Friday, August 26, 2016

What are the worst nightmare materials?

Not all materials are equal. Over the years I have noticed that there are certain materials that are rich, complex, and controversial.

Common problems (opportunities) are that it is extremely hard to control their chemical composition, they may have many competing ground states, tendency to inhomogeneity and instability, structural phase transitions, sensitivity to impurities (especially oxygen and water), and surface and bulk properties can be significantly different. One never knows quite which material system is being measured, regardless of what authors and enthusiasts may claim.

Consequently, these materials can be an abundant source of spurious experimental results leading to endless debates about their validity and possible exotic theoretical interpretation.

Pessimist's view: the material is a minefield for both experimentalists and theorists and with time the "exciting" results will disappear. They are a scientific nightmare. Be skeptical. Avoid.

Optimist's view: this is exciting science and there are promising technological applications. Jump in. With time we will sort out all the details.

Here are some of my candidates for the "best/worst" nightmare materials I have encountered.

Cerium oxides: controlling the stoichiometry is very tricky and chemical and physical properties vary significantly with oxygen content. Yet because of (or in spite of ?!) they have significant industrial applications...

water: polywater, "memory", and the "liquid-liquid" transition in supercooled phase....

1T-TaS2: it undergoes multiple charge density wave transitions as the temperature is lowered, there is "star of David" charge density wave order with a thirteen (!) site unit cell, a Mott insulator transition, superconductivity upon doping, and ultrafast electrical switching behaviour, ...

purple bronze, Li0.9Mo6O17: superconductivity, non-Fermi liquid, large thermopower, ...


What do you think are the "best" nightmare materials?


  1. many (claims of) dilute magnetic semiconducting transition metal oxides... (Co in TiO2 anyone?)

  2. I have dabbled in bismuth chalcogenides for the past few years and learned that the surfaces are not nearly as simple or forgiving as might be inferred from many beautiful ARPES papers on topological surface states. They actually don't always cleave as nicely or predictably as intuition suggests.

    Also, trying to magnetically dope them brings out the same pitfalls of studying weak magnetism that are found in the magnetically-doped oxides that pcs mentioned (how to rule out segregated trace impurity phases, e.g. of Cr2Se3 in Bi2Se3)