Zinc phosphide (Zn3P2) nanowires have recently been demonstrated as a promising earth-abundant photovoltaic absorber due to their optoelectronic properties (Eg=1.50 eV and optical absorptance comparable to gallium arsenide).[1,2] However, the epitaxial growth has been heavily reliant on indium in both the vapour-liquid-solid growth catalyst and the substrate (indium phosphide), taking away from the earth-abundant nature of the absorber material, in addition to low-throughput growth processes (molecular beam epitaxy).[3] To further advance this material and increase its applicability it is thus desirable to replace the scarce element from the process and utilise more scalable growth techniques.
In this work, we use metalorganic vapour phase epitaxy (MOVPE) to demonstrate the transferability of the epitaxial growth of zinc phosphide nanowires and characterise the structural properties using electron microscopy. Furthermore, we can deposit the catalyst material as a separate step in-situ, rather than relying on substrate decomposition, which allows for increased control over the interface condition and epitaxial relation and the use of alternative catalyst materials. One such alternative is tin, a more earth-abundant alternative to indium that has been demonstrated for non-epitaxial growth of zinc phosphide nanowires.[4] As shown in Figure 1, tin-catalysed epitaxial zinc phosphide nanowires show additional morphologies to those observed previously, which is also influenced by the growth surface (e.g. (100), (110) ir (111)). However, using tin also has an interesting effect on the defect formation in the nanowires. The zigzag nanowires, as shown in Figure 1, have been shown to rely on indium from the droplet in order to form a heterotwin to create the oscillating structure.[5] In this case, a single monolayer of tin phosphide will be formed at each heterotwin. Due to its much smaller bandgap compared to the previously observed indium phosphide inclusion, there is a possibility the tin phosphide may affect the optoelectronic properties of the nanowires.
The last step in removing indium from the process relates to the substrate. This substrate was initially used as the close overlap of the phosphorus sub-lattices is expected to facilitate epitaxial growth. However, as is well known, nanowires are more flexible in allowing for greater lattice mismatch in defect free epitaxial growth. As such, we are exploring the growth of zinc phosphide nanowires on silicon (100) and (111) substrates. This step completely removes scarce elements from the zinc phosphide nanowire growth process, making it exclusively reliant on earth-abundant materials.

References
[1] G. M. Kimball et al. Appl. Phys. Lett., 2009, 95, 112103.
[2] M. Y. Swinkels et al. Phys. Rev. Applied, 2020, 14, 024045.
[3] S. Escobar Steinvall et al. Nanoscale Horiz., 2020, 5, 274-282.
[4] S. B. Choi et al. J. Phys. Chem. C, 2019, 123, 4597-4604.
[5] S. Escobar Steinvall et al. Nanoscale, 2020, 12, 22534-22540.