 nm), and the defects in all ZnO nanoflower arrays are believed to be closely related to the defects induced by the complexes of antisite defect (521 nm) and interstitial oxygen (544 nm). If the intensities of ultraviolet and green light are represented by IUV and IG, respectively, and the IG/IUV ratio represents the number of defects existing in ZnO nanoflower arrays. Table displays the relationships of IUV and IG values and IG/IUV ratios of ZnO nanoflower arrays as a function of Zn(CH3COO)2 and C6H12N4 concentrations. As Figure and Table demonstrate, the IG value first increased and reached a maximum at 0.02 M and then decreased as Zn(CH3COO)2 and C6H12N4 concentrations were further increased. This result proves that the defects existing in ZnO seed layer decreases as ZnO nanoflower arrays are synthesized on it.

Table 2

Relationships of IUV and IG Values and IG/IUV Ratios of ZnO Nanoflower Arrays as a Function of Zn(CH3COO)2 and C6H12N4 Concentrations

concentration (M)	IUV	IG	IG/IUV
0.01	268	89.7	0.335
0.02	2597	163	0.063
0.03	2246	101	0.045
0.04	1955	84.8	0.043

The XPS technique was used to measure the chemical-bonding state of oxygen, which can be used to find the reasons for causing the variations in the defects and optical properties of ZnO nanoflower arrays. The drawings of O1s peaks of ZnO nanoflower arrays with the Zn(CH3COO)2 and C6H12N4 concentrations of 0.02 and 0.04 M, which are centered at 530.1 ± 0.1 eV, are shown in Figure a,b. The typical surface O1s peaks of ZnO nanoflower arrays were fitted by three Gaussian components of OI, OII, and OIII peaks. The divided three peaks of O1s are that OI, OII, and OIII were centered at 523.0 ± 0.1, 530.4 ± 0.1, and 532.1 ± 0.1 eV, respectively. The Zn2+–O2– bond will cause the component of the OI peak, the oxygen vacancies within ZnO nanoflower arrays will cause the bond energy for the component of the OII peak, and the chemical absorption of oxygen on the surfaces of ZnO nanoflower arrays will cause the bond energy for the component of the OIII peak.[22]

Figure 6

Drawings of the O1s peak of the ZnO nanoflower arrays with the Zn(CH3COO)2 and C6H12N4 concentrations of (a) 0.02 and (b) 0.04 M.

The areas of O1s peaks measured from the XPS spectra of ZnO nanoflower arrays are compared in Table as a function of Zn(CH3COO)2 and C6H12N4 concentrations. For ZnO nanoflower arrays, the area of the OI peak first decreased from 50.64% to 44.30%, the area of the OII peak decreased from 41.13% to 42.36%, and the area of the OIII peak increased from 8.84% to 13.34% as the Zn(CH3COO)2 and C6H12N4 concentrations increased from 0.01 to 0.02 M; As Zn(CH3COO)2 and C6H12N4 concentrations further increased from 0.02 to 0.04 M, the area of the OI peak first increased from 44.30% to 50.64%, the area of the OII peak decreased from 42.36% to 40.50%, and the area of the OIII peak increased from 13.34% to 9.62%. Apparently, the area of the OII peak of ZnO nanoflower arrays has a maximum value for the Zn(CH3COO)2 and C6H12N4 concentrations of 0.02 M, which prove that the oxygen vacancies of ZnO nanoflower arrays have a maximum for ZnO nanoflower arrays with the Zn(CH3COO)2 and C6H12N4 concentrations of 0.02 M, which match the measured results shown in Figure . These result also indirectly prove that the higher IUV and IG values of ZnO nanoflower arrays are caused by the high total surface area.

Table 3

Areas of O1s Peaks of ZnO Nanoflower Arrays as a Function of Zn(CH3COO)2 and C6H12N4 Concentrations

concentration	0.01	0.02	0.03	0.04
OI	49.25%	44.30%	46.63%	50.64%
OII	41.13%	42.36%	41.79%	40.50%
OIII	8.84%	13.34%	11.58%	9.62%

Conclusions

The results demonstrate that Zn(CH3COO)2 and C6H12N4 concentrations had a significant effect on the height, diameter, total surface area, total volume, and density, and PL properties of ZnO nanoflower arrays. The height and diameter increased and the density decreased with Zn(CH3COO)2 and C6H12N4 concentrations, but the total volume and aspect ratio had different trends in ZnO nanoflower arrays. For ZnO nanoflower arrays, as Zn(CH3COO)2 and C6H12N4 concentrations increased, all the IUV and IG values and the IG/IUV ratio first increased and reached a maximum at 0.02 M. From the analyzed and measured results, it was found that the total surface areas and PL properties of ZnO nanoflower arrays had the positive correlation. The XPS analyses also showed that the area of the OII peak first increased and reached a maximum at 0.02 M, which had the same trend with the variation of IG value. These results indicate that the total surface areas are the most important factor to affect the optical properties of ZnO nanoflower arrays.

