COMPOSITE LIGHT HARVESTING MATERIAL, DEVICE, AND METHOD

Abstract

A photovoltaic device and method utilizing a light harvesting device and a photovoltaic cell; wherein the light harvesting device includes an organic semiconductor photoactive layer capable of multiple exciton generation with a luminescent material dispersed therein; wherein the bandgap of the luminescent material is selected such that the triplet excitons, formed as a result from the multiple exciton generation in the organic semiconductor, can be transferred from the organic semiconductor into the luminescent material non-radiatively via Dexter Energy Transfer; a photovoltaic cell disposed in an emissive light path of the luminescent material and having a first photoactive layer, wherein the bandgap of the luminescent material matches or is higher than the bandgap of the first photoactive layer.

Description

BACKGROUND

The present invention relates, in general, to the composition of organic semiconductors capable of multiple exciton generation. Particular compositions can be used in photovoltaic and light emitting devices to give enhanced efficiencies.

Conventional solar cells are limited in efficiency to around 34%, mainly due to the thermalisation of above-bandgap photons and transmission of below-bandgap photons. The limit in efficiency is called the Shockley-Queisser limit.

A number of strategies have been used to make solar cells that exceed the Shockley-Queisser limit. Such strategies can involve the use of organic and inorganic tandem cells. However, it has been challenging to match the current of the sub-cells in these designs. Efficient transfer of energy between organic and inorganic semiconductors is a widely sought after property, with applications in photovoltaics (PVs), light emitting-diodes and sensors. To date, efforts to couple organic and inorganic semiconductors have focussed on the transfer of singlet excitons via Förster resonance and energy transfer (FRET).

US 2010/0193011 A (MAPEL ET AL) May 8, 2010 discloses solar concentrators to improve the efficiency of PV cells. The solar concentrator comprises an emitting chromophore effective to receive at least some energy by Förster energy transfer from another chromophore. The emitting chromophore emits some of the received energy at a wavelength that is red-shifted from the wavelength absorbed by the other chromophore.

Another strategy to make a solar cell that exceeds the Shockley-Queisser limit is disclosed in EHRLER, Bruno, et al. Singlet Exciton Fission-Sensitized Infrared Quantum Dot Solar Cells. Nano Lett. 2012, vol.12, p.1053-1057. Ehrler et al demonstrate an organic/inorganic hybrid photovoltaic device architecture that uses singlet exciton fission to permit the collection of two electrons per absorbed high-energy photon while simultaneously harvesting low-energy photons. Singlet exciton fission is a well-established process in organic semiconductors by which a photogenerated singlet exciton couples to a nearby molecule in the ground state, creating two triplet excitons. The transfer of triplet excitons is desirable because triplet excitons possess properties such as long lifetimes, up to several ms, and diffusion lengths up to several μm. However, the transfer of triplet excitons via Förster resonant energy transfer is spin forbidden as discussed in SCHOLES, G. D, et al. Long range resonance energy transfer in molecular systems. Annual review of physical chemistry. 2003, vol. 54, p. 57-87. In the device of Ehrler et al., infrared photons are absorbed using lead sulfide (PbS) nanocrystals. Visible photons are absorbed in pentacene to create singlet excitons, which undergo rapid exciton fission to produce pairs of triplets. Each of the two triplets can generate charge following dissociation at an organic/inorganic heterointerface meaning the direct effect of singlet fission in the solar cell is to double the photocurrent while halving the maximum possible photovoltage.

Accordingly, there is a need to further improve the coupling of organic and inorganic semiconductors and enable efficient energy transfer between them. Further there is a need to establish the conditions under which triplet excitons can undergo efficient energy transfer into inorganic semiconductors.

SUMMARY

It is therefore an object of the present invention to provide a composite light harvesting material capable of coupling organic and inorganic semiconductors together using energy transfer of triplet excitons. The composite light harvesting material has applications in photovoltaics (PVs), light emitting-diodes, lasers and sensors.

According to a first aspect of the present invention, there is provided a composite material comprising a host organic semiconductor material capable of multiple exciton generation dispersed with a luminescent material; wherein the bandgap of the luminescent material matches the energy of the triplet excitons formed as a result from the multiple exciton generation so that the triplet excitons are resonant with the bandgap of the inorganic luminescent material.

In the following the inventors demonstrate that the triplet excitons can be transferred from the organic semiconductor to the luminescent material via Dexter energy transfer. The triplet excitons, formed as a result of multiple exciton generation in the organic semiconductor are transferred from the organic semiconductor to the luminescent material via non-radiative energy transfer.

