The layers grow not WS2, but WO2. They become WS2 after subsequent sulfurization. There is no oxygen inside the liquid, that's why crystals WO2 do not grow there.

The layers grow not WS2, but WO2. They become WS2 after subsequent sulfurization. There is no oxygen inside the liquid, that's why crystals WO2 do not grow there.

Dear Alexander N Titov ,

Thanks for your comment, According to the article Precursor decomposes into WS3 and WS3 further reacts with water to form WO3. So the oxygen is coming from water. WO3 flakes should form everywhere inside the liquid precursor. But WO3 flakes forms only at the liquid surface. That was actually my question.

Obviously, the oxygen activity is higher at the border with air than inside the solution. Hence, there is a gradient of oxygen concentration and a gradient of its chemical potential. According to the Gibbs-Duhem rule, for binary tungsten trioxide, it is equal to the gradient of the chemical potential of tungsten with the opposite sign. Therefore, there must be diffusion towards the boundary with air. And there the chemical potential of oxygen and tungsten is already sufficient to form tungsten trioxide.