Understanding the Bias Introduced in Quantum Dot Blinking Using Change Point Analysis

The fluorescence intermittency of single CdSe/CdS quantum dots (QDs) with different shell sizes is studied using the conventional bin and threshold method and the statistically more rigorous method, change point analysis (CPA). The on-state truncation time (τ_c) is a critical value used to interpret the dynamics of charge trapping in single QDs; however, changing the bin size and threshold in blink traces significantly modifies τ_c. Herein, we use the CPA method to minimize the bias that binning and thresholding introduces and find that a widely used assumption that there is only one on and one off state is questionable. We observe that 12 out of 17 QDs exhibit more than two intensity levels and find that the τcvalues of individual levels differ from the values obtained when the levels are combined, i.e., when one assumes there is only one on and one off state as in the conventional bin and threshold method. For instance, one QD has τ_c values of 0.5 (0.1) and 2.0 (0.2) s from two different intensity levels, whereas when the levels are combined into only one on state, τ_c is found to be 7 (1) s. The CPA method is found to be more suitable for studying multilevel emission in QDs than the conventional bin and threshold method.