According to a second aspect of the present invention, there is provided a photovoltaic device, wherein a composite material is combined with a photovoltaic cell such that light emission from the composite material falls upon the photovoltaic cell. Therefore, according to the second aspect of the present invention, there is provided a photovoltaic device comprising an organic semiconductor photoactive layer capable of multiple exciton generation with a luminescent material dispersed therein; wherein the bandgap of the luminescent material is selected to match the energy of the triplet excitons formed as a result from the multiple exciton generation so that the triplet excitons are resonant with the lowest optical absorption band, termed here bandgap, of the luminescent material; a photovoltaic cell disposed in an emissive light path of the inorganic luminescent material and having a first photoactive layer, wherein the bandgap of the luminescent material matches or is higher than the bandgap of the first photoactive layer.

Preferably, the organic semiconductor photoactive layer is capable of singlet exciton fission. Examples of such organic semiconductor materials are polyacenes or oliogoacenes, optionally pentacene, tetracene or derivatives thereof selected from bis(triisopropyl-silylethynyl) pentacene (TIPS-P), diphenylpentacene (DPP), di-biphenyl-4-yl-pentacene (DBP), di(2′-thienyl)pentacene (DTP), and di-benzothiophene-pentacene (DBTP), bis(triisopropyl-silylethynyl) pentacene (TIPS-P), bis((triethyl)ethynyl)pentacene (TES-P) or rubrene, bis(triisopropyl-silylethynyl) tetracene (TIPS-T), di(2′-thienyl)tetracene (DTT).

Preferably, the organic semiconductor photoactive layer has a bandgap in the range 2.0 to 2.6 eV, preferably 2.2 to 2.5 eV, more preferably 2.4 eV.

The bandgap of the luminescent material is preferably within 0.4 eV of the bandgap of the energy of the triplet excitons, preferably within 0.3 eV, more preferably within 0.2 eV.

Preferably, the bandgap of the luminescent material is in the range of 0.6 eV to 1.6 eV, preferably 0.75 eV to 1.3 eV, more preferably 0.95 eV to 1.1 eV.

Advantageously, the luminescent material comprises an inorganic material, preferably a nanocrystalline semiconductor. In this case, the nanocrystal may comprise a lead chalcogenide nanocrystal such as lead selenide or lead sulfide. Other choices for nanocrystal semiconductor may include any one or more of nanocrystals comprising CdSe, CdS, ZnTe, ZnSe, PbS, PbSe, PbTe, HgS, HgSe, HgTe, HgCdTe, CdTe, CZTS, ZnS, CuInS2, CuInGaSe, CuInGaS, Si, InAs, InP, InSb, SnS2, Ge, CuS and Fe2S3.

A photovoltaic device may be arranged such that the light harvesting device comprises an organic semiconductor photoactive layer capable of multiple exciton generation with a luminescent material dispersed therein; wherein the bandgap of the luminescent material is selected such that the triplet excitons, formed as a result from the multiple exciton generation in the organic semiconductor, can be transferred from the organic semiconductor into the luminescent material, with at least one step mediated by non-radiative Dexter Energy Transfer. In this way the mediation can be provided by an intermediary step of different energy transfer mechanism provided that at least one step is mediated by non-radiative Dexter Energy Transfer.

As part of the photovoltaic the device, the light harvesting device may comprise an organic semiconductor photoactive layer capable of multiple exciton generation with luminescent nanocrystals dispersed therein; wherein the bandgap of the nanocrystals is selected such that the triplet excitons, formed as a result from the multiple exciton generation in the organic semiconductor, can be transferred from the organic semiconductor into the nanocrystals, where the last energy transfer step into the nanocrystals is mediated by non-radiative via Dexter Energy Transfer. As such, whilst the last energy transfer step into the nanocrystals is by non-radiative Dexter Energy Transfer, previous steps or hops or energy transfer from the organic semiconductor material may occur by alternative means including Förster resonance and energy transfer and via other materials.

Preferably a photon multiplier system comprising a film and containing the composite material described above is provided with at least one light-directing element to preferentially direct light emitted from the luminescent material towards one or a selection of the surfaces or edges.

Preferably, the organic semiconductor is an acene, an acene dimer, a perylene, a perylene dimer, a perylenediimide, a terylene, a terrylene, a thiophene, or a semiconducting polymer.

Preferably, suitable choices for the nanocrystal semiconductor comprises any one or more of nanocrystals comprising organometal halide perovskite or cesium lead halide perovskite.

In the device, the photovoltaic cell is preferably provided with the first photoactive layer comprising amorphous silicon. Alternatively, the photovoltaic cell is provided with the first photoactive layer comprising crystalline silicon, copper indium gallium selenide (CIGS), germanium, CdTe, GaAs InGaAs, InGaP, InP or perovskite semconductors such as organometal halide perovksite semiconductors and more specifically methylammonium lead iodide chloride (CH3NH3PbI2Cl).

Preferably, the nanocrystal semiconductor is passivated with ligands that solubilise them in solvents compatible with the organic semiconductor, preferably small molecules, more preferably amines or thiols.

