Efficient energy transfer mitigates parasitic light absorption in molecular charge-extraction layers for perovskite solar cells

Abstract

Organic semiconductors are commonly used as charge-extraction layers in metal-halide perovskite solar cells. However, parasitic light absorption in the sun-facing front molecular layer, through which sun light must propagate before reaching the perovskite layer, may lower the power conversion efficiency of such devices. Here, we show that such losses may be eliminated through efficient excitation energy transfer from a photoexcited polymer layer to the underlying perovskite. Experimentally observed energy transfer between a range of different polymer films and a methylammonium lead iodide perovskite layer was used as basis for modelling the efficacy of the mechanism as a function of layer thickness, photoluminescence quantum efficiency and absorption coefficient of the organic polymer film. Our findings reveal that efficient energy transfer can be achieved for thin (≤10 nm) organic charge-extraction layers exhibiting high photoluminescence quantum efficiency. We further explore how the morphology of such thin polymer layers may be affected by interface formation with the perovskite.

Fig. 1. PL transients revealing energy transfer between polymer and MAPbI₃ thin films.

a Samples examined were either thin films of polymers deposited on MAPbI₃ films on quartz (schematic diagram on the left) or polymers deposited directly on quartz to serve as references (schematic diagram on the right). ET may occur between the polymer and the MAPbI₃ layer, leading to PL quenching, as depicted schematically in b. The left column shows time-resolved PL transients as a function of time after excitation for polymers c Super Yellow, e F8BT and g P3HT, when deposited as films of thickness L on either MAPbI₃:quartz (orange lines) or just on quartz (red lines). The dashed lines are monoexponential fits to the data, yielding the PL lifetime of each polymer on quartz in the absence of ET (lifetime values provided in Supplementary Table 4). The solid black lines are fits to data based on an ET model described in the main text, with the grey shaded area indicating the error margin. The insets show equivalent data for thicker polymer films on either MAPbI₃:quartz (grey lines) or just on quartz (blue lines). The right column depicts the expected PL-decay traces for polymers d Super Yellow, f F8BT and h P3HT for different film thickness L, as determined from the ET model. PL transients for the scarcely emitting PTAA samples are provided as Supplementary Fig. 15.

Fig. 2. Schematic representation of the model describing ET in polymer:MAPbI₃ two-layer films.

The top right shows the two-layer sample architecture on quartz substrates and the typical excitation geometry. Polymer excitation from the front surface causes an exponentially decaying profile in the polymer layer. ET between the polymer and the underlying MAPbI₃ results in accelerated PL decays (top right). The bottom graphic shows the hierarchy of the modelling and its dependence on input parameters: polymer film thickness L, PL lifetime τ_d of the polymer in the absence of MAPbI₃, polymer PLQE and polymerabsorption coefficient at the wavelength of excitation α(λ_exc). The Förster radius $R_0^{\mathrm{SY}}$ for ET between Super Yellow and MAPbI₃, and the nominal acceptor concentration C_a that applies to MAPbI₃, are first determined from fits to PL-decay transients for Super Yellow:MAPbI₃ samples, and used to calibrate modelled transients for samples involving the other polymers.

Fig. 3. Modelled ET efficiency between a semiconducting charge-transport layer and a MAPbI₃ layer.

ET efficiency as a function of the PLQE, film thickness and absorption coefficient of the CTL. a False-colour plot indicating the efficiency of ET between a photoexcited CTL and a MAPbI₃ layer underneath, as a function of organic CTL thickness L and PLQE. b ET efficiency as a function of semiconductor CTL thickness L for distinct PLQE values, and absorption coefficient α(λ_exc) = 13 × 10⁴ cm⁻¹. c ET efficiency as a function of the CTL absorption coefficient α(λ_exc) at the excitation wavelength, for different film thicknesses and a value of 50% PLQE. Additional false-colour plots mapping out this parameter space with wider ranges are shown in Supplementary Fig. 17.

Fig. 4. Changes in PL spectral shape when P3HT films of different thicknesses are deposited on MAPbI₃ or quartz.

Normalised PL intensity from P3HT films deposited through spin coating from solution with varying spin speeds onto either quartz (dashed lines) or MAPbI₃:quartz (solid lines) substrates. The line colours indicate spin speeds that anti-correlate with the P3HT film thicknesses achieved. Spin speed was increased from 2000 to 6000 rpm in steps of 1000 rpm, yielding P3HT film thicknesses from 92 to 40 nm for the films on MAPbI₃:quartz and 77 to 42 nm for the P3HT films on quartz (film thickness does not scale linearly with spin speed; the full list of values is provided in Supplementary Table 3). In addition, the normalised PL spectrum of a MAPbI₃ film on quartz is shown as a dotted brown line.