Beginner’s guide to kesterite (CZTS) solar cells

Life was comfortable back in 2007. While starting to work on quaternary semiconductors for thin-film solar cells, there was very little literature. It was possible to read all of the papers in the field. Since the report of a 10% efficient solar cell made from Cu2ZnSn(S,Se)4 in 2010, interest in the field exploded and there now stands over 1000 publications. I feel sorry for any new graduate student beginning a project…

CZTSFrom a Web of Science search for “Kesterite OR CZTS solar cell” with 1155 results – 28th May 2015

From reading quite a few of these papers and attending conferences and workshops over the years, here is a decent reading list to get started:

1. Development of CZTS-based thin film solar cells (Thin Solid Films, 2009) – An important historical overview of the development of the field. Like many technologies, it all started in Japan.

2. New routes to sustainable photovoltaics: evaluation of Cu2ZnSnS4 as an alternative absorber material (Physica Status Solidi B, 2008) – An important paper from Jonathan Scragg (whose PhD thesis turned into the first book on kesterite solar cells) with layers made by electrodeposition.

3. The crystal structure of kesterite type compounds: A neutron and X-ray diffraction study (Solar Energy Materials and Solar Cells, 2011) – X-ray diffraction has trouble distinguishing between Cu and Zn. Neutron diffraction confirms the ground-state crystal structure (not stannite) and the tendency for cation disorder.

4. Kesterite Thin-Film Solar Cells: Advances in Materials Modelling of Cu2ZnSnS4 (Advanced Energy Materials, 2012) – The complexity of these materials has provided a fertile ground for theory and simulation, with early efforts reviewed here on structure, defects and band energies.

5. 8.6% Efficient CZTSSe Solar Cells Sprayed from Water–Ethanol CZTS Colloidal Solutions (Journal of Physical Chemistry Letters, 2014) – simple, clean and easy to scale up, with more recent reports of reproducible 10% efficiency from this approach.

6. Device Characteristics of CZTSSe Thin-Film Solar Cells with 12.6% Efficiency (Advanced Energy Materials, 2014) – The current record device with 12.6% efficiency. The one to beat!

7. Influence of compositionally induced defects on the vibrational properties of device grade CuZnSnSe absorbers for kesterite based solar cells (Applied Physics Letters, 2015) – Precision Raman spectroscopy is becoming increasingly useful for identifying secondary phases and quantifying structural disorder in kesterites. The team at IREC are leading the way.

8. Suns-VOC characteristics of high performance kesterite solar cells (Journal of Applied Physics, 2014) – What is limiting performance to less than 20%? This is one of several important detailed charactertisation papers that point to issues with the back contact.


They are all naked (how to survive scientific presentations)

I prefer writing to speaking, but from time to time the latter is necessary. When I quite young, I joined a speech and drama club that included mime and poetry. I still remember being on stage at Father Mathew Hall (Dublin City) for a competition. There was no fear in an eight year old me.

That confidence was lost, and I clearly recall my final year presentation as a Chemistry undergraduate, with quivering voice and shaking hands. As a postgraduate, my first conference presentation was at Oxford. I made the mistake of sitting at the back of the room. It was a long and terrifying walk to the front of the auditorium, but I survived. I knew that first talk by heart (and probably sounded like a nervous robot).

Of course presentations play an important role in science, so I had to work at it. I read quite a few books on public speaking (from the dreaded self-help section of the book shop). There were some good general speaker tips:

  • Visit the room early to get used to the presenter view
  • Speak slowly and clearly
  • Face the audience as much as possible
  • Try not to walk around
  • Avoid repetitive words or hand movements
  • Always bring your own pointer and laptop
  • Practice

and some general content tips:

  • Tell a story: a beginning, middle, and end
  • Have a single message per slide
  • Know your audience (tailor the content for them)
  • Use the minimal amount of text
  • Simple is best (no fancy transitions)
  • Large fonts (for the people at the back)

No matter how much you read, it can be difficult to overcome your default behavior and reactions in the moment. Standing in front of your peers is not easy – I have seen senior professors from top universities being as nervous as the first year students. I find it depends as much on the event itself (room layout and setup, audience interest and atmosphere). The worst case scenario is a half empty room of disinterested people with a microphone that keeps clipping. So I find that even with a similar set of slides, a talk can be a disaster, okay or great.

Despite all of the words above, it really comes down to the results. If you have something interesting to present, it doesn’t matter how you do it. One of the most memorable talks I have seen is a student who stood up and delivered a 15 minute monologue (no slides or data). Rules can always be broken.

Hybrid Perovskites Go Bananas?

There have been discussions regarding hysteresis in the performance of hybrid halide perovskite solar cells since the MRS Fall Meeting in 2013 (a brave presentation from the group of Mike McGehee and supplementary slides from the presentation of Henry Snaith). Since then, there has been a flurry of papers reporting and attempting to characterise the behaviour (see a news piece in Chemistry World this week).

A related phenomenon is the low frequency dielectric dispersion of these materials (mentioned in my recent stream of consciousness), where large polarisation features emerge due to build up of charge (e.g. see the Maxwell-Wagner effect).

This effect reminded me of some arugments in the literature several years ago regarding the characterisation of ferroelectric materials (from Bananas go Paralectric to Ferroelectrics go Bananas). The response observed for a banana is remarkably similar to the “giant dielectric effect” reported for the (inedible) hybrid halide perovskites. Quite a chunk of literature can be rationalised through this anology: “With simple experiments, the response of a banana to electric fields is revealed as characteristic for an inhomogeneous paraelectric ion conductor.”

