Growing crystals for your PhD – The trials and tribulations Part 2

In yesterday's post I reminisced about a not-so-successful crystal growth of the defect perovskite material Sr_0.8Ti_0.6Nb_0.4O₃. However, my crystal growing adventures did not end there! I had begun a new crystal growth which would prove to be much more successful. In fact, it was to be my first crystal. The material in question was Sr₃TiNb₄O₁₅.

Sr₃TiNb₄O₁₅ and Sr_0.8Ti_0.6Nb_0.4O₃ are actually very closely related. In fact they are a part of the same series derived from SrTiO₃. If substitution of niobium for titanium is continued beyond the Sr_0.8Ti_0.6Nb_0.4O₃ composition, the defect perovskite structure is no longer stable and a tungsten bronze type structure is formed instead. The exact composition at which a pure tungsten bronze type structure resulted was Sr_0.6Ti_0.2Nb_0.8O₃, which corresponds to Sr₃TiNb₄O₁₅.[1]

Figure 1: Crystal structure of Sr₃TiNb₄O₁₅ generated in VESTA. Structural information can be found in reference 1.

Initially, Sr₃TiNb₄O₁₅ was an impurity to me. It was something I was trying to avoid in order to synthesise a highly defect perovskite structure. However, Sr₃TiNb₄O₁₅ itself is interesting as a relaxor ferroelectric material. Traditional ferroelectric materials exhibit a permanent reversible dipole moment which makes them extremely useful in capacitors, non-volatile data storage and other applications. Relaxor ferroelectrics differ slightly compared with traditional ferroelectric materials due to their complex compositions. Rather than exhibiting a strong ferroelectric response at a specific temperature (like traditional ferroelectrics), the response is spread out over a range of temperatures. This means that the dielectric constant of the material will be very close to its maximum for a wide range of temperatures, making it much more useful in applications.

Much like traditional ferroelectrics, the crystal symmetry plays a large role in determining the properties of the material. Specifically, the crystal symmetry can tell us whether a permanent dipole moment is possible within a structure. The best way to determine the structure of a crystal is diffraction from a single crystal! So once more I set out to grow crystals using the floating zone furnace.

Contrary to my rather lofty plans for Sr_0.8Ti_0.6Nb_0.4O₃, I did not require an especially huge crystal. Anything sub-millimetre scale would have been great. However, in comparison with the defect perovskite, a crystal of Sr₃TiNb₄O₁₅ was really quite easy to grow. The main difficulty encountered was that the crystal growth was a little bit temperamental and unless it was "baby-sitted" it would tend to destabilise after about an hour. So I spent a nice quiet afternoon in the basement of the chemistry building slowly growing my crystal. Every now and then I would adjust some growth parameters to ensure that the melt wouldn't destabilise. Eventually I obtained a lovely transparent yellow crystal.

Figure 2: Close up of the tungsten bronze crystal.

Although the resulting column was comprised of multiple crystals, the crystal fragments were big enough for both X-ray and neutron diffraction experiments (however I was very hesitant to break up my crystal to do these!). While the work to determine the exact symmetry of the structure has now been passed down to another student, I will always be proud of my first ever successful crystal growth.

[1] T. A. Whittle, W. R. Brant & S. Schmid (2013). S. Schmid, R. L. Withers, R. Lifshitz (Eds.), Aperiodic Crystals, Proceedings of the 7th Conference on Aperiodic Crystals. Dordrecht: Springer. pp. 179 – 186.

Tags: floating zone   crystal growing   inorganic