Solar cells commercial success is based on an efficiency/cost calculation. Nanowire solar cells is one of the foremost candidates to implement third generation photo voltaics, which are both very efficient and cheap to produce. This thesis is about our progress towards commercial nanowire solar cells.

Resonance effects between the light and nanowire causes an inherent concentration of the sunlight into the nanowires, and means that a sparse array of nanowires (less than 5% of the area) can absorb all the incoming light. The resonance effects, as well as a graded index of refraction, also traps the light. The concentration and light trapping means that single junction nanowire solar cells have a higher theoretical maximum efficiency than equivalent planar solar cells. We have demonstrated the built-in light concentration of nanowires, by growing, contacting and characterizing a solar cell consisting of a single, vertical, gallium arsenide(GaAs) nanowire grown on silicon with a radial p-i-n-junction. The average concentration was ~8, and the peak concentration was ~12.

By increasing the number of junctions in solar cells, they can extract more energy per absorbed photon. In ideal multi junction (MJ) solar cells each junction absorb the same number of photons. Current MJ solar cells efficiency is hampered by the fact combining the most complimentary materials, from an absorption standpoint, is impossible due to mismatches in the crystal structure. Nanowires solve this problem, since the small footprint of grown nanowires relaxes the crystal matching constraint. 1.7eV is the ideal bandgap for a top junction in a dual junction solar cell, where silicon is the bottom junction. This can be obtained with GaAs0.8P0.2. We have demonstrated how to incorporate phosphorous(P) into Ga-catalyzed nanowire growth, and grown GaAs1−xPx nanowires with different inclusions of P(x) directly on silicon. The incorporation of P was generally higher in nanowires than for planar growth at identical P flux percentage. More interestingly, the percentage of P in the nanowire was found to be a concave function of the percentage of P in the flux, while for planar growth it was a convex function. We have demonstrated GaAs0.8P0.2 nanowires and further grown a shell surrounding the core with the same composition.

The lattice matched GaAsP core-shell nanowire were doped to produce radial p-i-n junctions in each of the nanowires, some of which were removed from their growth substrate and turned into single nanowire solar cells (SNWSC). The best device showed a conversion efficiency of 6.8% under 1.5AMG 1-sun illumination. In order to improve the efficiency a surface passivating shell consisting of highly doped, wide bandgap indium gallium phosphide was grown. The best SNWSC from this batch had an efficiency of 10.2%.