A different kind of refrigerator

Say your house on the ground-floor is ankle deep in water – a common sight in most Indian cities during the monsoons. So you grab a pail and start throwing the water out in a four-step process:

  1. You dip the empty pail into the water and fill it up
  2. You carry the pail to the drain
  3. You empty the pail
  4. You bring it back to the flooded place

You repeat these four steps over and over until all the water has been removed.

The refrigerator has a similar working principle. Instead of water, there’s heat. The more heat you remove from inside the machine, the more it cools down. And instead of a pail, there’s a fluid called the refrigerant. These are the four steps it follows through to do its job:

  1. You flow the cool refrigerant through a pipe that wraps around the box where the food is kept; it absorbs heat from the box
  2. You pump the refrigerant to a component that will cool it back down by removing the heat
  3. The component does its thing
  4. You pump the heat back to the pipes wrapped around the box

Your air conditioner works the same way, except instead of flowing the refrigerant around the whole room, it interacts with small quantities of air.

This cycle of four steps used with refrigerators and air-conditioners is called the vapour compression cycle. It is a type of heat-pump cycle, which is the broader class of cycles that machines use to move heat from a cooler environment into a warmer environment. (Heat flows naturally from warmer to cooler environments so you don’t need a machine to do that.)

The vapour-compression cycle is employed by many millions of machines around the world – from small household refrigerators to industrial scale warehouses. However, its popularity is slowly declining because the most common refrigerants are environmental pollutants, and their manufacture and use involve processes and other materials that are polluting in their own right.

Scientists and engineers have been looking for more climate-friendly alternatives along different lines of inquiry. Such multiplicity exists because the way heat-pump cycles are conventionally executed is improvident and, in some ways, contrived. There are lots of moving parts, each with its own failings, that interact with each other to give rise to multiple ways in which the system can lose energy. As a result, these machines have low efficiency. This also means there is a lot that can be improved.

One promising class of alternatives is materials that exhibit caloric effects. Typically, this means that when the material is exposed to an external energy field, like a magnetic field, it releases/absorbs heat into/from its surroundings, and vice versa when the field is removed. Materials that respond like this to a magnetic field are said to exhibit the magnetocaloric effect. The element gadolinium is a famous example.

Other types of caloric effects include the electrocaloric effect, the mechanocaloric effect, the barocaloric effect and, of particular interest in the current case, the elastocaloric effect. It is exactly what it sounds like: the external ‘field’ applied to elicit the caloric effect takes the form of mechanical strain. That is, when the material is strained, it heats up; when the strain is released, it cools down.

As with all the other caloric effects, executing the elastocaloric effect doesn’t require multiple parts. The hardware that acts on the refrigerant is the same as the refrigerant itself: the material. And instead of the refrigerant undergoing energy-intensive phase transitions through different states of matter, from liquid to gas and back again, the heat is moved through changes in the way the material’s atoms are arranged.

These are called structural phase transitions. The shape and architecture of the atomic lattice confers different mechanical and electrical properties, among others, on the overall material. The way they are arranged also determines the amount of potential energy contained in the arrangement as a whole. Different structural phases stand for different amounts of energy. So a material that can easily move between two arrangements with different potential energies can be used to absorb and release heat.

Scientists have known of such materials since the early 1980s. The real challenge today is to find a material that matches the efficiency and the lifetime of the vapour compression cycle together with a mechanism that applies the strain as well as possible. In other words, the material should be able to undergo the heat-pump cycle millions of times while being at least as efficient as the vapour compression cycle before it fails. Second, the device used to apply and release the stress should do its job with the least consequence for the overall energy efficiency.

To the end of a material with an appreciable temperature performance, researchers from China, Spain and the US have created an alloy of nickel, magnesium and titanium that exhibits a colossal elastocaloric effect. The advantage here is that this material is greener than the vapour compression cycle. Its underlying structural phase transition is called the martensitic transformation: when a mechanical strain is applied, the atoms slide into a different arrangement, and the material absorbs heat from its surroundings in the process.

An illustration of a martensitic transformation. Source: http://www.materia.coppe.ufrj.br/sarra/artigos/artigo10308

The alloy’s composition on paper is as follows: in every 1,000 atoms, 500 are of nickel, 315 are of manganese and 185 are of titanium; the researchers also added a little bit of boron to improve stability. When they applied 700 MPa of stress to an ingot of the alloy, its volume changed by 2% and warmed by 26.9 K. When they removed the stress, it cooled by 31.5 K. These numbers, the researchers write in their paper, “far [exceed] that directly measured in all elastocaloric, electrocaloric, and barocaloric materials in any form (thin film, wire, bulk, etc.).” The numbers are also nearly equal in value, which means the elastocaloric effect is reversible: the alloy can alternatively gain and lose similar amounts of heat.

Figure (a) shows the temperature differences during loading and unloading at 700 MPa of stress, at 32-37º C, which is room temperature in many places. Source: https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.122.255703

The paper was published on June 26, 2019.

In effect, the elastocaloric four-step heat-pump cycle would go like this:

  1. Expose the alloy to the volume to be cooled
  2. Apply the stress so the material absorbs heat from the volume
  3. Drain the heat from the material
  4. Re-expose it to the volume to be cooled

The researchers have also worked out a way to figure which materials can exhibit such a large elastocaloric effect, or in fact just large caloric effects. The secret is a combination of three factors, all of which depend on the atomic arrangement. First, the material has to be ferroelastic: when mechanical stress is applied, the configuration of atoms needs to change spontaneously. Second: it needs to have “good mechanical properties” (quoted from the paper).

The third factor depends on the unit cell volume. The unit cell is the smallest repeating unit of the arrangement. In the martensitic transformation shown above, it is one square in the undeformed grid; in an actual material, it would be a cube. According to the researchers, the more the volume of the unit cell changes during the martensitic transformation the better.

So by maximising the contribution of each of these three factors, and then determining what the material composition would have to be for all of them to occur together, the researchers have given their peers some new tools to use to uncover other materials that display giant caloric effects. This way, they hope, new materials can be discovered to build the perfect elastocaloric refrigerator or air-conditioner. It is just another way to make our world a better place, though one you probably haven’t heard of.