Self-assembly of ZnO nanodots on glass substrates by a facile sol-gel method

. In this work, we report the self-assembly of ZnO nanodots on the glass substrates by a sol-gel method. Firstly, the precursor solution consisted of zinc acetate and the solvent of ethylene glycol or propylene glycol is employed to fabricate precursor thin film by spin coating. Secondly, the as-grown precursor thin film is annealed at a low temperature to form the self-assembled ZnO nanodots. The effects of solvent type, precursor solution concentration and annealing time on the morphologies of the self-assembled ZnO nanodots were investigated. In addition, the self-assembly evolution of the ZnO nanodots is proposed in order to helping people to understand the formation of the nanodots by mean of sol-gel method.


Introduction
In recent years, ZnO nandots arrays have been widely utilized as high-performance electrodes for lithium-ion batteries [1], fluorescent paper sensor [2], gas sensor [3,4], humidity sensor [5], quantum reinforcing material [6], template for synthesis of nanorods/nanowires [7] and so on [8][9][10][11], owing to their excellent physical and chemical properties [8,9,12,13]. In order to achieving and applying the ZnO nanodots, many methods have been created to synthesize the ZnO nanodots [3,14,15]. Gruzintsev et al. reported that two-dimensional ordered array of the ZnO quantum nanodots can be realized by means of the synthetic opal with a diameter of 279 nm [16]. It has been demonstrated by Choi et al. that highlyordered ZnO nanodot arrays can be obtained by electrosynthesis with colloidal polystyrene spheres and electrodeposited polypyrrole templates (double template) [17]. Metal-organic chemical vapor deposition (MOCVD) also can be used to selectively grow ZnO nanodots on focused-ion beam nanopatterned substrates [18]. Interestingly, by mean of the step edges on R-face sapphire, ZnO nanodot arrays can be formed by MOCVD [19]. Similarly like MOCVD, pulsed laser deposition can be deployed to prepare ZnO naondots, too [20]. Amazingly, it was reported by Kaupužs et al. that Zn and ZnO nanoparticles can be formed by laser radiation on ZnO single crystals [21]. Erol demonstrated that ZnO nanoparticles synthesized by sol-gel process reveal outstanding humidity sensing properties [5]. Similarly like sol-gel process, Riedel et al. employ spray methods to fabricate ZnO nanodots [7], which are subsequently applied for synthesizing ZnO nanorods array. Nowadays, environmental protection is getting more and more people's attention. As a result, green synthesis of ZnO nanodots is necessary. It has been demonstrated by Jacob et al. that stable, ultra-small ZnO quantum dots (nanodots) can be achieved by a facile, green, solution approach, which have radius less than 2 nm [6]. Based on the concept of green synthesis, it is superior to prepare ZnO nanodots by sol-gel method with environmentally friendly materials. Here, we focus on the ZnO nanodots for synthesizing ZnO nanorods. Uniform size and uniform distribution ZnO nanodots are considered as the best template for growing ZnO naonrods. According previous studies, sol-gel method would be a better approach for preparing ZnO nanodots due to its mild reaction system and low growth temperature that is usually lower than 100 oC [5,10,12,22]. It is believed that both solvent and annealing temperature play an important role in sol-gel method [5,10]. However, systematic research on this aspect is still relatively rare. Furthermore, the formation evolution of ZnO nanodots is often neglected. In this work, we report the effect of solvent type, solution concentration and annealing temperature on the morphological properties of ZnO nanodots. Furthermore, the formation evolution of the ZnO nanodots is also proposed.