Experimental Procedures

The facile hydrothermal method was used to synthesize ZnO nanoflower arrays on the template-assisted deposited ZnO seed arrays. For the preparation of the ZnO gel, ethylene glycol monomethyl ether (CH3OCH2CH2OH), 2-aminoethanol (monoethanolamine) (C2H7NO), and zinc acetate dihydrate (Zn(CH3COO)2·2H2O) were used to prepare a solution of 0.75 M Zn+ ions. The solution was heated to 60 °C, stirred for 2 h, and left for 48 h. For the preparation of the Al sacrificial layer, the evaporation method was used to deposit the Al sacrificial layer at a thickness of approximately 120 nm on the sapphire substrate. Then, the sapphire substrates were annealed at 500 °C for 1 h, and a mixing gas of 90% Ar + 10% H2 was introduced during the annealing process.

The processes performed to synthesize ZnO nanoflower arrays consisted of the following steps: (1) A sapphire substrate with patterned concave structure, which was provided by Shun Haw Technology Ltd. (Taiwan), was used as a template to prepare the ZnO seed layer with a patterned convex structure, as the schematic diagram of Figure a shows. The cross-sectional observation of used 2 in. sapphire substrate was prepared using focus ion beam (FIB) for observation, and the image is depicted in Figure a. The sapphire substrate had a concave structure, and its average bottom width and average height were 0.37 and 0.48 μm. As Figure a shows, because the sapphire substrate, which was provided by Shun Haw Technology Ltd. in Taiwan, has the structure of the patterned concave arrays. Therefore, we used it as the substrate to prepare a patterned protrusion ZnO seed layer and then we could synthesize ZnO nanoflower arrays on patterned concave arrays. The sapphire substrate was cleaned using acetone, isopropyl alcohol, and deionized water in an ultrasonic cleaner for 10 min at each cleaning step, then they were dried using nitrogen flow. (2) The thermal evaporation was used to deposit a sacrificial Al layer on the sapphire substrate, and the thickness of Al was about 120 nm by controlling the deposition time, as the schematic diagram of Figure b shows. (3) A spin-coating process was used to deposit ZnO gel and fully cover the surface of patterned sapphire template, 2000 rpm, and 30 s were used as spin speed and time, as the schematic diagram of Figure c shows. After the ZnO gel was coated, the template was baked at 300 °C for 10 min, and the baked process was used to remove and volatilize the organic solvent and to harden and dry the coating film. The spin-coating and baking processes were repeated six times, then the ZnO seed layer with a thickness of approximately 190 nm was obtained. (4) The deposited ZnO seed layer was annealed at 500 °C for 1 h in an Ar ambient, then we coated an optical grade silicone A-B glue on the ZnO seed layer and baked it at 100 °C for 1 h, as the schematic diagram of Figure d shows. (5) A mixed solution with K3Fe(CN)6/KOH/H2O = 10:1:100 (in weight) was used to etch the sacrificial Al electrode, and we found that the etching rate of the Al electrode was approximately 10 nm/s, as the schematic diagram of Figure e shows. (6) The multilayer ZnO seed layer/silicone A-B glue was lifted off and transferred to the silicon substrates, and the convex ZnO seed arrays were formed, as the schematic diagrams of Figure f,g show. A focus ion beam microscope (FIB) was used to observe the cross-sectional morphologies of prepared novel seed arrays for synthesizing ZnO nanoflower arrays, and Figure b and c represent their observed results with low magnification and high magnification, respectively. The thickness of the ZnO seed layer was 198 nm, which was close to the value of the seed layer deposited on the P-type Si ⟨100⟩ wafer. (7) Finally, the hydrothermal method was used to synthesize ZnO nanorods (ZnO nanoflower arrays) at 90 °C for 60 min with different concentrations of Zn(CH3COO)2 and C6H12N4 (0.01, 0.02, 0.03, and 0.04 M) in 200 mL of deionized water, as the schematic diagram of Figure h shows. When the flower-like (nanoflower arrays) ZnO nanomaterials were well synthesized, their crystal properties were analyzed using X-ray diffraction (XRD) with Cu Kα radiation (λ = 0.154056 nm), their morphological properties were observed using field-effect scanning electron microscopy (FESEM). The measurement system of the optical properties was Horiba Jobin iHR550 fluorescence spectrophotometer with a single-wavelength 325 nm He Cd laser as the excitation light source to excite the ZnO nanoflower arrays. Their optical properties were investigated using the iHR550 fluorescence spectrometer in the wavelength range of 350–650 nm and at room temperature.[23]

Figure 7

Schematic diagrams of the fabrication of ZnO nanoflower arrays. (a) Sappire substrate, (b) coating Al sacrificial layer, (c) coating ZnO-seed layer, (d) coating AB glue, (e) etching Al sacrificial layer, (f) get seed layer, (g) seed layer put on Si substrate, and (h) growth ZnO nanoflowers.

Figure 8

SEM images of the cross-sectional images for (a) patterned sapphire prepared using FIB and prepared novel templates with (b) low magnification and (c) high magnification.