Posted by 
The rapid technological advancements of our world have been enabled by our capacity to design and fabricate ever smaller electronic chips. These underpin computers, mobile phones and every smart device deployed to date.
One of the many challenges is that electronic components generate increasingly more heat as they are miniaturised. A significant issue lies in making the wires which connect the transistors on the chip thinner while ensuring that the minimum amount of heat is released.
These interconnects are typically made from copper, and as we start to scale them down to nano-scale thicknesses, their electrical resistance increases rapidly because the electrons moving along the wires have a higher probability of colliding into the surface of the wire. Known as scattering, this leads to energy being released in the form of waste heat, meaning you need more power to maintain the same level of performance.
A group at Stanford University has published a new paper showing that thin films of a material known as niobium phosphide (NbP) exhibit much higher conductivity than copper below a thickness of 5 nanometres (nm) (the typical thickness of the wiring in today’s chips is about 10nm-30nm). This improvement is because NbP is a material with unique quantum properties.
NbP is in a class of materials known as topological materials, which have unique electronic properties due to their atomic structure and were the subject of the 2016 Nobel prize in physics. Their conductivity along the surface is extremely high and stays the same regardless of changes to the shape or size of the material (physicists refer to this as a “topologically protected surface state”).
In other words, it is robust and not easily destroyed by reshaping it. Even when the material is made thinner, bent or given ridges, the special conductive properties that can be observed on the surface remain present.
This is important, as the ability to decrease the thickness of the electrical interconnects while still allowing electricity to flow easily is key to balancing the fabrication cost and complexity against energy efficiency. This gives NbP a significant advantage over other materials.
Disorder vs order
What is particularly remarkable about the discovery by the Stanford team is that these quantum properties were observed in disordered films, meaning the niobium phosphide wasn’t fabricated in the most controlled way to maximise its conductive properties.
There’s a good analogy in making bars of chocolate: it’s vital to control the cooling process of melted chocolate to create the glossy end-product that you’re used to. This process is known as tempering. Anyone who has tried this likely knows how dull and soft the untempered version can be – and how it affects the taste.
The difference between tempered and untempered chocolate is the size of the individual crystals within the chocolate’s structure (or to put it another way, the amount of ordering). Untempered chocolate contains multiple small crystals that are disordered with respect to one another. To make tempered chocolate, a manufacturer will cool it in a controlled way and also add a piece of tempered chocolate to promote the growth of a uniform crystal structure.
Now imagine how useful it would be if you could simply skip this tempering. Not only would it make the production process more efficient, it could also decrease the energy required. This is essentially what the Stanford team have shown in relation to their material.
To put it in technical terms, the niobium phosphide had no long-range order. This means that while it might have contained some of the small crystals that you would associate with an optimum version, the material was considered a disordered alloy overall.
Despite this, it was still an excellent conductor for an ultrathin film, seemingly due to the exotic quantum phenomenon of topologically protected surface states. If this effect can occur even in a disordered material, it greatly simplifies the manufacturing process – with positive implications for the costs.
This shows how quantum effects could lead to such significant improvements in conductivity, potentially enabling us to produce more powerful and energy-efficient computer chips in the near future. One important question is whether enough niobium phosphide could be available for manufacturing purposes.
While this is not our area, phosphorus is almost as plentiful as carbon in the Earth’s crust, while niobium is about one third as plentiful as copper, roughly comparable with cobalt and nickel. That suggests sufficient supply, though specialists would have to weigh the relative cost of extracting these elements from their ores.
For those of us working in this field, the discovery also raises the question of whether other topological materials such as tantalum phosphide and tantalum arsenide could behave similarly. Also, what is the essential ingredient that the NbP samples possess for them to act in this way? These questions will continue to engage scientists in their search for solutions to enable lower-power computing technologies of the future.
Mark Greenaway receives funding from the Engineering and Physical Sciences Research Council (EPSRC), UK.
Fasil Dejene receives funding from the Engineering and Physical Sciences Research Council (EPSRC), UK.
Kelly Morrison receives funding from the Engineering and Physical Sciences Research Council (EPSRC), and Royal Society, UK and is an active Member of the Institute of Physics (MInstP).
