Figure 1(a) shows the typical XRD pattern of the as-synthesized nanocrystals (sample 1). It could be observed that all diffraction peaks can be indexed to the anorthic-phase Sn₃O₄ (JCPDS No. 16-0737). No additional impurity diffraction peaks are observed, which indicates that the single-phase Sn₃O₄ is obtained in the present work. As shown in the higher-magnification SEM and TEM images in Fig. 1(b), the Sn₃O₄ shows a sheet-like shape with a thickness of about 10 nm∼ 20 nm. The selected-area electron diffraction (SAED) pattern shows a single-crystal-like diffraction corresponding to the [010] zone axis. It is shown in Fig. 1(c) that the plate has a (101) interplanar distance of 0.372 nm and a (002) interplanar distance of 0.242 nm. The XPS measurements demonstrate that the area ratio of S(Sn²⁺ )/[S(Sn⁴⁺ ) + S(Sn²⁺ )] is 0.66 deduced from the Sn 3d spectra (Fig. 1(d)). This is in good agreement with the theoretical value of 0.67 in Sn₃O₄. The Brunauer– Emmett– Teller (BET) surface area of the single-crystal-like nanosheet is further investigated by N₂ adsorption– desorption measurements in Fig. 2. The representative nitrogen isotherm shows a uniform pore size distribution with an average pore diameter of 2.3 nm and a high surface area of 67 m²/g. The regions of low contrast between individual crystallites (Fig. 1(c)) indicate mesopores with diameters of about 1.2 nm∼ 2.7 nm, ^{[18, 19]} which is in good agreement with the pore size distribution results. Considering all the above results, it can be concluded that nanoporous anorthic-phase Sn₃O₄ nanosheets have been successfully fabricated by the hydrothermal process without any additives.

Fig. 1.

	Figure Option ViewDownloadNew Window
	Fig. 1. XRD pattern (a), SEM (b), HRTEM image (c), and XPS (d) of nanoporous Sn3O4 nanosheets. The upper left inset in panel (b) shows the TEM image, the upper right inset is the SEM image, and the lower right inset is the SAED pattern.

Fig. 2.

	Figure Option ViewDownloadNew Window
	Fig. 2. N2 adsorption– desorption isotherms of nanoporous anorthic Sn3O4 mesocrystals. The solid and empty symbols represent adsorption and desorption data, respectively. The inset shows the pore size distribution obtained from the adsorption branch.

It should be noted that SnO mesocrystals are synthesized in the same fabrication progress when the pH is adjusted to 11.8 by sodium hydroxide (sample 2).^[20] For investigating the formation mechanism of the Sn₃O₄, the precursors of different pH values during the reaction are monitored. Powder XRD patterns in Fig. 3 indicate that the precursor prepared with a pH value of 11.8 is consistent with the standard Sn₆O₄(OH)₄, and the precursor prepared with a pH value of 2.0 is similar to Sn₆O₄(OH)₄ with peaks shifting toward higher angles. According to the XPS measurements (Fig. 4), chemical states of the tin elements of the precursor prepared in an alkaline condition (sample 2) are all bivalent while those prepared in an acidic condition (sample 1) are mixed valence with a ratio of Sn²⁺ to Sn⁴⁺ being 2.7:1.

Fig. 3.

	Figure Option ViewDownloadNew Window
	Fig. 3. Typical XRD pattern of the precursor of samples 1 and 2.

Fig. 4.

	Figure Option ViewDownloadNew Window
	Fig. 4. The Sn 3d and Sb 3d XPS spectra of the precursors for samples 1, 2, and 3.

To clearly identify the synthesis mechanism of Sn₃O₄ nanosheets, a comparative experiment is carried out. A controlled-valence progress is introduced to regulate the valence of the precursor to synthesize Sn₃O₄ nanosheets in an alkaline condition. Verwey et al. have successfully changed the valence of Ti⁴⁺ or Mn⁴⁺ in CaTiO₃ or CaMnO₃ respectively by La dopant.^[21] In our comparative experiment, antimony (6 at.%) is considered to be doped into the precursor and the pH value is adjusted to 11.8 by sodium hydroxide (sample 3). The chemical states of the two principal elements (Sn and Sb) in sample 3 are characterized by XPS. In the precursor of sample 3, the ratio of Sn²⁺ to Sn⁴⁺ is 2.6:1 and the ratio of Sb³⁺ to Sb⁵⁺ is 1.6:1 (Fig. 4). Figure 5 shows the XRD pattern of sample 3. As-synthesized particles exhibit a perfectly anorthic-phase Sn₃O₄ structure (JCPDS No. 16-0737). According to the data mentioned above, the valence state of Sn species in the precursor might be the key factor of the formation of Sn₃O₄ nanosheets.

Fig. 5.

	Figure Option ViewDownloadNew Window
	Fig. 5. Typical XRD pattern of sample 3 with 6 at.% Sb doping and 11.8-pH value.

The hydrolysis process of sample 3 is monitored by UV-vis. The reaction is stable for about 20 min (Fig. 6(a)) and the satellite peak below the absorption edge evolves from one to two near 225 nm and 228 nm, which correspond to the formation of Sn⁴⁺ species.^[22] However, for sample 2 (Fig. 6(b)), the hydrolysis process of SnCl₂ lasts for 5 min with no Sn⁴⁺ spices existing in the precursor. It indicates that the SbCl₃ not only slows the hydrolysis rate of SnCl₂ but also facilitates the oxidization of Sn(2). The reaction scheme of the dissolution of SnCl₂ in an alkaline condition is as follows:

With the addition of SbCl₃ in sample 3, the rate of hydrolysis of SnCl₂ is slowed down according to the following reaction:^{[23, 24]}

The system becomes unstable and the Sn(2) and Sb(3) ions are slowly oxidized to Sn(4) and Sb(5) in air under magnetic stirring. After the precursor transforms into a Teflon-lined autoclave, the Sn(2) and Sb(3) are continuously oxidized by oxygen molecules in the hydrothermal environment. The oxygen molecules come from the air (30 mL) in the Teflon container and are dissolved in the precursor solution ([O₂] ≈ 1.4 mM).^[25] However, the oxygen molecules in the system are insufficient for totally oxidizing the Sn(2) species, with an upper limit yield of 35.8 mol%. Sn₃O₄ has the most stable structure in SnO_{2− x} (x = 0.73– 0.60) and has a formation energy of − 9 meV/Sn.^[14] Thus, Sn₃O₄ is formed finally in an alkaline condition.

Fig. 6.

	Figure Option ViewDownloadNew Window
	Fig. 6. UV-vis spectra of the hydrolysis process of samples 3 (a) and 2 (b) with different reaction times.