Processing of thin film photovoltaics from chalcogenide nanoparticles

Kevin Wayne Brew, Purdue University


Over the last few decades, it has become evident that current energy production for humanity since the industrial revolution has incurred the emission of greenhouse gases (GHGs) into the Earth’s atmosphere, resulting in rampant pollution, global warming, ocean acidification and other disastrous environmental effects. The continued emission GHGs is a direct result of the predominant use of fossil fuels to meet an exponentially increasing global energy demand. Development of sustainable energy technologies is a global imperative to avoid future catastrophe. Photovoltaics (PV) are an ideal resource that allows us to convert our greatest supply of energy, sunlight, directly into our greatest source of energy consumption, electricity.^ In the last four decades, PV research for new solar materials and fabrication methods to compete with crystalline silicon (c-Si) modules has expanded in an effort to reach $1/Watt solar energy. Thin films of chalcogenide semiconductors such as CdTe, Cu(In,Ga)(S,Se)2 and Cu2ZnSn(S,Se)4 (CZTSSe) have ideal band gaps for solar absorption, require 100 times less material than c-Si, and do not require high levels of purity, thus lowering material costs and processing. The lower material cost from earth-abundant elements and ability to solution process CZTSSe make this material an ideal competitor with current PV technologies. Solution based methods, such as roll to roll printing of nanoparticle inks, are a more scalable method to deposit the absorber film. Once coated, only the selenization sinters sulfide nanoparticles into micron sized grains, which are required for high efficiency PV devices. Selenization and sulfurization equipment were engineered to study the chalcogenization of CZTS and CZTSe nanoparticle films. Due to the volatility of sulfur, liquid assisted sintering does not occur. However, abnormal grain growth above 550°C may occur as the kinetics of nanoparticle thermolysis become appreciable, thus forming nuclei at the film surface. Recent literature has shown the necessity of CuSe nuclei to induce sintering. Sulfurization of CuS doped CZTS films sinter, thus expanding these claims beyond the selenide.^ Throughout this work, experiments on the chalcogenization of varying nanoparticle materials, film architectures, and processing times, were studied to understand the sintering process. One disadvantage of annealing nanoparticle thin films is the formation of the carbon rich fine-grain layer due to pyrolysis of oleylamine nanoparticle ligands present from hot-injection synthesis. Oleylamine capped CZTS nanoparticles form a suitable film morphology that enables the percolation of liquid selenium throughout the nanoparticle film, and diffusion of copper to the film surface. Ligand exchange procedures with carbon-free diammonium sulfide suppress sintering by altering the nanoparticle film morphology with agglomerate formation, resulting in the fabrication of porous nanostructures that are more suited for thermoelectric application. Pre-annealing ligand exchanged films in air breaks the nanoparticle agglomerates, resulting in a higher degree of coarsening and improved device performance.^ Beyond diammonium sulfide, the effect of surface ligand on film sintering was researched. Hard soft acid base (HSAB) concepts qualitatively explain why soft Lewis bases such as Cu2S, Se, (NH4)2S bind strongly to copper and prevent film sintering. In contrast, nanoparticles recapped with hard Lewis bases like KOH, NH4OH, and NaNH2 show more sintering, presumably by allowing copper to react with selenium more readily. Thiourea which comprises both soft sulfide and hard amide results in both large and small grain sintering.^ As a lower cost alternative to CZTS nanoparticles, Cu2FeSnS 4 (CFTS) nanoparticles were synthesized via hot injection and fabricated for the first time into a PV device. Selenization of hot-injection synthesized tetragonal P4 CFTS nanoparticles sinter into ~1-2 µm stannite Cu2FeSn(S,Se)4 (CFTSS) grains. PV devices demonstrated photoconductivity; however, tin loss during annealing produced binary and ternary phases that shunted the devices.^ The aim of this work is to develop new solution processing methods and materials to lower the cost per watt of chalcogenide solar absorbers. The use of Corning® Willow® glass (CWG) enables roll-to-roll printing and provides a unique opportunity for processing flexible chalcogenide solar cells. Doping with NaF was demonstrated as a viable method to improve device efficiency, reaching a record 6.9% CZTSSe device. Thus this work establishes a plethora of solution processing tools towards achieving $1/Watt PV energy. ^




Rakesh Agrawal, Purdue University.

Subject Area

Chemical engineering|Materials science

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server