Experimental
ZnO nanodots were prepared on glass substrates by a solgel method. Firstly, zinc acetate (Zn(Ac)2, purity of 99.5%) put into the solvent (ethylene glycol or propylene glycol) to prepare 0.15 mol/L mixture precursor solution by mean of magnetic stirring for 60 min at room temperature. Subsequently, 0.75 mL precursor solution was put on the middle position of the clean glass substrate, which was set in the spin coater. Afterwards, the glass substrates were rotated for 30 s at 500 rpm and then rotated for 120 s at 1500 rpm to form the precursor thin films. The as-grown precursor thin films were transferred into the annealing furnace, following an annealing process at 70 oC for 3.0 h. During the annealing process, ZnO nanodots were gradually formed. In the second part, precursor solutions with solution concentrations of 0.01, 0.015, 0.02 and 0.025 mol/L were prepared respectively. And then ZnO nanodots were grown in the same process but annealing at 90 oC for 30 min.
In the third parts, similarly, precursor solutions with solution concentration of 0.01 mol/L were utilized to synthesize ZnO nanodots by annealing at 90 oC for 0.5, 1.0, 2.0 and 3.0 h, respectively. The morphology of ZnO nanodots samples was measured by mean of scanning electron microscope (SEM, ZEISS Sigma 500). Fig. 1(a)-(d) represent the SEM images of ZnO nanodots self-aggregating on glass substrates, and the solvent is varied, and the annealing treatment of 70 °C for 3 h remains unchanged. As expected, when ethylene glycol is chose as the solvent, ZnO nanodots are self-aggregated on the glass substrates, which are confirmed by Fig. 1(a). ZnO nanodots, the white points in Fig. 1(a), can be seen clearly. Further study by the high magnification SEM image shows that ZnO nanodots with a half-spheric morphology are almost fully covered on the glass substrate, the size of ZnO nanodots is not uniform and most of them land in the range from 7.7 to 18.5 nm, as shown in Fig. 1(b). Furthermore, there is a significant height difference between the large size ZnO nanodots and the small size ones. On the contrary, when it turns to the solvent of propylene glycol, uniform sized ZnO nanodots can be observed in Fig. 1(c). High magnification SEM image further confirms that ZnO nanodots are more uniform distribution in the size with a diameter of 12.8-21.5 nm (Fig. 1(d)), and no significant height difference can be found in Fig. 1(d). Distinctly, there is a large or small space between the ZnO nanodots. The different result caused by the solvent type may be due to the difference in the volatility and molecular weight between thylene glycol and propylene glycol. Compared with ethylene glycol, propylene glycol has larger viscosity and molecular weight value, which makes it suitable for creating space between ZnO nanodots, possibly resulting in uniform size of the ZnO nanodots. Therefore, the ZnO nanodots made from the solvent of propylene glycol are better match with the demands for the subsequent growth of ZnO nanorods/nanowires [7]. formation process is described as above, but the annealing time is reduced to 0.5 h. Fig. 2(a) displays the SEM image of ZnO nanodots that made from the 0.01 mol/L precursor solution. The coated precursor film initiates its selfaggregation to form a long strip during the annealing process. Due to uneven film thickness precursor', quite a lot of large size nanosheets also can be seen in Fig. 2(a). In contrast, with the increase of the precursor solution concentration, it is found that the morphology of the selfassembled ZnO nanodots on the glass substrates is significantly different. By increasing the precursor solution concentration from 0.01 to 0.015 mol/L, the coherent aggregation of ZnO nanodots with flat tops can be observed in Fig. 2(b). When further increasing the precursor solution concentration to 0.02 mol/L, the size of ZnO nanodots with a half-spheric morphology becomes apparently smaller than those made from the small precursor solution concentration, whereas the density of ZnO nanodots greatly increase, as shown in Fig. 2(c). If the precursor solution concentration is increased to the value of 0.025 mol/L, the glass substrate is covered by ZnO thin film (nanofilm) and some ZnO nanodots (white points) are found on the ZnO thin film' surface ( Fig. 2(d)). Based on above results, it can be concluded that the suitable precursor solution concentration range for selfassembling ZnO nanodots with the solvent of propylene glycol is from 0.01 to 0.02 mol/L, which is mainly due to the increased number of crystallized core during annealing process. It is easily believed that the number of crystallized core is increased with the precursor solution concentration. Once it is over than a certain value, taking the case of 0.025 mol/L precursor solution as an example, the ZnO nanodots will become into ZnO nanofilm. In addition, compared with the annealing time of 3 h, it is also found that shortening annealing time can induce ZnO nanodots too. Therefore, it is necessary to optimize the annealing time for self-assembling ZnO nanodots with a better morphology. This will be studied in the next part. In fact, the size and distribution of the self-aggregated ZnO nanodots on the glass substrates are mainly depended on the precursor solution concentration, the annealing process, and the thickness of the precursor thin film. It has been demonstrated that annealing temperature plays a critical role in the size and uniformity of the selfaggregated nanodots. In details, higher annealing temperature is able to result in a reduced uniformity in the size distribution of the nanodots. On the contrary, lower annealing temperature certainly takes a longer time for self-assembling nanodots. However, lower annealing temperature means more safety and higher controllability in the morphology of nondots. For the self-aggregated ZnO nanodots, 90 oC is considered as a superior value of annealing temperature based on our previous experiments. By comparing Fig. 1(d) and Fig. 2(b), it can be concluded that the morphology of the self-aggregated ZnO nanodots is significantly affected by the annealing time. In order to finding the suitable annealing time and studying formation evolution for self-assembling ZnO nanodots,

