GaN and related III-nitrides have attracted considerable attention as promising materials for application in optoelectronic devices, in particular, light-emitting diodes (LEDs). At present, sapphire is still the most popular commercial substrate for epitaxial growth of GaN-based LEDs. However, due to its relatively large lattice mismatch with GaN and low thermal conductivity, sapphire is not the most ideal substrate for GaN-based LEDs. Therefore, in order to obtain high-performance and high-power LEDs with relatively low cost, unconventional substrates, which are of low lattice mismatch with GaN, high thermal conductivity and low cost, have been tried as substitutes for sapphire. As a matter of fact, it is not easy to obtain high-quality III-nitride films on those substrates for various reasons. However, by developing a variety of techniques, distincts progress has been made during the past decade, with high-performance LEDs being successfully achieved on these unconventional substrates. This review focuses on state-of-the-art high-performance GaN-based LED materials and devices on unconventional substrates. The issues involved in the growth of GaN-based LED structures on each type of unconventional substrate are outlined, and the fundamental physics behind these issues is detailed. The corresponding solutions for III-nitride growth, defect control, and chip processing for each type of unconventional substrate are discussed in depth, together with a brief introduction to some newly developed techniques in order to realize LED structures on unconventional substrates. This is very useful for understanding the progress in this field of physics. In this review, we also speculate on the prospects for LEDs on unconventional substrates.

Highly active and stable Pt-free electrocatalysts for hydrogen production via water splitting are of great demand for future energy systems. Herein, we report a novel hydrogen evolution reaction (HER) catalyst consisting of rhodium phosphide (Rh2P) nanoparticles as the core and N-doped carbon (NC) as the shell (Rh2P@NC). In a wide pH range, our catalyst not only possesses a small overpotential at 10 mA cm−2 (∼9 mV in 0.5 M H2SO4, ∼46 mV in 1.0 M PBS and ∼10 mV in 1.0 M KOH), but also demonstrates high stability. Importantly, all these performances are far superior to commercial Pt/C catalysts for HER. To the best of our knowledge, this is the highest HER performance reported so far in acidic and basic media. Density functional theory (DFT) calculations reveal that the introduction of phosphorus can significantly lower the proton adsorption energy of Rh/NC, thereby benefiting surface hydrogen generation. Moreover, this synthetic strategy for Rh2P@NC is also applied to other transition metal phosphides (TMPs)/nitrogen-doped carbon heterostructures (such as Ru2P@NC, Fe2P@NC, WP@NC etc.) with advanced performance toward HER and beyond.