The nanocrystal-ligand interface of colloidal semiconductors has attracted significant research interest owing to their dominant role in tuning and tailoring the physio-chemical properties of such functional nanomaterials. For metal chalcogenide quantum dots (QDs), a comprehensive picture has emerged over the past decade on surface passivation and photoluminescence relationship. The surface of QDs is metal rich passivated by X-type ligands. Displacing them from the surface results in a marked drop of PLQY, which could be reversed by readsorption of ligands. DFT studies indicate that this is linked to the formation of under-coordinated surface Se which behave as traps for electrons or holes. QDs are nanometre-size crystallites with crystal facets of few square nanometres. This differs markedly from 2-D nanocrystals which are few monolayers thick with atomically flat top and bottom surfaces measuring hundreds of square nanometre e.g. CdSe nanoplatelets. A pronounced interplay between the ligand capping and optical properties of nanoplatelets has been reported. But, research on CdSe nanoplatelets is often hampered by their limited colloidal stability and unwanted stacking in bundles, two aspects that are linked to nanoplatelet-ligand interface. This combination of promising properties and cumbersome processing calls for an in-depth study of their surface chemistry. Here, the question as to whether concepts developed for multifaceted QDs can be transferred to nanoplatelets, which are terminated solely by atomically flat interfaces, stands out.

In this work, detailed study of the surface-chemistry of CdSe nanoplatelets is reported. We first introduce an improved synthesis strategy for nanoplatelets that yields colloidally stable and aggregation free nanoplatelets suspensions. Despite large surface area, these core-only nanoplatelets have a PLQY as high as 55%. Elemental analysis show nanoplatelets are Cd rich.¹H NMR analysis show that Cd excess comes with a surface termination by X-type carboxylate ligands, a binding motif similar to QDs. Addition of an L-type ligand displaces cadmium-carboxylate complexes from surface. The displacement isotherm unravelled that the surface features adsorption sites with different binding energies for CdX₂. We further emphasize the validity of a multiple-site model by DFT calculations, which yield a variation in binding sites from edges to the centre of the facet. Moreover, we analysed different types of mid-gap trap states and site dependent surface reconstruction the nanoplatelet undergoes with displacement of the ligands. Finally, this surface-model for CdSe nanoplatelets is most likely not restricted to CdSe only and should be considered when analysing the surface reactions of another 2-D nanoplatelet system.