The process of photosynthesis starts with an absorption of a photon followed by an excitation energy transfer to the reaction centre, where charge separation occurs. Quantum efficiency of these primary processes of photosynthesis is high, ~80–95% [1]. Photosynthesis therefore serves as an inspiration for new ways of solar energy utilization, including artificial photosynthesis, various biohybrid devices, or novel types of solar cells [[2], [3], [4], [5]]. The high efficiency of photosynthetic light harvesting is a result of complex interactions between pigment molecules and proteins, which are difficult to imitate. An exception is a chlorosome, the main antenna of green photosynthetic bacteria [[6], [7], [8]]. The major pigment is bacteriochlorophyll (BChl) c, d or e, which self-organizes into aggregates without any assistance from protein. The aggregates are organized into various lamellar structures, ranging from disordered curved lamellae to well organized multilayered cylinders [9,10]. The baseplate of the chlorosome contains BChl a through which the absorbed energy is extracted from the interior and transported further towards the reaction centre. The whole chlorosome forms an oblong body attached to the inner side of the cytoplasmic membrane. With its large size (~100 nm) and a huge amount of pigments (tens of thousands), chlorosomes enable phototrophic life at extremely low light conditions [11,12]. The overall efficiency of energy trapping in green photosynthetic bacteria is about 85% [13].

Aggregates of chlorosomal BChls can also be prepared artificially, either in non-polar solvents like hexane [14,15], or in aqueous buffers. The latter environment requires the use of an aggregation-inducing lipophilic agent. Various lipids, pigments, or quinones were shown to induce aggregation [[16], [17], [18]]. Spatially well defined, chlorosome-resembling aggregates can be prepared in aqueous buffers with amphiphilic diblock copolymer [19] In all cases, the aggregation properties are determined by the length of the esterifying alcohol of the BChl molecule [20]. The ability of BChl c, d and e to form self-assembling structures with unique light harvesting properties is often mimicked in artificial systems [[21], [22], [23], [24]].

As with other (B)Chls, aggregates of chlorosomal BChls do not absorb in the green-yellow region of the spectrum. Should aggregates of BChls be utilized for artificial light harvesting, an increase in spectral coverage is highly desirable. Such an extension of the absorption spectrum is possible and several types of pigment molecules have already been tested and incorporated into the structure of the aggregates [[25], [26], [27]]. The hydrophobic space between the layers of polar chlorin rings in the lamellar structure of the aggregates is the place where non-polar molecules are incorporated in chlorosomes [28] and it can be used for the same purpose also in vitro. In addition to the increase of the spectral coverage, artificial light-harvesting system also requires a terminal acceptor of the excitation energy from which it could be collected for further use. A few were already tested. Zinc chlorins co-aggregated with energy traps led to 70% energy-transfer efficiency [29] and aggregates of BChl e with BChl a were shown to exhibit energy transfer efficiency approaching 100% [30]. Rapid energy equilibrium between aggregated BChls and BChl a on a timescale of tens of picosecond was observed in the diblock copolymer, chlorosome-like particles [22].

Our goal in this study was to create a self-assembling, multichromophore system with high energy-transfer efficiency. For this purpose we used the mixture of BChl c epimers from Chloroflexus aurantiacus with different esterifying alcohols in the same ratio as they are found in chlorosomes. β-carotene was added to the structure as an aggregation-inducing agent and to increase the spectral coverage in the blue-green region. In addition, BChl a was added as a final acceptor of the excitation energy. An advantage of this approach is that all pigments are purified from photosynthetic organisms and do not require expensive synthesis.

In addition, we have chosen and tested a set of methods to determine the efficiency of energy transfer, as we found that the traditional method of fluorescence excitation spectra can't be used without modifications (for the β-carotene-to-BChl c energy transfer) or at all (for the BChl c-to-BChl a energy transfer). These methods may be useful also for study of other complexes where a simple use of excitation spectra is not possible.