Preferably, the mean distance between the luminescent components is chosen to be similar to the triplet exciton diffusion length in the organic semiconductor; where a low concentration of the luminescent component is necessary to minimise self-absorption by the luminescent component.

More preferably, the mean distance between the organic semiconductor and the luminescent material is between 10 nm and 2000 nm, more preferably between 20 nm and 200 nm.

More preferably, the mean distance between the luminescent components is between 10 nm and 2000 nm, more preferably between 20 nm and 200 nm.

In order to further enhance the efficiency of the device and harvest any photons emitted out of direction towards the photovoltaic cell, the organic semiconductor photoactive layer is preferably provided a layer to guide the light towards the photovoltaic cell. Preferably this layer is a selective wavelength reflecting layer or where the refractive indices of the device layers are tuned to guide the emission from the composite light harvesting device to the photovoltaic cell.

According to third aspect of the present invention, there is provided a light emitting device comprising an organic semiconductor layer with an inorganic luminescent material dispersed therein; wherein the bandgap of the inorganic luminescent material is selected to match the energy of the triplet excitons formed as a result of electrically injected charges into the organic semiconductor layer so that the triplet excitons are resonant with the bandgap of the inorganic luminescent material.

In each aspect of the invention and the preferred embodiments described herein, bandgap is taken to mean that the triplet excitons are resonant with the lowest optical absorption band.

BRIEF DESCRIPTION OF DRAWINGS

Embodiments of the invention will now be described, by way of example only, and with reference to the accompanying drawings of which:

FIG. 1 is a schematic diagram of singlet fission down conversion;

FIG. 2 is a schematic diagram of a device structure;

FIG. 3 is a graph of external quantum efficiency of a device according to a first embodiment of the invention compared to various controls;

FIG. 4A is a schematic diagram of singlet exciton fission in pentacene

FIG. 4B is a schematic diagram to illustrate how inorganic solar cells can be singlet fission sensitized using triplet transfer from a thin organic singlet fission layer;

FIG. 4C is a schematic diagram of possible processes a triplet exciton can undergo at an organic/inorganic interface;

FIG. 4D is a graph comparing kinetics at a pentacene ground state in pristine pentacene and bilayers with lead selenide nanocrystals;

FIG. 5A is a Transient Absorption spectra (TA) of pentacene with lead selenide bilayer, the data in FIG. 5A is decomposed into three excited state species, with spectra shown by the solid lines in FIG. 5B and corresponding population kinetics shown in FIG. 5C;

FIG. 5D illustrates blue-shifting of the Ground State Bleach (GSB) peaks of the lead selenide nanocrystals;

FIG. 6A is a graph of normalized kinetics of a pentacene spectral component extracted via the GA from transient absoprtion data of pentacene and lead selenide bilayers with varying nanocrystal bandgaps;

FIG. 6B is a graph of the corresponding PbSe spectral component of FIG. 6A;

FIG. 6C is a schematic diagram of the range of nanocrystal bandgaps for which triplet transfer was observed for a pentacene to lead selenide bilayer;

FIG. 7A is a graph of photoluminescence of lead selenide, pentacene and lead selenide bilayer films;

FIG. 7B is a graph showing photoluminescent enhancement from lead selenide is correlated to pentacence absorption; and

FIG. 8 is a schematic diagram of a light emitting diode based upon triplet transfer to inorganic nanocrystals.

DESCRIPTION OF EMBODIMENTS

According to FIG. 1 and first embodiment of the present invention, a composite material 10 comprises a thin film of an organic material 12 capable of singlet fission, mixed with inorganic nanocrystals 14. Light absorption of high-energy photons 16 in the organic material 12 creates singlet excitons 18 that rapidly undergo singlet fission to form two triplet excitons 20. The triplets excitons 20 cannot emit because of spin selection rules. However, the triplet excitons 20 will diffuse and undergo efficient Dexter energy transfer into the inorganic nanocrystals 14. The electron-hole pair in the inorganic nanocrystal 14 can then recombine radiatively, emitting a photon 22. Thus every high-energy photon 16 absorbed by the organic material 12 can lead to the emission of two low energy photons 22 by the nanocrystals 14. Low energy photons 24 pass through the organic material 12.

The photons 22 emitted by the nanocrystals 14 can be absorbed by an adjacent solar cell 26 such as presented in FIG. 2. A light trapping layer 28 such as a selective reflector 28 can be used such that it reflects photons 22 emitted by the nanocrystals 14 to enhance light incoupling into the adjacent solar cell. This is a way of converting the energy from one high-energy photon 16 into two low-energy photons 22 that match the bandgap of the solar cell 26, hence doubling the current generated from high-energy photons 16.