Results and discussion
annealing time values of 0.5, 1, 2 and 3 h are set to synthesize ZnO nanodots with a precursor solution concentration of 0.01 mol/L, as shown in Fig. 3(a)-(d). When the annealing time is shorter than 0.5 h, precursor thin film is still retained and no ZnO nanodots can be found. By increasing the annealing time to 0.5 h, the numbers of the self-assembled ZnO nanodots are apparently enhanced, and the nanostrips are still retained as shown in Fig. 4(a). It should be noted that these nanostrips are made of the nanodots, which means that the nanostrips would eventually break into the self-assembled ZnO by increasing the annealing time. The nanostrips disappear completely and only self-assembled ZnO nanodots have been obtained on the glass substrates when using an annealing time of 1 h (Fig. 2(b)). The average size and density of the self-assembled ZnO nanodots are 15 nm and about 6.6×1010 cm-2, respectively, as illustrated in Fig. 3(b). These small ZnO nanodots would gradually merge with large one, which lead to a reduction of density accordingly. For 2 h annealing, a few selfassembled ZnO nanodots become into large square or triangular prism nanodots from the half-spheric nanodots while others change into smaller ones (Fig. 3(c)). The largest square ZnO nanodot has a size of ~141 nm. In addition, the density of the self-assembled ZnO nanodots is decreased obviously. If the annealing time is further lengthened to 3 h, the size of the large sized ZnO nanodots will keep enlarging, while small nanodot will get smaller and smaller until it disappears completely, and accordingly, the density of the self-assembled ZnO nandots will keep reducing, as shown in Fig. 3(d). The largest sized of ZnO nanodot is further increased to 306 nm that is about 2 time of the value annealed for 2 h. In this work, our aim is to synthesize ZnO nanodots for growing ZnO nanorods, especially core-shell ZnO-based nanorods. Therefore, annealing time of 1 h would be an appropriate value for synthesizing ZnO nanodots that matches our demands. Interestingly, after 1 h annealing, if the annealing time is further lengthened, a few ZnO nanodots become much larger while other become smaller than those annealed for 1 h. From what has been mentioned above, the self-assembly evolution of ZnO nanodots can be summarized as shown in Fig. 4. Firstly, the precursor solution is made into precursor thin film by mean of spin coating. Secondly, in the initial state of annealing at 90 oC, zinc acetate may hydrolyze accompanying the evaporation of alcohol. Taking ethanol solvent as an example, this reaction would be summarized as equation (1). Indeed, the Zn(OH) will be broke into ZnO and H2O, which could be displayed as equation (2).
Zn(AC)2 + (CH2OH)2 → Zn(OH)↓+ CH2(AC)-CH2(OH)↑ (1) Zn(OH)2 → ZnO (s) + H2O (g) (2) When the equation (1) mainly occurs, it is possible to observe the nanostrips on the glass substrates, as shown in Fig. 2(a) and Fig. 3. Thirdly, as the annealing time is lengthened, the equation (2) turns up and results in the formation of self-assembled ZnO nanodots, and gradually it will be able to become the main reaction that produce more and more ZnO nanodots, as shown in Fig. 1(c

Conclusion
By mean of sol-gel method, ZnO nanodots have been selfassembled on the glass substrates at a low annealing temperature. It is found that propylene glycol solvent is beneficial to forming uniform sized self-assembled ZnO nanodots with distinct space between nanodots when compared with ethanol solvent. The precursor solution concentration plays an important role in the formation of the assembled ZnO nanodots. 0.01-0.015 mol/L would be a superior value for self-aggregating ZnO nanodots that are good templates for synthesizing ZnO nanorods in a short annealing time. Furthermore, annealing time has demonstrated to be a key factor that affects the morphology, size and uniformity of the self-assembled ZnO nanodots. To match the aim for synthesizing coreshell ZnO-based nanorods, 1 h may be considered as a suitable annealing time value. In addition, the selfassembly evolution of the ZnO nanodots is proposed.