Food for thought…


One chain does not make a crystal

Having time to read was one of the few perks of feeling ill over the Christmas holidays. I had a chance to finish the 1964 classic text by J. M. Ziman, Principles of the Theory of Solids. He writes with a certain eloquence and authority that you don’t often find in physical science. It is not a text for true beginners, but addresses many fundamental concepts in condensed-matter physics with a unique perspective and a clear narrative.

A few choice quotes to stimulate the mind:

Lattice periodicity and k-points:
“There are exactly as many allowed wave-vectors in a Brillouin zone as there are unit cells in a block of crystal.”

Electrostatic summations of ionic solids:
“The problem of computing such a sum becomes very serious. The best direct method is that of Evjen, where one treats successive shells, going outward from the origin, each one being exactly neutral in charge.”

Special points in an electronic band structure:
“For a given amount of computing, we can get more accurate values of E(k) at points of high symmetry, than we can at an arbitrary point in the [Brillouin] zone.”

Ionic semiconductors:
“It is doubtful whether a hole can move like a free particle in an ionic crystal. The small overlap between valence orbitals on neighbouring ions implies a very narrow valence band with a correspondingly high effective mass.”
Carrier mobility in semiconductors:
“The scattering of carriers by lattice vibrations in semiconductors is, in general, a much simpler problem. Because the carriers are usually thought of as concentrated in a small region of k-space, near a minimum in E(k), the possible change of k-vector in the scattering is small.”
Ising model:
“The importance of the Ising model is not, however, in the description of particular physics effects; it is a mathematically tractable model of a system that should exhibit co-operative phenomena and phase transitions.”

31557600 seconds of work

For research, 2014 has been an extremely fun year. There have been many new projects going in unexpected directions, which keeps things fresh and interesting. My research group composition has been changing too (Out: Lee to Kytoto, Davide to Oxford and Rachel to Queen Mary; In: Suzy from York, Ruoxi from Fudan, Katrine from Aarhus), which alters the dynamic. There is no such thing as a quiet or normal week.

Publications from 2014:

We have done better for fully open access (OA) publications this year, but still room for improvement. Generally, we don’t pay for gold open access with the American Physical Society (e.g. Physical Review B or Physical Review Letters) because they have the most generous policy for hosting on personal websites and institutional databases.

Too many papers, too little time

[fade to black and white] I remember hunting down a series of 1950s papers on sterochemical lone pairs by L. E. Orgel at the start of my PhD. There was the wonderful satisfaction of finding the right volume of the journal, photocopying the paper, and then curling up in the corner of the library basement to read it. If I started this year, a quick web search just sends me to the right place.

Immediate access to information is useful, but it makes it increasingly difficult to navigate the expanding literature. Even in my general area of computational materials chemistry, there are too many journals, papers and authors to keep track of. My current workflow involves the following web services:

  • Google Scholar. The commercial Web of Science and Scopus search engines are quickly becoming redundant. Google is faster and more effective. There are some nice features such as direct export to BibTeX, and access to pdfs that you may not have subscriptions to (e.g. stored on personal websites or online databases). There is also a surprisingly accurate alert system, which gives you recommended reading based on the papers you have published and cited.


  • Mendeley. On one hand, Mendeley is useful for sharing papers. I use it for maintaining a list of publications in the emerging field of hybrid perovskites, for keeping track of our journal club, and an essential reading list for new students. The desktop client is also very useful for synching pdfs across machines (including notes and annotations), and for maintaining a bibliography for LaTeX or Word documents. The cite-as-you-write feature has now made Endnote  redundant (which has always been a clunky and error prone piece of software). Mendeley is particularly smart at importing missing database entries when you edit a collaborator’s document.
  • Old Reader. Since the early death of Google Reader, I tried out many options for tracking RSS feeds from journals in my field. Eventually, I settled with the Old Reader. As the name suggests it maintains the functionality of Google Reader. It is fast and displays TOC art quite nicely. The alternative is weekly alert emails from journals, but I enjoy my morning coffee browsing through the new articles of the day (caffeine and new science are equally addictive).

Help I can’t swim

The absence of regular postings isn’t due to a lack of things to say, simply a lack of free time.

Time management is probably my biggest challenge these days. I have a dozen wonderful group members, doing great research. It is a full time job just to keep up with them, and then I have to find time for my own research, presentations and the bane of all academics… bureaucracy. I am not complaining, I have never enjoyed science as much as I do now.

My research group is now focused on three areas: photovoltaic materials; metal-organic frameworks, and metastable states. Some topics, such as hybrid halide perovskites, are bridging all three themes due to the complexity of their chemistry and physics. A major driving force for our current work is temperature dependent properties (see Jonathan Skelton’s pro-tip guide for Phonopy) and disorder. When I have some free time, I can be found reading some dusty statistical mechanics (one essential read) or thermal physics texts (don’t tell my chemistry colleagues).

Time now to prepare for another trip to my second home (South Korea). This time there is a workshop between Yonsei University and the University of Bath to expand the range and depth of collaborations, followed by an exciting Royal Society collaboration with the Korean Institutes for Basic Science (to be held at Seoul National University). No doubt the next few days will pose exciting adventures in culture, food and functional materials.