Referring to FIG. 3, we exemplify the first embodiment of the present invention using TIPS-tetracene molecules. PbSe quantum dots of 1.1 eV bandgap were embedded in the TIPS-Tetracene matrix (10% by weight) and the EQE of a silicon solar cell was observed. Trace 1 is external quantum efficiency of the silicon solar cell and trace 2 is a silicon solar cell with a film of composite material in the light path. Using a filter between the composite layer and the solar cell, low energy photons, including the photons emitted from the nanocrystals, can be filtered out (trace 3) and the ratio of traces 2 and 3 with and without a filter in trace 4 reveals that additional current generated by the nanocrystal emission in the infrared originates from absorption of the organic singlet fission sensitizer in the visible.

Trace 2 spectrum of FIG. 3 shows a clear dip where the tetracene absorbs and when a filter to remove the light emitted from the nanocrystals is placed between the tetracene film and the silicon solar cell as shown in trace 3, the overall efficiency goes down. Crucially, it reduces more where the tetracene absorbs, identifying the additional current from the triplet transfer in trace 4.

Accordingly therefore in the first embodiment of the present invention, the inventors have demonstrated efficient resonant-energy transfer of molecular spin triplet excitons from organic semiconductors to inorganic semiconductors. In the following description, we further demonstrate the physical process behind the transfer using ultrafast optical absorption spectroscopy to track the dynamics of triplets, generated in pentacene via singlet exciton fission, at the interface with lead selenide (PbSe) nanocrystals. We show that triplets transfer to PSe rapidly (<1 ps) and efficiently, with 1.8 triplets transferred for every photon absorbed in pentacene. The triplet transfer is most efficient when the bandgap of the nanocrystals is close to resonance (±0.2 eV) with the triplet energy. Following triplet transfer, the excitation can undergo either charge separation, allowing photovoltaic operation, or radiative recombination in the nanocrystal, enabling luminescent harvesting of triplet exciton energy in light emitting structures.

EXAMPLE

Singlet exciton fission (SF) is a process in organic semiconductors, by which a single photogenerated spin-singlet exciton is converted to two spin-triplet excitons on nearby chromophores. As the process is spin allowed, it can occur on sub 100 fs timescales with an efficiency of 200%, when the energetics of the system are favourable, i.e. the energy of the spin-singlet exciton is greater than or equal to twice the energy of the spin-triplet exciton. SF is a promising route to overcome the Shockley-Queisser limit in single-junction photovoltaics, if a SF material could be suitably combined with a low-bandgap inorganic semiconductor.

Therefore referring to FIG. 4A, singlet exciton fission in pentacene is demonstrated schematically wherein photogenerated singlet excitons S₁ undergo singlet fission SF into two triplet excitons 20 with a time T₁ of within 80 fs.

In this configuration, illustrated in FIG. 4B, the low-bandgap semiconductor 26 generates one electron-hole pair for each low-energy photon 24 absorbed, while the SF material 12 generates two triplet excitons 20 for each high-energy photon 16 absorbed. By distributing the energy of high-energy photons 16 into two excitations, SF could allow solar cells to overcome the otherwise dominant thermalisation losses. Ideally, the energy of the triplet excitons 20 could be directly transferred into the inorganic semiconductor 26, for which charge generation and collection are already optimized. This approach allows the SF material to act as an energy-funnelling layer on top of a conventional photovoltaic cell, rather than being an active part of the circuit.

Pentacene (Pc) is a model system for singlet exciton fission. Previous transient optical absorption (TA) measurements on Pc determined a fission rate of 80 fs, outcompeting alternative decay mechanisms. The fission-generated triplets can be efficiently dissociated at a heterojunction using the fullerene C60 as the acceptor, allowing for external quantum efficiencies (EQE) of 126%, the highest for any photovoltaic technology to date. There are two possible pathways for charge generation at such organic/inorganic interfaces, as shown in FIG. 4C. The first is electron transfer (ET) from Pc to PbSe. The second is energy transfer of triplet excitons (triplet transfer TT) into PbSe, followed by back-transfer of holes (hole transfer HT) into Pc to obtain charge separation.

Results

To investigate the dynamics of Pc triplet excitons at the interface with PbSe, the inventors performed TA measurements on thin PbSe/Pc bilayers, consisting of 1-2 monolayers of spin-coated PbSe, onto which 5 nm of Pc (3 molecular layers) was evaporated. The thinness of the pentacene layer ensures that all triplet excitons are generated close to the interface with PbSe. This allows the inventors to probe interfacial dynamics, which are normally masked by bulk diffusion processes. The samples were investigated with femtosecond (fs) TA spectroscopy, using a narrowband pump pulse centred at 550 nm and broadband probe pulses. In order to amplify the signal from the extremely thin layers, we use an optical cavity, which allows for multiple passes of collinear pump and probe beams through the sample. A series of PbSe nanocrystals with bandgap energies between 0.67 and 1.61 eV were compared against Pc.

