Nanostructured manganese dioxides (MnO₂) have been considered as an ideal electrode material for energy storage, such as supercapacitors (also known as electrochemical capacitors (ECs)) [1-4], high-capacity lithium ion batteries [5], lithium-air batteries [6-8] for their advantages of low cost, earth abundance, environmental friendliness and superior performance in energy capacity. So far, numerous efforts have been devoted to synthesize MnO₂ nanostructures and a variety of strategies have been developed, including thermal decomposition, coprecipitation [9], simple reduction [10,11], solid-phase process, hydrothermal method [4], sol-gel [12], microwave process [13], etc. Among these methods, hydrothermal synthesis has attracted more attention because it is easily controlled on the shape of materials, which are simple processed and in large scale. For example, Li et al. [14] used hydrothermal route to obtain 3D urchinlike β-MnO₂ constructed of self-assembled nanorods; Qiu et al. [15] synthesized MnO₂ nanomaterials by hydrothermal treatment and investigated their catalytic and electrochemical properties. However, the phase and morphology of the MnO₂ nanostructrues are still not well controlled. Since the properties of electrochemical devices extremely rely on the crystalline phase and morphology of MnO₂ nanostructures [16], developing a simple route to synthesize various phases and shape for MnO₂ nanostructures is of fundamental importance. Herein, we demonstrate a onestep hydrothermal route to synthesize MnO₂ nanostructures with well controlling of their phases and morphologies, including α-MnO₂ nanorods, nanotubes, nanocubes, nanowires and β-MnO₂ nanosticks, which are simply tuned by changing the molar ratio of Mn precursor solution to HCl, Mn(Ac)₂∙4H₂O or C₆H₁₂O₆·H₂O, surfactants as well reaction temperature and time. We also propose the formation mechanism of MnO₂ nanostructures. These MnO₂ products would be potentially used in supercapasitor applications and other energy storage devices.

To study the role of the molar ratio of KMnO₄ to HCl, we made three different samples with the molar ratio of 1:8, 1:4 and 1:2, respectively. The reaction was carried out at the temperature of 140˚C for 12 h. Figure 1 shows the morphology of the as-prepared products. As it shows (Figures 1(a)-(c)), the products consist of nanorods with the length ranging from 1 to 3 μm. But when we take a closer look at Figure 1(b), as revealed in the picture inserted, these nanorods are hollow in the center with open ends, more like nanotubes. We found that the nanorods synthesized at the molar ratio of 1:8 were aggregated

to some extent with relatively small diameter (30 - 50 nm) and some of these nanorods were entangled to form stablike spheres with sharp tips, as shown in Figure 1(a). However, this phenomenon was not observed in the ones synthesized at the molar ratio of 1:4 or 1:2, as shown in Figures 1 (b) and (c), in which the diameters are wider ranging from 80 to 120 nm. It is likely that a larger amount of HCl (lower ratio of KMnO₄ to HCl) leads to the aggragation of nanorods. From the reaction process point of view, the reactions for the formation of MnO₂ use KMnO₄ and HCl according to the following reactions: [17].

It is obvious that more HCl would accelerate the reaction proceeding to the right, thus more MnO₂ nuclei would be produced within the given time, which is more likely to lead to an aggregation.

The powder X-ray diffraction (XRD; D/max-2550 PC) pattern was shown in Figure 1(d) for the sample synthesized with the molar ratio of KMnO₄ to HCl of 1:4. The peaks were shown up at the 2θ angle of 12.6˚, 17.9˚, 28.7˚, 36.5˚, 41.8˚, 49.7˚ and 60.2˚. According to the standard value (JCPDS: 44 - 0141), those as-prepared products can be indexed to a tetragonal α-MnO₂ and there is no characteristic peak from impurities. The sharp shape and narrow line widths of the diffraction peaks indicate that the MnO₂ material is highly crystallized. We also performed XRD measurement for another two samples and found out they are in the same crystalline structure.

