Imaging the photoionization of individual CdSe/CdS core-shell nanocrystals on n- and p-type silicon substrates with thin oxides

The low-intensity photoionization of individual semiconductor nanocrystals, at 23 degreesC in dry nitrogen, is time-resolved over many hours for both S (532-nm excitation) and P (395-nm excitation) nanocrystal excited states using electrostatic force microscopy. Over 7000 calibrated charge measurements have been made on 14- and 21-Angstrom-thick oxide layers. Photoexcited electrons tunnel across the oxide into the silicon, and multiple charges can build up on individual nanocrystals at intensities of only 0.1-0.01 W/cm(2). The silicon dopant type influences the net nanocrystal charging via the interfacial band bending; P-type subtrates show a faster nanocrystal reneutralization rate due to their higher interfacial electron concentration. There is a huge range of photoionzation behavior for individual nanocrystals. This behavior is different for 395- and 532-nm excitation in the same nanocrystal. This individuality seems in part to reflect tunneling through spatially localized defect states in the oxide. The line widths of spatial charge images of individual nanocrystals and the semicontinuous rate of charge re-neutralization after excitation suggest that we observe trapped electron motion in the adjacent oxide and/or on the nanocrystal surface, in addition to the ionized nanocrystal. On average, tunneling of the excited P electron is faster by 1-2 orders of magnitude than that of the S electron; the data show direct photoionization from the excited P state. A kinetic model is developed, including the effect of charging energy on tunneling rate, and applied to ensemble average behavior. There is no quantitative agreement of the tunneling-rate dependence on oxide thickness and excitation energy with the simple I D effective mass tunneling model. However, overall observed trends are rationalized in light of current thin-oxide tunneling literature.