FIG. 4D compares the TA kinetics, at 670 nm, in pristine Pc and Pc/PbSe bilayers with nanocrystals of varying bandgap. 670 nm is the position of the peak of the Pc ground state bleach (GSB). The lowest lying molecular triplet exciton (T1) in Pc has been reported to lie close to 0.86 eV. The two bilayers with nanocrystals of bandgap far below (0.67 eV) and above (1.61 eV) this energy, show almost identical kinetics to the pristine Pc film. In contrast, the bilayer with nanocrystals of bandgap 0.78 eV, which is close to resonance with T1, shows a significant drop in signal within the first 2 ps followed by a revival at later times. As we develop below, this kinetic behaviour is due to energy transfer, de-exciting the chromophore, followed by back charge transfer, re-exciting the chromophore.

However, the presence of excited state signals from the PbSe also needs to be taken into account, before we can quantify the populations of excited state species. FIG. 5A shows TA spectra of the Pc/PbSe (0.78 eV) bilayer. In order to extract the individual spectra and population kinetics of the various excited state species we use a genetic algorithm. The data in FIG. 5A can be decomposed into three excited state species, with spectra shown by the solid lines in FIG. 5B and corresponding population kinetics shown in FIG. 5C. One component 50 matches the signal expected for Pc, as demonstrated by the agreement with the spectrum of a pristine Pc film (dashed spectra 52). We note that after the ultrafast fission process (80 fs), there is no spectral evolution in the signal from Pc. Thus, in the experiments presented here, with a time resolution >300 fs, only a single spectrum is expected for Pc. The nanocrystals show two distinct spectra in this region (solid 54, 56), associated with relaxation from higher to lower energy excited states, as demonstrated by the early and later time signals from pristine PbSe films (dashed spectra 58, 60).

We now turn to the kinetics extracted by the genetic algorithm, shown in FIG. 5C, which represent the weight of the spectral component of an excited state species in the whole data set, rather than kinetics at a particular spectral point. We observe that the Pc component in the Pc/PbSe(0.78 eV) film (50) drops faster than in pristine Pc films (dashed 52), over the initial 0.3-3 ps, but subsequently rebounds, and at later times (>30 ps) has a larger signal than pristine Pc films. The population in the Pc, at the earliest times we can observe (300 fs) consists entirely of triplet excitons, due to the ultrafast (80 fs) fission process. Thus the reduction in signal between 0.2-3 ps represents a loss of triplet excitons. The increase in signal at later times is consistent with hole-transfer from the PbSe to Pc. The rise of the lower-energy excited state component of PbSe, curve 62 is associated with the reduction of the higher-energy PbSe excited state component (curve 54), as well as the loss of triplets in Pc. This suggests that both processes could populate the lower-energy PbSe excited state.

Lastly, we observe a blue-shifting in the GSB peaks of the PbSe nanocrystals, FIG. 5D. We consider this shift to be caused by charge transfer across the Pc/PbSe interface, which sets up microscopic electric fields, causing a Stark shift of the transition energy of species near the interface. This result in a derivative-like feature at the red-edge of the absorption feature, termed electroabsorption. The blue-shift of the GSB occurs over tens of ps, and hence, the associated charge-transfer process occurs over tens of ps. This timescale is consistent with the rebound in the Pc component, confirming the assignment of this rise to hole back transfer from the PbSe. No electroabsorption features or shifts in GSB were observed for pristine PbSe layers or Pc/PbSe bilayers in which the triplet transfer was not observed. Thus, the model that emerges from the data is one of forward triplet transfer from Pc to PbSe followed by back hole transfer from PbSe to Pc.

Therefore as illustrated at FIG. 5A, we show transient absorption (TA) spectra of Pc/PbSe (0.78 eV) films. Spectra are averaged over the indicated pump-probe delays. The pump fluence is 30 μJcm⁻² for the first pass and 0.4 μJcm-2 for the last pass. The TA data can be numerically decomposed into 5 components using a genetic algorithm (GA) and are shown in FIGS. 5A to 5D.

FIG. 6A shows normalized kinetics, extracted via the GA, of the Pc component in both pristine a Pc film and Pc/PbSe bilayers with varying nanocrystal bandgaps. FIG. 6B shows the corresponding PbSe component in both pristine nanocrystal films and Pc/PbSe bilayers. The initial triplet transfer and subsequent hole transfer are observed only with nanocrystal bandgap of 0.78 eV, FIG. 6A. In the other two cases, for nanocrystals of bandgap much higher or lower than the Pc triplet energy (0.86 eV), the kinetics are almost identical to pristine pentacene films, indicting very little or no triplet transfer.