In order to further explore other parameters that might make impacts on the morphology of the products, we studied the synthesis at different temperatures or time. Moreover, the role of surfactant was also examined. Figures 2(a) and (b) show the morphology of the as-prepared MnO₂ synthesied at the molar ratio of KMnO₄ to HCl of 1:2 at the temperature of 160˚C and 180˚C for 12 h. Similar to the MnO₂ synthesized at 140˚C, the proucts are made of nanorods with length ranging from 1 to 4 μm and diameter from 50 to 200 nm. Compared with the MnO₂ synthesized at 140˚C (Figure 1(b)), there are few fine particles on the surfce of MnO₂ nanorods and the higher the temperature is, the fewer the particles on the surface are, which were replaced by a few short nanorods, as shown in Figure 2(b). It is commonly known that nanostructrues start from forming nuclei and then these nuclei would grow up to ressemble into different nanostructures under different conditions. The formation of rods is favored over that of spherical-shaped nanocrystals under the high growth rate regime which usually results from high temperature [18]. This is why we observed that higher temperature leads to short nanorods forming on the surface but not the particles. To study the effect of time, we chose the sample synthsized at 180˚C for 12 h and elongate the reaction time to 18 h. As reveals in Figure 2(c), increasing time doesn’t lead to a big variation in the morphology of MnO₂ but the dispersity and uniformity are becoming better with the reaction time increasing. And the surface of the MnO₂ nanorods is more uniform and smoother, and no other impurities on the surface were observed. Additionally, the diameter of these nanorods increases to 50 nm but the length is the same as the ones obtained under lower temperature. Figure 2(d) reveals the morphology of the as synthesized MnO₂ by adding PVP as surfactant and the reaction was carried out at the molar ratio of KMnO₄ to HCl of 1:2 at 140˚C for 12 h. It is interesting that the MnO₂ nanorods were changed into shorter nanostructures in different shapes, more like cylinder-like and spindle-like nanosticks with the diameter around 1.2 μm. We noticed that the surface of these nanostructures was not as smooth

as the one made before but wrinkled.

XRD was examined to identify the structure for the product obtained by using PVP as surfactant. As Figure 3 shows, the peaks appear at the 2θ angle of 28.6˚, 37.3˚, 42.7˚, 56.6˚, 59.3˚ and 72.4˚. According to the standard value (JCPDS: 65 - 282), the as-prepared product can be indexed to a tetragonal β-MnO₂ and there are no other characteristic peaks from impurities. The possible reason for the β-MnO₂ formation is proposed as follows: PVP would be absorbed on the surface of MnO₂ nuclei at the beginning of the reaction, resulting in smaller possibility that K+ could take up the 2 × 2 tunnel site in α-MnO₂. Thus, K+ was not able to get into the tunnel to serve as the tunnel stabilizer, finally leading to the formation of small tunnel size β-MnO₂.

Since the molar ratio of KMnO₄ to HCl, the temperature and time doesn’t change the shape of the MnO₂ nanosturcture significantly, we used Mn(Ac)₂·4H₂O and C₆H₁₂O₆·H₂O replacing of HCl to explore the effect of precursors on the shape of MnO₂. In order to make a parallel comparison, all reactions were carried out at 180˚C for 24 h. As suggests in Figures 4(a) and (b), using HCl results in forming nanorods with length around 3 μm, wich is consistent with the previous results. But when use Mn(Ac)₂·4H₂O instead of HCl, long nanowires with the length longer than 5 μm and the diameter of 40 nm were formed, as revealed in Figures 4(c) and (d). Interestingly, when C₆H₁₂O₆·H₂O was used, the as-prepared samples were formed into hexahedron nanocubes with diameter around 2 µm which are uniform and well dispersed (Figures 4(e) and (f)).

Figure 5 reveals the XRD pattern of the assynthesized products (Figures 4(e) and (f)) prepared by using KMnO₄ and C₆H₁₂O₆·H₂O as the precursors. According to the standard value (JCPDS: 83-1763), the peaks shown in Figure 5 are consistent with MnCO₃ but not MnO₂ which we previously obtained. This is similar to the previous