ZnO is a wide‐bandgap (3.37 eV at room temperature) oxide semiconductor that is attractive for its great potential in short‐wavelength optoelectronic devices, in which high quality films and heterostructures are essential for high performance. In this study, controlled growth of ZnO‐based thin films and heterostructures by molecular beam epitaxy (MBE) is demonstrated on different substrates with emphasis on interface engineering. It is revealed that ultrathin AlN or MgO interfacial layers play a key role in establishing structural and chemical compatibility between ZnO and substrates. Furthermore, a quasi‐homo buffer is introduced prior to growth of a wurtzite MgZnO epilayer to suppress the phase segregation of rock‐salt MgO, achieving wide‐range bandgap tuning from 3.3 to 4.55 eV. Finally, a visible‐blind UV detector exploiting a double heterojunction of n‐ZnO/insulator‐MgO/p‐Si and a solar‐blind UV detector using MgZnO as an active layer are fabricated by using the growth techniques discussed here.

A phenomenon of wurtzite (w), zincblende (zb), and rock-salt (rs) phase separation was investigated in ZnCdO films having Cd contents in the range of 0%-60% settling a discussion on the phase stability issues in ZnCdO. First, low-Cd-content (17%) ZnCdO was realized preferentially in a w matrix determining optimal Zn-lean conditions by tuning the precursor decomposition rates during synthesis. However, more detailed analysis of x-ray diffraction and photoluminescence (PL) data revealed that the w single-phase stability range is likely to be as narrow as 0%-2% Cd, while samples containing 7%-17% of Cd exhibit a mixture of w and zb phases. Second, high-Cd-content (32%-60%) ZnCdO samples were realized, supplying more of the Cd precursor utilizing Zn-lean growth conditions, however, resulting in a mixture of w, zb, and rs phases. Characteristic PL signatures at 2.54 and 2.31 eV were attributed to zb-ZnCdO and rs-CdO, respectively, while the band gap variation in w-Zn 1−x Cd x O is given by (3.36-0.063x) as determined at 10 K. The phase separation is interpreted in terms of corresponding changes in the charge distribution and reduced stacking fault energy.

No abstract