Turning to the PbSe component, FIG. 6B, for the 0.67 eV nanocrystals, where no triplet transfer is observed, there is only a small difference between pristine PbSe (dashed 72) and Pc/PbSe (solid 74). However, for the 0.78 eV nanocrystals, where triplet transfer is observed, there is a large enhancement in signal for the Pc/PbSe sample (solid 70) in comparison to the pristine PbSe (dashed 68). The signal peaks at a later time, 4-5 ps, consistent with the timescale for triplet transfer (FIG. 6A). This shows that the increase in signal is due to transfer of triplets to the PbSe. The population is also found to be longer-lived, consistent with the formation of longer-lived charges following back hole transfer. In summary, in FIG. 6C, we find that triplet transfer (TT) from Pc occurs most efficiently in nanocrystals with bandgap of 0.78 eV or 0.93 eV. Nanocrystals whose bandgap lie within a narrow range (less than ±0.2 eV) of the Pc triplet energy (0.86 eV) show the most efficient triplet transfer.

The narrow energy range in which triplet transfer most efficiently occurs indicates the importance of the overlap of the density of states of donor and acceptor. The coupling integral for the energy transfer process contains contributions both from the Coulomb interaction and exchange interaction. The negligible oscillator strength of the S0→T1 transition for Pc means that the Coulomb interaction plays no significant role in the process. For the exchange interaction, DEXTER, D. L. A Theory of Sensitized Luminescence in Solids. The Journal of Chemical Physics. 1953. derived that:

$k\approx e^{-\frac{2R}{L}}J$

where, k is the rate of transfer, L is the orbital radius of donor and acceptors site, R is the separation between them and J is the normalised spectral overlap between donor emission and acceptor absorption.

Importantly, J is independent of the oscillator strengths of the optical transitions. Thus, triplet transfer would only be efficient to nanocrystals whose lowest-energy absorption feature, which has a large density of states, overlapped with the S0→T1 transition for Pc, at about 0.8 eV. The width of the lowest-energy absorption feature for the nanocrystal studied here is about 0.15 eV which corresponds well with the narrow range in which triplet transfer is observed, less than ±0.2 eV.

For the Pc/PbSe system studied here, triplet transfer can be followed by back hole transfer. But a fraction of the excitations may not undergo hole-transfer and recombine within the PbSe. Also at later times electron-hole recombination, of states previously separated across the interface, will occur. Both cases allow for radiative recombination and hence enhanced emission from the PbSe, whenever triplet transfer is possible. FIG. 7A shows the PL spectra, for PbSe and Pc/PbSe films, for excitation above (532 and 650 nm) and below (780 and 808 nm) the Pc bandgap. The red-shift of the PL peaks in the bilayer results from a red-shift in the absorption onset of the PbSe in Pc/PbSe films in comparison to pristine PbSe films. In all cases the Pc/PbSe shows enhanced PL in comparison to pristine PbSe. We consider the enhanced PL for below gap excitation (780 and 808 nm) to arise from increased radiative recombination in the presence of Pc, due to a change in the local environment of the PbSe. We normalize the PL signals using the ratio of the PL of Pc/PbSe to PbSe at 808 nm (factor of 36.5). This allows us to account for the increased radiative recombination, which should be independent of pump wavelength, and isolate enhanced PL arising from triple transfer. Following normalisation, no change in relative PL is seen at 780 nm, but much stronger PL is seen at 650 and 532 nm.

As shown in FIG. 7B, the stronger PL is correlated to the Pc absorption, confirming that it arises due to triplet energy transfer from the Pc. Near the peak of the Pc absorption, at 650nm, the Pc/PbSe (0.93 eV) bilayer absorbs 47% more light than the PbSe (0.93 eV) film (see S10). Crucially, at this wavelength the PL is enhanced by 85%. This means that for every photon absorbed by Pc 1.8 excitations contribute to the PL. As shown above, triplet transfer is the dominant process in the bilayers. Assuming that singlet fission proceeds with a 200% yield of triplets, this implies a minimum triplet transfer efficiency of 90%. No PL enhancement arising due to excitation of Pc was observed for Pc/PbSe samples where the nanocrystal bandgap blocked triplet transfer, which further demonstrates the enhancement does not arise from early time (sub 100 fs) transfer of singlet excitons via FRET. We note, that by suitable choice of SF material and inorganic acceptor, it would be possible to arrange energetics such that back hole transfer is blocked, allowing only for triplet transfer.

In conclusion, we have reported the first demonstration of triplet energy transfer from organic to inorganic semiconductors. Our studies of the photophysics of thin bilayer samples of pentacene/PbSe nanocrystals demonstrate that triplet energy transfer from pentacene to PbSe is efficient only when the nanocrystal bandgap is resonant with the molecular triplet energy. This result opens new avenues to couple organic and inorganic semiconductors and new possibilities for devices. For instance, to harness non-radiative triplet excitons generated via electrical injection of charges in to organic LEDs. The triplets could be harvested via transfer into inorganic nanocrystals where the electron-hole pair could recombine radiatively, allowing for white-light emission without the need for phosphorescent molecules. As demonstrated here, this process can also be used to harness triplet excitons generated via singlet exciton fission, allowing the energy of the triplets to be directly funnelled in to conventional inorganic solar cells. This offers a very promising method to overcome the Shockley-Queisser limit.

