Materials known as plastic crystals are composed of molecules that interact through weak long-range forces. As a result, these materials are highly compressible and can be deformed in a reversible manner — hence the adjective ‘plastic’. Under certain pressure and temperature conditions, molecules in plastic crystals can start rotating almost freely around their centres of mass. The centres of mass remain localized at well-defined and ordered positions in the crystal lattice, but the rotation leads to orientational disorder. In a paper in Nature, Li et al.1 report that the phase transition between molecular order and disorder in plastic crystals can be used for cooling purposes through the application of small pressures.
Conventional refrigeration technologies are based on cycles in which greenhouse gases are alternately compressed and expanded. One kilogram of a typical refrigerant gas contributes as much to the greenhouse effect in our planet’s atmosphere as two tonnes of carbon dioxide, which is the equivalent of running a car uninterruptedly for six months (see go.nature.com/2ffbqvt). In addition, current cooling technologies cannot be scaled down to the dimensions of microchips, which hinders the development of faster and more-compact computers and portable electronic devices. There is therefore a pressing need to find eco-friendly and highly scalable cooling methods for improving many crucial technologies, as well as for protecting the environment.
Solid-state cooling is an environmentally friendly, energy-efficient and highly scalable technology that could solve most of the problems associated with current refrigeration methods. It relies on applying cycles of external magnetic, electric or mechanical fields to compounds called caloric materials. These compounds undergo temperature variations as a result of field-induced phase transitions that involve large changes in entropy — a measure of disorder.
Examples of caloric materials include ferroelectrics2, organic–inorganic hybrid perovskites3 and fast-ion conductors4. However, most known caloric materials are not ideal. Some have only modest refrigeration performances, for example, or operate at temperatures different from ambient conditions. And for others, durability and cycling rate can be affected by material fatigue and phase-transition hysteresis — in which the conditions required for completing a phase transition depend on the direction of the transition. Consequently, progress in solid-state cooling has been limited.
Li and colleagues’ work offers exciting prospects for the field of solid-state cooling. The researchers discovered extremely large entropy changes associated with the molecular order–disorder phase transitions that occur in plastic crystals (Fig. 1a). Moreover, they found that these transitions could be triggered near room temperature by applying small pressures (about 10–100 megapascals), and so could be used for refrigeration purposes (Fig. 1b).
The typical field-induced entropy changes measured in archetypal caloric materials are of the order of 10 joules per kilogram per kelvin. By comparison, those found by Li et al. in plastic crystals are of the order of 100 J kg−1 K−1. For instance, the entropy change reported for the representative plastic crystal neopentylglycol near room temperature is about 390 J kg−1 K−1, which leads to a large temperature change (roughly 50 K).
The molecular mechanisms that underlie the colossal pressure-induced entropy changes in plastic crystals can be understood intuitively. When pressure is applied to the disordered (high-entropy) phase, the molecular rotations are geometrically frustrated; that is, competing interactions between the molecules limit their possible orientations. As a result, the ordered (low-entropy) phase is stabilized. The accompanying reduction in entropy is huge — similar in magnitude to the entropies typically associated with the melting of a crystal. Conversely, when pressure is removed, the molecules resume their rotations and the disordered phase is re-established, causing an equally large increase in entropy.
Previous studies have already reported giant temperature changes associated with pressure-driven order–disorder phase transitions in caloric materials. For example, in fast-ion conductors, such effects accompany the transition from a normal phase to a superionic phase, in which the conductivity of ions is extremely high5–7. However, plastic crystals are quite different from other caloric materials, and not just because of their huge entropy changes near room temperature: they are cheap and easy to produce, lightweight, non-toxic and flexible. They therefore seem to be especially well suited for the integration of solid-state cooling in electronic devices and mobile applications.
Nevertheless, plastic crystals are not perfect caloric materials. For instance, given their organic nature, they have relatively low melting points (typically about 300–400 K)8, which is not desirable for refrigeration applications. In addition, the properties that make plastic crystals highly deformable mean that these materials lack the mechanical resilience to endure many refrigeration cycles. Perhaps most importantly, hysteresis and phase-coexistence effects are likely to weaken the cooling performance of plastic crystals. These technical issues will need to be analysed, and solutions found, if we are to pursue the use of plastic crystals in commercial refrigeration.
Li and colleagues propose combining external pressure and electric fields as a way of avoiding the hysteresis problems associated with plastic crystals. A similar strategy has been demonstrated to work well for compounds called magnetocaloric materials, in which mechanical and magnetic fields have been combined to remove unwanted hysteresis effects9.
However, the fact that most molecules in plastic crystals are polar does not guarantee that applying electric fields to these materials will eliminate the effects of hysteresis. The reason is that incredibly large electric fields (about 1,000 kilovolts per centimetre) might be needed to induce any effect on the molecular rotations, as has been shown theoretically for organic–inorganic hybrid perovskites10. A possible solution to this problem might be to find or engineer ferroelectric plastic crystals, in which the ordered phase already exhibits collective polar order11. Despite these challenges, Li and colleagues’ work represents a step towards finding other caloric materials that have advantageous properties.
Nature 567, 470-471 (2019)