Semiconductors materials and devices are essential building blocks for many of the technologies deeply embedded in modern life. Improving the performance of semiconductor devices requires a deeper understanding of the fundamental mechanisms controlling the crystalline structure of the semiconductor materials. This thesis focuses on two compound semiconductor systems in the form of 𝑍𝑛3𝑃2 thin films and 𝐺𝑎𝐎𝑠 nanowires and tries to establish methods for minimizing the defects in these structures by the careful control of order at the atomic scale. In the first part of this work, we have focused on Zn3 P2 thin films. The combination of suitable material properties and the abundancy of its constituting elements makes this platform a promising candidate for large-scale and scalable photovoltaic applications. Despite this, Zn3 P2 is not adopted by the solar industry due to the many technical challenges in the growth of this material. In this work, by careful control of the state of the interface between the thin films and the substrates, we have provided a method for the growth of thick and monocrystalline Zn3 P2 thin films with superior quality. This is crucial for the realization of any successful photovoltaic cell based on this material. Although molecular beam epitaxy is used in this study, the process provided is generalizable and could also be exploited for the other growth methods. A host of different characterization techniques, including electron microscopy and spectroscopy, Raman spectroscopy, photoluminescence spectroscopy and x-ray diffraction are used to assess the different aspects of the thin films. It is observed that monocrystalline films have better optical properties compared to polycrystalline thin films, making them more suitable for photovoltaic applications. In the second part of the thesis, we have focused on improving the crystalline quality of 𝐺𝑎𝐎𝑠 nanowires. Similar to other III-V semiconductors, 𝐺𝑎𝐎𝑠 hosts an internal electrical dipole known as polarity. Arsenide and phosphide nanowires often grow along (111)𝐵 direction, which implies the termination of (111) bilayer by group V elements. However, this configuration is often defective and results in polytypism. On the other hand, (111)𝐎 growth, which refers to group III termination, results in a crystalline structure that is largely free of defects. However, growth in this polarity is rarely reported. In this work, we try to understand the reason for this elusiveness and the superior crystalline quality associated with A-polar growth. To do that, we have established a new framework that explicitly focuses on the atomic structure and order of liquid-solid interface in nanowires grown by vapor-liquid-solid method. The atomic structure of this interface has been ignored up to now in the literature due to the technical challenges for investigating such fragile atomic order. Here, we employ a combination of experimental observations via electron microscopy and spectroscopy and machine-learning-based molecular dynamics simulations that were developed by external collaborators. The results of this study provide a unique insight into the fundamental aspects of nanowire growth. Key words: compound semiconductors, photovoltaic cells, molecular beam epitaxy, thin films, nanowires, growth polarity, liquid-solid interface, liquid ordering