Great question! To my knowledge, the Purdue Solar Lab holds the nanocrystalline based solar cell world record efficiencies at 9.2% and 14.3% for CZTSSe and CIGSSe based devices respectively. We also hold the record for highest CZTGeSSe nanocrystalline based devices, which is also above 9% conversion efficiency. Apart from better materials processing, there are several ways to improve our devices. For CZTSSe, higher sulfur content would help expand the band gap and increase the open circuit voltage. Some groups are looking into extrinsic dopants such as sodium for improving device performance as well. Our group has a strong focus on electrical characterization; by ascertaining the flaws of our devices, we hope to overcome any efficiency barriers for our materials.

To add to Kevin’s comments, the holy grail of many thin-film film technologies is determining the fundamental cause of the relatively low open-circuit voltage (Voc) compared to the devices’ built-in voltage (Vbi) or bandgap. This inherent limitation has vexed many reserach groups including our own. Through materials characterization and device modeling/simulation we are working to understand this limitation and improve the materials synthesis and fabrication to eliminate it.

While we have researchers working towards synthesizing novel photovoltaic materials, many members are focusing on CIGSSe and CZTSSe at Purdue and CdTe at UTEP. CIGSSe is an attractive material because it is commercially produced and has surpassed 20% power conversion efficiency on a lab scale. However; this efficiency was achieved using physical deposition processes. Our study on CIGSSe with a nanocrystalline approach has the opportunity to produce the same thin film for a much lower cost. As for CZTSSe, the use of earth-abundant materials over indium and gallium have made it an attractive alternative. CdTe is another high efficiency material that is commercially produced; however, we believe higher efficiencies can be achieved by tuning the device structure on a nanoscale. The final grain size, composition and structure of the solar absorber material are some of the greatest factors in dictating the final solar performance. By using gradient concentrations of gallium through CIGS, researchers have been able to control the bandgap through the material and improve performance.

To comment on phase space probing, our group at Purdue is divided into three subgroups (new materials, CIGSSe, and CZTSSe). The new materials group works under the hypothesis that certain families of crystal structure result in good solar performance. We then take a structure, look at the periodic table, and imagine what substitutions might be made. We then consider how such substitutions should theoretically affect the bandgap of the material and position of the conduction and valence bands. If an attractive potential material is identified, we try to make it and characterize its properties. In this manner new absorber materials can be found.

An added level of difficulty is then pairing the material with an appropriate buffer layer to form the device. CdS has been the default for many years, but may not have the required band alignment for new materials. Thus, the hunt begins for an appropriate buffer layer.

To add to the last question about the post-sintered material, a common trend we observe is that device performance increases as the sintered grain size increases. Thus, the sintering conditions are chosen to promote large grains (>1 micron), while limiting unwanted reactions at the back contact of the device (the formation of MoSe2) and the loss of volatile material from the film surface (such as SnSe2).

Great question. The most interesting fundamental problem is with the kesterites; CZTSSe has an incredibly low open circuit voltage (around 0.42 V) for its bandgap (around 1.05 eV). All the kesterites suffer from very low open circuit voltages. Whether this dearth of performance is due to the absorber material or due to the overall solar diode architecture is still debated. We employ a plethora of electrical characterization techniques coupled with device simulations that are helping us elucidate why this material is falling short of expectations.

Indeed, the low Voc (compared to the bandgap) reported for CZTSSe throughout literature is arguably the most interesting fundamental problem we are working to tackle in our group. This problem is particularly interesting since CIGSSe – which is in many ways very similar to CZTSSe – does not generally exhibit as detrimental a Voc deficiency. The characterization techniques Kevin mentioned include:

- temperature-dependent current-voltage measurements, which provide insight into advanced diode characteristics such as the ideality factor, dark saturation current, and series and parallel resistances (given an assumed equivalent circuit model);
- temperature-dependent capacitance-voltage and capacitance-frequency measurements, from which we can calculate energies and densities associated with defects and the free carriers in our absorber materials;
- optoelectronic measurements such as time-resolved photoluminescence and temperature-dependent, steady-state emission photoluminescence, which provide carrier lifetimes and information about the dominant radiative recombination mechanisms present in our absorbers.

Many issues we see for CZTSSe were reported for CIGSSe when it was a “young” material, but have since been resolved to a large degree. By comparing measurements taken for both CIGSSe and CZTSSe devices, we hope to learn about what limits our CZTSSe cells. Once we understand the fundamental factors inhibiting device performance, we can modify our fabrication processes to overcome these limitations and improve our solar cell efficiencies.

If I have learned one thing from working with solar cells, it is that there are an infinite number of ways to make a device perform worse. Each step is crucial towards the final performance. All the way from the glass you use, to the grids you lay, everything you do is important. This process takes roughly 3 days start to finish, and the best results are when the full fabrication is done quickly.

From my earlier studies, I have learned that the environment in which you dry a coated film, and the medium in which you disperse your nanoparticles to form the ink, are crucial towards the final device performance.

I’ve learned that by varying the sulfur:selenium content, we can tune the bandgap of our materials and increase the open circuit voltage. Chuck Hage’s research has also shown that other elements like germanium effectively adjust the band gap as well. His CZTGeSSe devices are our newest technology.

We still have a long way to go to achieve our goal of 15% PCE for kesterite materials, but with our approach of focusing on the physics of these materials, we hope to pinpoint the fundamental flaws in CZTS so we can rectify them.

Thank you. We suspend our nanocrystals in 1-hexanethiol for doctor blade coatings. The balance is in finding a solvent that will suspend the particles and uniformly wet the substrate. On the other hand, spray coatings are often done with toluene, though other solvents may be used.

The future is very promising. Several techonologies exist for mass production of thin films, but a key constraint will be the efficiency of module scale films, as most work with CZTS-based materials is currently on a lab scale. Scaleup usually takes a toll on device efficiency, but we are confident this can be overcome as it has to a large extent with other PV technologies.

Thanks! After their AMA on Reddit, I was pretty sold on using their music, and Helix just fit so well with what I wanted to do.

The primary advantage of the nanoparticle approach is the ability to make printable inks. For the CIGSSe material system, nanocrystal-based methods fail to reach the record efficiencies achieved via vacuum deposition processes (~15% vs. 20%). Currently, the world record CZTSSe device efficiency (11%) belongs to a solution-processed solar cell; our champion nanocrystal-based CZTSSe efficiency of 9.2% is comparable to the best values reported for vacuum-deposited CZTSSe cells. If there is a slight performance advantage to nanocrystal-based or solution-processed CZTSSe devices over vacuum-deposited cells, it may be due to marginally better control of compositional uniformity throughout the CZTSSe film afforded by these methods. This advantage could be irrelevant for CIGSSe, a ternary material system, since ensuring compositional uniformity is arguably less difficult for CIGSSe than for quaternary CZTSSe.