Methods

Nanocrystal fabrication: All chemicals were purchased from Sigma Aldrich, if not stated otherwise, and were anhydrous if available. PbSe nanocrystals were synthesized following standard methods 20. Briefly, Pb(OAc)2H2O (3.44 mmol; 1.3 g), oleic acid (OA; 8.58 mmol; 2.7 ml) and loctadecene (ODE; 75 mmol; 24 ml) were degassed at 100 ° C. under vacuum (10-2 mbar or better) for 2 h. In order to form the Pb-oleate precursor complex the temperature was raised to 160° C. under nitrogen atmosphere and subsequently changed to the desired Se-precursor injection temperature (120° C.-180° C.). In parallel, Se (Alfa Aeser, 10.8 mmol; 852.8 mg), diphenylphosphine (DPP; 15 μmol; 26.1 μl) and trinoctylphosphine (TOP; 24.2 mmol; 10.8 ml) were combined and stirred under nitrogen atmosphere to form the Se-precursor. PbSe nanocrystal growth was initiated by the rapid injection of the Se-precoursor into the prepared Pb-oleate solution. After the desired nanocrystal size was reached (20 sec-5 min) the reaction was quenched by injecting 20 ml of hexane and by placing the flask into an ice-cooled water bath. Subsequent purification steps were carried out in an argonfilled glovebox. The nanocrystals were extracted via repeated precipitation with a mixture of lbutanol and ethanol.

Sample Fabrication: Samples were fabricated on 0.13 mm thin cover glass slides. Nanocrystal films were deposited by a layer-by-layer method, in an inert environment, using 1,3-benzenedithiol as a crosslinking molecule, from a 5 mg/mL solution of PbSe. Subsequently, 5 nm of pentacene was evaporated on the nanocrystal films, in a vacuum better than 2×10-6 mbar. The samples were encapsulated with a second 0.13 mm thin glass slide and an epoxy glue before exposing to air.

Steady State Optical Measurements: The absorption spectra of the nanocrystals were taken in solution at 0.05-1 mg/mL using a PerkinElmer Lambda 9 UV VisIR spectrophotometer. PL was measured by illuminating a spot of ca. 2 mm in diameter with a diode lasers (MGL-III-532 for 532 nm, SMFR-R0004 for 650 nm, Lasermax-MDL for 780 nm, IQu1C135 for 808 nm). Lenses project the PL emitted to a solid angle of 0.1π onto an InGaAs detector (Andor DU490A-1.7) which has a cut-off at 1600 nm.

Transient absorption (TA) spectroscopy: In this technique a pump pulse generates photoexcitations within the film, which are then studied at some later time using a broadband probe pulse. A portion of the output of a Ti: Sapphire amplifier system (Spectra-Physics Solstice) operating at 1 KHz, was used to pump a TOPAS optical parametric amplifier (Light Conversion), to generate narrowband (10 nm FWHM) pump pulses centered at 550nm. Another portion of the amplifier output was used to pump a home built non-collinear optical parametric amplifier (NOPA). The probe beam was split to generate a reference beam so that laser fluctuations could be normalized. Pump and probe beams were made collinear with a beam splitter and entered an optical cavity, consisting of two concave mirrors (focal length f) placed 4f apart from each other with the sample in the center. The beams underwent multiple bounces in the cavity, making multiple passes in the sample, thus allowing for the weak signal from the thin layers to be amplified. After exiting the cavity a long pass filter was used to block the pump beam, while allowing the probe beam to pass. The probe and reference beams were dispersed in a spectrometer (Andor, Shamrock SR-303i) and detected using a pair of 16-bit 512-pixel linear image sensors (Hamamatsu). The probe was delayed using a mechanical delay stage (Newport) and every second pump pulse was omitted using a mechanical chopper. Data acquisition at 1 kHz was enabled by a custom-built board from Stresing Entwicklunsbüro. The differential transmission (ΔT/T) was calculated after accumulating and averaging 1000 “pump on” and “pump off” shots for each data point.

Due to the group velocity mismatch between pump and probe wavelengths there is a reduction in time resolution of the experiment. From the rise time of the signal we estimate the time resolution of the experiment to be about 300 fs at 670 nm. While this is insufficient to study the initial singlet fission process in Pc, which proceeds on sub 100 fs timescales, it is sufficient to study the triplet transfer process.

