Single-crystal semiconductor thin films for biohybrid photovoltaic devices

Although solar energy is freely available, the utility cost for electricity generated from photovoltaic (PV) modules is higher than that from coal or natural gas. Existing silicon (Si)-based PV devices are efficient, but the high manufacturing costs eventually contribute to the high utility cost. Energy generation from artificial photosynthesis, which borrows partial steps from natural photosynthesis, is a promising strategy for future energy supply.¹ Photocatalytic metal oxides, such as titanium dioxide (TiO₂) thin films, are used for artificial photosynthesis to harvest solar energy. However, for the wide-scale use of metal oxide solar cells, it is essential to have low production costs and high efficiency.

Thin-film morphology is a very important determinant of solar energy conversion efficiency. A 1D structure is particularly advantageous in photoelectrochemical applications. Several methods have been introduced to synthesize 1D metal oxide electrodes,² but most of them are multistep processes that are hard to scale up. Our lab recently developed a novel thin-film coating system called aerosol chemical vapor deposition (ACVD)³ that is simple and operates at atmospheric pressure. Conventional chemical vapor deposition takes place at low pressure, and particle formation is a nuisance for homogeneous and uniform deposition. In contrast, ACVD involves intentionally forming particles and depositing them on a substrate to form thin films with high surface area.

ACVD involves feeding the precursor, for example, titanium tetra isopropoxide (TTIP) for TiO₂ films, into a heated reactor (see Figure 1). The precursor decomposes at high temperatures into monomers that collide to form particles. The concentration gradient results in particles diffusing toward the heated substrate, and the particles get deposited. The high temperature causes the deposited particles to sinter and form nanostructures on the substrate. Thin films with various morphologies can be synthesized by ACVD. The morphology is mainly affected by the deposited particles' sintering rate, which in turn depends on the substrate temperature and the size of the particles being deposited. The particles' size can be altered by changing the precursor concentration or the residence time in the reactor.

Figure 1. Aerosol chemical vapor deposition (ACVD) (left) and particle deposition in ACVD (right). N₂: Nitrogen gas. ITO: Tin-doped indium oxide.

Figure 2(a) and (b) depicts the side view of scanning electron microscopy images for a columnar and a granular TiO₂ film, respectively. The relatively large depositing particles do not sinter to the film, resulting in granular morphology. In that morphology, the grain boundary between particles accelerates the recombination of photogenerated electrons and holes. However, by choosing the right process conditions to form relatively smaller particles, and the right substrate temperature, the arrival rate of particles is balanced with their sintering rate. That results in a 1D single-crystal columnar morphology. We also found in simulations for nanoparticle deposition that the sintering rate plays an important role in determining the film morphology: see Figure 2(c--d). In photoelectrochemical devices based on metal oxide nanostructured electrodes, we observed that the columnar morphology exhibits the best performance.^{3, 4} This is because of the single-crystal structure, which results in lower electron-hole recombination at the grain boundaries.

Figure 2. Scanning electron microscopy images of TiO₂films showing different morphologies generated by ACVD: (a) columnar and (b) granular films. Simulation images highlighting differences in morphology obtained by changing sintering rates: (c) columnar and (d) granular films.

Unlike Si-based PV devices, metal oxide-based PV devices require a light absorber that can effectively harvest photon energy under sunlight illumination. TiO₂ or other metal oxide electrodes with a wide bandgap are used mainly for electron transfer rather than for light absorption. There are various light absorbers, including dyes, quantum dots, and wires. However, their light absorption range is narrow and often limited to the visible region. Light-harvesting complexes, present in living organisms, have evolved over time to absorb and transfer light efficiently. A novel device combines natural light-harvesting antenna and an artificial reaction center (see Figure 3).⁵ The black dye molecules, which act as an artificial reaction center, were adsorbed onto the surface of TiO_2, followed by electrospray deposition of natural antenna chlorosomes. On illumination of the biohybrid device, chlorosomes absorb light and transfer it to the black dye, which separates charge and inject electrons into TiO₂. The biohybrid PV device has shown 30 times enhancement in incident photon conversion efficiencies at peak absorption wavelengths compared to a normal black-dye-sensitized PV device.⁵

Figure 3. (a) Biohybrid photovoltaic (PV) device and (b) photocurrent results with respect to wavelengths of incident light with and without chlorosome deposition by electrospray. (Reproduced by permission of the Royal Society of Chemistry.) TiO₂: Titanium dioxide. P3OT: Poly(3-octylthiophene-2,5-diyl). MA/cm²: Milliamperes per square centimeter.

Chlorosomes consist of self-assembled bacteriochlorophyll c (BChl c) pigments that are closely packed. The chlorosomes' absorption spectrum cannot be changed since they are extracted from natural organisms. However, it can be altered by controlling the pigments' self-assembly. Electrospray can be used to synthetically assemble BChl c pigments or their analogs to mimic chlorosomes and deposit them in one step. The goal is to deposit various layers of red-shifted dyes, which can transfer energy by Förster resonance energy transfer to fabricate an energy funnel that absorbs light over the complete solar spectrum.

In summary, we have developed an ACVD system to synthesize single-crystal semiconductor columnar films. We successfully deposited chlorosomes using an electrospray system over black dye molecules, which are adsorbed onto TiO₂ film with an ester linkage to form a biohybrid device. We showed that that aerosol processes are promising both for reducing production costs and enhancing device performance. Going forward, we will work to further improve the efficiency of the device by synthetically assembling absorbers similar to chlorosomes using BChl c analogs and span the solar spectrum with multiple layers of self-assembled dyes.

This work was financially supported by the Photosynthetic Antenna Research Center, an energy frontier research center funded by a Department of Energy project (DOESC0001035). This work was conducted collaboratively with Robert Blankenship and Dewey Holten at Washington University in St. Louis.