Quantifying Förster Resonance Energy Transfer from Single Perovskite Quantum Dots to Organic Dyes

Abstract

Colloidal quantum dots (QDs) are promising regenerable photoredox catalysts offering broadly tunable redox potentials along with high absorption coefficients. QDs have thus far been examined for various organic transformations, water splitting, and CO₂ reduction. Vast opportunities emerge from coupling QDs with other homogeneous catalysts, such as transition metal complexes or organic dyes, into hybrid nanoassemblies exploiting energy transfer (ET), leveraging a large absorption cross-section of QDs and long-lived triplet states of cocatalysts. However, a thorough understanding and further engineering of the complex operational mechanisms of hybrid nanoassemblies require simultaneously controlling the surface chemistry of the QDs and probing dynamics at sufficient spatiotemporal resolution. Here, we probe the ET from single lead halide perovskite QDs, capped by alkylphospholipid ligands, to organic dye molecules employing single-particle photoluminescence spectroscopy with single-photon resolution. We identify a Förster-type ET by spatial, temporal, and photon-photon correlations in the QD and dye emission. Discrete quenching steps in the acceptor emission reveal stochastic photobleaching events of individual organic dyes, allowing a precise quantification of the transfer efficiency, which is >70% for QD-dye complexes with strong donor-acceptor spectral overlap. Our work explores the processes occurring at the QD/molecule interface and demonstrates the feasibility of sensitizing organic photocatalysts with QDs.

Keywords: FRET; energy transfer; perovskite; photocatalysis; photoluminescence; quantum dot.

Figure 1

Perovskite QDs as photocatalysts. (a) Illustration of a QD with a nonpermeable ligand shell and its ET pathways to organic dyes. Direct charge/energy transfer to the reactant is slow (pale blue arrow), whereas the FRET to a dye (orange wave) and subsequent charge transfer (CT) to the reactant (blue arrow) can be faster. (b) Normalized PL and absorption spectra of the donor (CsPbBr₃ QDs, blue) and acceptor (cyanine 3 NHS ester, red) used in this study. The yellow area indicates the spectral overlap of donor PL and acceptor absorption that is required for FRET.

Figure 2

Spatial and temporal donor–acceptor correlations. (a, b) Wide-field images selectively probing the green emission of QDs (a) and the red emission from dyes (b) show a spatial correlation of bright spots in the two channels. Scale bars correspond to 2 μm. (c, d) Time traces (10 ms time-binning) of the green QD donor (c) and red dye acceptor (d) emission in a bright spot demonstrating strong temporal correlation. Three distinct types of correlation are highlighted via gray bars, namely, (i) the presence of FRET if the QD is in a bright state, (ii) the absence of FRET if the QD is dark, and (iii) photobleaching events of the dye leading to an increase in the QD emission. Note: Vertical axes do not include zero counts.

Figure 3

Intensity and photon–photon correlations of the QD and the Cy3 emission. (a) Correlation of the intensities (10 ms time-binning) in the green QD donor and the red dye acceptor channel. The probability is indicated by a logarithmic heatmap, and the types of correlation are indicated by arrows: bright QD and bright dyes due to ET (light red, 1), dark QD and dark dyes due to quenched QD emission with the absence of ET (red, 2), brighter QD and darker dyes after photobleaching of the first (light blue, 3) and second dye (dark blue, 4). Note: Axes do not include zero counts. (b) Second-order correlation function (g²(τ)) of the donor and acceptor photon arrival times. The antibunching (dip at a zero-delay time) indicates anticorrelated emission of donor and acceptor.

Figure 4

Number of acceptors and FRET efficiency for a single QD. (a, b) Intensity in the red acceptor channel for a single QD with one (a) and three molecules (b) undergoing FRET with the QD. The transparent blue lines serve as a guide to the eye and indicate the stepwise photobleaching of the molecules. Note that vertical axes do not start at zero counts. (c) Histogram of the counted acceptor molecules per single QD (gray bars). The black line corresponds to a fitted Poisson distribution with a mean number of molecules of 0.90(9). (d) Representative time-resolved PL traces of single QDs in samples without dye (gray line) and with increasing amounts of dye (blue to red lines). (e) Dependence of the sample-averaged single-particle FRET efficiency on the number of accepting molecules obtained from single-particle QD excited-state lifetimes estimated as 1/e decay times. The black line corresponds to a fit of eq 1 to the data, yielding a single-QD-to-single-dye efficiency E₀ = 0.12(2), and error bars correspond to 95% confidence intervals.

Figure 5

Energy transfer from CsPbBr₃ QDs to a perylene bisimide (PBI) dye. (a) Structural formula of the PBI dye N,N′-bis(2,6-diisopropylphenyl)-1,6,7,12-tetraphenoxy-3,4,9,10-perylenetetracarboxylic diimide. (b) Ensemble PL spectra of CsPbBr₃ QDs at increasing dye concentrations. (c) Time-resolved PL of the QD donor at increasing dye concentrations. (d) Time traces (10 ms binning) of green donor (top) and red acceptor (bottom) intensity in a bright spot of a single QD. (e) Intensity correlation map corresponding to the traces in (d), displaying anticorrelation of donor and acceptor counts. (f) Second-order correlation function (g²(τ)) of green donor and red acceptor photons exhibiting antibunching.