Designing New Atomic Layer Deposition Precursors for Third-Generation Photovoltaics

Atomic layer deposition (ALD) holds a critical position in nanofabrication because of its uniformity, conformality, and subnanometer thickness control. Its abilities arise from its chemical mechanism: precursors saturate the surface with a self-limiting monolayer and prevent continuous growth, then react with another co-reactant to deposit a thin film of material. Consequently, ALD first requires precursors with suitable volatility, chemical reactivity, and thermal stability. Motivated by the prospects of using ALD to fabricate next-generation solar cells, we set out to design and develop new Pb-containing ALD precursors. We began by attempting to prepare Pb(II) analogues of N-heterocyclic carbenes, eventually finding the stable lead(II) rac-N,N′-di-tert-butylbutane-2,3-diamide (1Pb) that displayed the best volatility of any reported Pb(II) complex (1 Torr vapour pressure at 94◦C). With 1Pb and the well-known lead(II) bis[bis(trimethylsilylamide)] (0Pb), we developed an ultralow-temperature ALD process for PbS using H2S. Crystalline PbS films were deposited with as low as 45◦C with 1Pb and 65◦C with 1Pb, enabling deposition on thermally sensitive substrates like CH3NH3PbI3 (MAPI) perovskites. Not only did the process encapsulate MAPI thin films, the films also had tunable p-type conductivity, hole mobility, and carrier concentrations. An unusual reduction to elemental Pb(0) of 1Pb by H2S was found to begin at 95◦C and its mechanism was studied experimentally and theoretically. ALD at higher temperatures with 0Pb and 1Pb was limited by their thermal stability (<155◦C), so we designed the more stable Pb-containing precursor bis[lead(II) N,N′-di-tert-butyl-1,1-dimethylsilanediamide] [(4Pb)2]. An in-depth theoretical study of its bonding, thermodynamics, and molecular electrostatic properties gave us a more nuanced view of the chemical nature of volatility with implications for future precursor design. The success of (4Pb)2 led us to adapt its ligand for Co, yielding the unprecedented spirocyclic low-spin Co(IV) bis[N,N′-di-tert-butyl-1,1-dimethyldiamidosilane] (4Co). Finally, we explored Sn trifluoroacetates as green ALD precursors to the transparent conductor F-doped SnOx. We analyzed their structure and bonding in the solid state and vapour phase, showing that their polymeric structures actually increased their volatility, which may represent a new precursor design strategy altogether.