Numerical Methods: We use numerical methods based on a genetic algorithm to deconvolute the overlapping spectral signatures of individual excited states and obtain their kinetics. In summary, a large population of random spectra are generated and bred to form successive generations of offspring, using a survival of the fittest approach. The best spectra are returned as optimized solutions. For a given solution, the fitness is calculated as the inverse of the sum of squared residual with a penalty added for non-physical results. The parent spectra are selected using a tournament method with adaptive crossover. The offspring are generated using a Gaussian-function mask of random parameters

According to a third embodiment of the present invention, in organic light-emitting diodes, excitons are generated from charges, electrically injected into the active layer. These charges (electrons, e−, and holes, h+) form only 25% of the emissive singlet excitons and 75% of the non-emissive triplet excitons. Hence, without phosphorescence, only 25% of the charges can be converted into light. With a small fraction of nanocrystals in the active layer, those triplets can be converted into an emissive species and generate additional light as illustrated in FIG. 8. Also, such an LED would emit at two different wavelengths, opening the possibility for white-light emission from a single layer of composite material.

Claims

1. A method of luminescent harvesting of triplet exciton energy in a photovoltaic device comprising: a light harvesting device; anda photovoltaic cell;wherein the method includes:providing a light harvesting device comprising an organic semiconductor photoactive layer formed of an organic semiconductor donor and capable of multiple exciton generation with a luminescent material dispersed therein;wherein the luminescent material comprises a nanocrystalline semiconductor passivated with ligands that solubilise the nanocrystalline semiconductor in at least one solvent compatible with the organic semiconductor;wherein the bandgap of the nanocrystalline semiconductor is selected to be resonant with triplet excitons such that the organic semiconductor is operative upon absorption of light of a first wavelength to transfer triplet excitons, formed as a result from the multiple exciton generation in the organic semiconductor, into the nanocrystalline semiconductor non-radiatively via Dexter Energy Transfer, wherein at least one energy transfer step of transferring the triplet excitons into the nanocrystalline semiconductor acceptor from the organic semiconductor donor is mediated by the non-radiative Dexter Energy Transfer, whereby to generate emission of light of a second wavelength, longer than the first wavelength, from the nanocrystalline semiconductor;wherein the photovoltaic cell is disposed in an emissive light path of the luminescent material and has a first photoactive layer; andwherein the bandgap of the first photoactive layer matches or is lower than the bandgap of the nanocrystalline semiconductor.

2. The method of claim 1, wherein the organic semiconductor photoactive layer is capable of singlet exciton fission.

3. The method of claim 2, wherein the organic semiconductor is an oligoacene.

4. The method of claim 3, wherein the oligoacene is pentacene, tetracene or derivatives thereof bis(triisopropylsilylethynyl)tetracene (TIPS-T) or di(2′-thienyl)tetracene (DTT).

5. The method of claim 1, wherein the organic semiconductor photoactive layer has a bandgap in the range 2.0 to 3.0 eV.

6. The method of claim 1, wherein the bandgap of the nanocrystalline semiconductor is within 0.4 eV of the bandgap of the energy of the triplet excitons.

7. The method of claim 1, wherein the bandgap of the nanocrystalline semiconductor is in the range of 0.6 eV to 1.6 eV.

8. The method of claim 1, wherein the nanocrystalline semiconductor comprises a lead chalcogenide nanocrystal.

9. The method of claim 8, wherein the lead chalcogenide nanocrystal is lead selenide or lead sulfide.

10. The method of claim 8, wherein the nanocrystalline semiconductor comprises any one or more of nanocrystals comprising CdSe, CdS, ZnTe, ZnSe, PbS, PbSe, PbTe, HgS, HgSe, HgTe, HgCdTe, CdTe, CZTS, ZnS, CuInS2, CuInGaSe, CuInGaS, Si, InAs, InP, InSb, SnS2, CuS, Ge, and Fe2S3.

11. The method of claim 1, where the mean distance between luminescent components of the luminescent material is chosen to be similar to the triplet exciton diffusion length in the organic semiconductor, wherein a concentration of luminescent components minimizes self-absorption by the luminescent components.

12. The method of claim 1, wherein the mean distance between luminescent components of the luminescent material is between 10 nm and 2000 nm.

13. The method of claim 1, wherein the photovoltaic cell is provided with the first photoactive layer comprising silicon.

14. The method of claim 1, wherein the photovoltaic cell is provided with the first photoactive layer comprising one or more of crystalline silicon, amorphous silicon, copper indium gallium selenide (CIGS), germanium, CdTe, GaAs, InGaAs, InGaP, InP, quantum dot, metal oxide, organic polymer or small molecule or perovskite semiconductors.

15. The method of claim 1, wherein the emission from the luminescent material is guided to the photovoltaic cell.

16. The method of claim 1, wherein a last energy transfer step of transferring the triplet excitons into the-nanocrystalline semiconductor is mediated by non-radiative Dexter Energy Transfer.

17. The method of claim 1, wherein the organic semiconductor is an acene, an acene dimer, a perylene, a perylene dimer, a perylenediimide, a terrylene, a thiophene, or a semiconducting polymer.

18. The method of claim 1, wherein the nanocrystalline semiconductor comprises any one or more of nanocrystals comprising organometal halide perovskite or caesium lead halide perovskite.