Quantum dots: The pros and cons in PV

Market penetration of competitive PV technologies is limited mainly by cost, which in turn depends on device efficiency, material availability and the energy requirement of production process. PV device researchers focus on incorporating novel concepts and device modification to optimize the solar cells which have already reached commercial production. Plan B is to come up with an alternative PV technology starting with noble material selection and working all the way up to full-scale production. One promising technology involves quantum dots (QD). Besides PV, they have applications in areas of photocatalysis, image sensors, light emitting diodes, high efficiency transistors and optical communication. Their name comes from their appearance as little dots which are roughly around 1 to 10 nanometers in size. QDs have a couple of significant features which attracted PV researchers in the first place. Commercially established PV devices based on silicon and thin film depend on micrometer-thick layers of materials for sufficient light absorption. QD nanoparticles, having high absorption, hold the promise of significantly reducing the absorber material requirement. Commonly investigated QD materials like cadmium selenide, lead sulphide, cadmium sulphide and indium phosphide typically have a couple of orders’ higher absorption than silicon and thin film materials. This, as a rule of thumb, would generate higher photocurrents for the QD-based devices. However, the catch with QDs is their size-dependent spectral response and photoluminescence.
Nanoscale semiconductors can exhibit shifts in absorption spectra with varying particle sizes. Having the right sizes and combination of QDs leads to broader coverage of the incident spectrum. This is fairly challenging in conventional solar cells since it would require alloys as absorbers with varying compositions to begin with. Another smart use of QDs is down converters which make use of the size-dependent photoluminescence characteristic. In this process, QDs absorb photons of a shorter wavelength (blue region) and emit them back at a longer wavelength (red region) where the device has high quantum efficiency. This enhances the photo response and efficiency of the cell.
In the early 90s, dye-sensitized solar cells (DSC) made a breakthrough by having devices work with 7% efficiency. Within two years, the efficiencies reached the 10% mark required for a PV device to be considered viable. Till now the DSCs work with an accepted efficiency of 11% and have made their way into PV markets. Lower photocurrent densities (~21mAcm-2) chiefly explains why they are lagging behind silicon (42.7mAcm-2) and thin film solar cells (26-35mAcm-2). Some of the best performing dyes have lower absorption than QDs by at least one order. Stronger light absorbers inevitably make better candidates to replace these dyes. Again, most dyes do not have response beyond the visible region, leaving the entire infrared part of the spectrum (49% of total solar radiation) unutilized. Tapping into that energy would boost photocurrents. QD semiconductors with band gap energies in the range of 0.5eV-1.6eV could perform this trick.

The potential problems

Despite these advantages, some of the best laboratory efficiencies of quantum dot-sensitized solar cells (QDSC) are around 4-5%, approximately half that of a DSC. One common argument presented is that there has been two decades’ worth of research on DSC compared to the emerging QD-sensitized cells. That, of course means detailed material and device characterization has been done for DSC which will help shed light on where QDSCs are falling short.
The first problem arises with QD materials replacing dyes in the system. Unlike first and second generation solar cells, dye or QD-based solar cells work on different principles. Light absorbers – commonly referred to as sensitizers – are at the heart of the device which helps to convert photons to electrons. These sensitizers need hole-transporting materials (HTM) like liquid or solid state electrolytes to regenerate the sensitizers as well as to collect the electrons at the counter electrode. Typically-used iodine/iodide electrolyte for DSCs is corrosive to the QD semiconductor. Researchers came up with alternative liquid electrolytes amongst which polysulphide liquid electrolyte proved to be a fairly good HTM for QDSCs, “fairly good” here referring to picoseconds hole injections from QDs to electrolytes. A new problem arose due to polysulphidenot having good catalytic activity with platinum, the doping material for DSC counter electrodes. Low catalytic activity has the undesired effect of high internal series resistance. Until now, copper (I) sulphide on brass counter electrode and polysulphide electrolyte have been accepted staples for QDSC. This makes DSC and QDSC devices similar in architecture but different in material selection. This incompatibility has dealt a blow to the possibility of dyes and QDs being used together as co-absorbers in one system.
The other problem arises from the deposition method. QDs can be fabricated and deposited either simultaneously or separately. In situ, methods like successive ionic layered absorption and reaction (SILAR), chemical bath deposition (CBD) and electrodeposition let the QDs grow directly onto the porous substrates. Ex situ, QDs are fabricated separately and deposited onto the substrate with a linker. In situ methods are the preferred techniques for better electron transport and higher QD absorption. For the cadmium sulphide deposition via SILAR method, only around 20% of the available titanium dioxide porous film surface area is covered. The QD particles get deposited in the early stages of fabrication and later cycles only add to those particle growths, ceasing increased surface coverage. Concentration of QDs not only limits the photocurrents but also, having a large surface area uncovered leaves a higher chance of electron recombination with the electrolyte. This effectively decreases the internal shunt resistance of QDSCs resulting in low fill factors (40%-60%) compared to that of DSCs (~70%). Recently electron recombination via surface defects in the QDs itself has been identified as a viable pathway for electron recombination. This is exclusive to QDSCs and raises the question of the limitations of deposition methods. Finally, most QD semiconductors used for QDSCs are heavy-metal-based compounds making them unlikely candidates for green energy.
Recent reviews for QD-sensitized solar cells have focused on narrowing down the performance-limiting factors and possible solutions. Photocurrents achieved for cadmium sulphide/cadmium selenide-sensitized cells are surprisingly close (~17mAcm-2) to those obtained for DSCs. Concentration of the deposited cadmium selenide is higher than that of cadmium sulphide for the same deposition cycles. This shows that use of dual sensitizers to some extent bypasses the surface coverage issue of QDSCs. Proper surface treatments and the right protective layers would reduce electron recombination and stabilize the QDs. That would contribute to the much-needed optimization of internal resistances. One of the most talked-about concepts is the generation of multiple electron-hole pairs (EHP) from one photon. Although this has been proven experimentally for QDs, it seems a little far-fetched for QDSCs. This process requires at least twice the amount of energy for particles to be generating more than one EHP per photon. The energy limit of our solar spectra to around 3.5eV narrows down the QD materials available to be implemented experimentally. Recent research into plasmonic metals could bring in a fresh perspective for improving QDSC performance. Metals like copper, gold and silver have response to visible light which triggers a highly concentrated electric field around these particles. First and second generation solar cells depend solely on electric field for charge separation. Existence of electric fields inside the absorption layer would improve separation of EHP. Incorporation of plasmonic metals with QDs is relatively new and has the potential to build better devices.
Based on the current state of research, the Achilles heel for QDSCs has not been identified yet. Some of the initial hype about quantum dots replacing dyes was overshadowed by limited device performance. There have also been investigations of silicon QDs as down converters and in tandem-cell application. While there have not yet been any tandem cells based on QDs, the highest success for down conversion technique was a 10% efficiency improvement for polycrystalline silicon cells. Coming back to QDSCs, accounting for the fabrication limitations together with correct and optimized device configuration is needed as a first step to reach a credible performance. As a secondary step, the introduction of back reflectors and a light scattering layer, further optimization with down conversion – incorporating plasmonic metals or even the generation of multiple EHPs, can push the efficiency limit even higher. Things can pick up for QDSCs only if the optimized efficiency is reasonably high, which would make it cost effective to go for the additional performance enhancing techniques. Reflecting on the current volume of research, the question as to whether or not QD-based devices will reach commercial production level efficiency is likely to be answered soon.