Deterministic growth of tungsten disulfide via liquid precursor chemical vapor deposition

Transition metal dichalcogenides (TMDs) are a family of materials with different properties ranging from insulators, to semiconductors, to metals. Their chemical composition is in the form \ce{MX2} where \ce{M} is a transition metal (group IV, V, VI, VII, VIII, IX or X) and \ce{X} is a chalcogen atom (such as S, Se or Te). Some of them are found in a layered structure, consisting of atomic layers held together by weak van der Waals interactions (vdW) forming bulk materials. In TMDs each layer is composed by three sublayers of atoms: one transition metal sublayer (M) is sandwiched between two sublayers of chalcogen atoms (X). In the present work we focus our attention on one of the semiconducting TMDs in which M is from group VI transition metal (M = Mo or W), and more precisely on tungsten disulfide (\ce{WS2}). Semiconducting TMDs are of special interest since the position of the valence band edges changes with the number of layers: the indirect band gap in semiconductor bulk material changes into a direct band gap in semiconductor monolayer. The band gap of bulk TMD materials is in the range of near infrared, while if reducing the number of layers down to monolayer the band gap shifts to the range of the visible light, making these materials interesting for optoelectronic applications. For bulk \ce{WS2} the indirect band gap is 1.35 eV while the direct gap of monolayer WS$_{2}$ is 2.1 eV. Strong photoluminescence, giant spin-orbit and spin-valley coupling, excellent preservation of polarization at room temperature, giant electro-refractive modulation of light, make \ce{WS2} a promising material for next generation applications such as spintronics, valleytronics, flexible electronics and optical photonics. \newline To date, different methods to obtain monolayer crystals have been explored, chemical vapor deposition (CVD) is the growth technique that has yielded the most promising results for the large scale synthesis of 2D TMDs, allowing one to finely control grain size and to obtain uniform thickness. In CVD, the deposition of a solid material occurs by adopting gaseous reactants which chemically interact near or on a heated substrate surface. However, a systematic growth technique to obtain such material with high-crystallinity and over large scale is still missing. Indeed when trying to scale-up in size \ce{WS2} only polycrystalline material is obtained, where grain boundaries and defects negatively affect the electronic and optical properties of the material. Also, such wafer-scale growth of polycrystalline thin layers of \ce{WS2} complicates the transfer procedure due to the difficulty to transfer over large areas 2D materials while preserving their integrity and avoiding wrinkling. Here, we develop for the first time the deterministic growth of \ce{WS2} using liquid precursor CVD (LqP-CVD) on \ce{SiO2}/\ce{Si} substrate. The implemented method consists of several subsequent preparation steps: namely, liquid precursor deposition, aggregation and thermal decomposition. Growth recipe is finally performed placing a cleaned \ce{SiO2}/\ce{Si} substrate face to face to the growth substrate in order to reduce the vertical growth and the nucleation in undesired zones. The as-grown flakes quality and thickness are characterized by using optical microscopy, Raman spectroscopy, photoluminescence (PL) measurements and atomic force microscopy (AFM). Our results demonstrate for the first time that seeded growth of \ce{WS2} crystals with lateral size of tens of micrometers can be achieved by LqP-CVD. These results are an important first step towards the wafer-scale growth of high-quality monolayer \ce{WS2}. In principle, the present growth method allows one to have high-quality single-crystals exactly where they are needed, in accordance with the final design of the electronic, photonic or spintronic devices. Also, single-crystals facilitate the transfer process so that the quality of the as-grown material is preserved. \newline We believe that further optimization of such growth approach could lead to consistently obtain crystals of monolayer thickness in all the matrix and not only in some specific locations. Possible improvements that shall be pursued in the future to augment the thickness homogeneity in the matrix are: (i) the optimization of the deposition process in such a way that the right quantity of precursor is deposited lowering its consumption during the growth ramp; (ii) the definition of the optimal quantity of \ce{NaOH} so that horizontal growth is more favourable than vertical growth