Co-authored by Rodney Pelzel, CTO, IQE
INTRODUCTION
For most of us, it is impossible to imagine a life without digital technology; devices have become indispensable to every part of our lives, shaping how we work, play, eat, sleep, learn, even how we choose a soul mate.
Device technology is underpinned by the now ubiquitous semiconductor. Over the past 40 years, a single semiconductor material, silicon, has been dominant in what we therefore know as the silicon chip.
However, even this extraordinary material has its limits, and to satisfy the needs of the devices of the 21st century, new materials are required. We are at a technology crossroads where the materials foundation of the semiconductor industry must be upgraded.
Fortunately, there is a solution.
This upgrade comes in the form of Compound Semiconductors (CSs). These materials have characteristics which can “turbo-charge” the capabilities of our devices when combined with silicon-based semiconductors in a process known as heterointegration. New CS-powered devices will enable the Metaverse, upgrade wearables, simplify 5G, enhance autonomous vehicles, continue to power data communications and through much improved power usage, help us on the path to Net Zero.
The global need for CS materials is great news for the UK, since it already leads in development and manufacture of these materials and has the opportunity now to cement, consolidate and extend its leadership. This is particularly welcome as the dominance of Asia, primarily Taiwan and South Korea, in silicon-based semiconductors is unlikely to change any time soon. Therefore, leadership in CS technology (which enables heterointegration) is a compelling and timely opportunity for the UK.
If acted upon, this opportunity will strategically position the UK as a vital component of the worldwide semiconductor supply chain. This is, of course, in addition to the UK’s proven leadership in chip design through companies like Arm and Graphcore, and our leading materials research in UK universities. However, for the purposes of this article, we will focus on the less visible world of Compound Semiconductors where our opportunity for growth lies. Timing is critical and the UK needs to act rapidly and decisively to cement its leadership position and reinforce its criticality in the future of semiconductors.
The current geopolitical climate makes the need for investment even more urgent. There is significant talk of a possible intervention in Taiwan by China. The vast majority of semiconductor chips rely on Taiwan, therefore, any disruption or change of control could possibly be catastrophic to the global semiconductor industry. The best way to mitigate this risk is to create a domestic (or at least a “friendly country”) supply chain. The UK’s position in CSs gives it significant bargaining power in the creation of such a supply chain.
COMPOUND SEMICONDUCTORS
As the name implies, Compound Semiconductors are different from silicon in the fact that they consist of compoundsrather than a single element. For the purposes of this article, two types of CSs will be discussed, Group III-V CSs (sometimes called “Three-Fives”) and Group IV CSs.
Three-Fives are chemical compounds made up of one or more materials from Group III of the periodic table (B, Al, Ga and In) combined with one or more materials from Group V of the periodic table (N, P, As, Sb and Bi), cf. Figure 1. Examples of Three-Fives are gallium arsenide (GaAs), indium phosphide (InP) and gallium nitride (GaN).
Group IV CSs are comprised of chemical combinations of Group IV elements from the periodic table (C, Si, Ge and Sn), cf. Figure 1. Silicon carbide (SiC) and silicon germanium (SiGe) are examples of Group IV CSs.
CSs outperform standard silicon in three key ways:
- EFFICIENCY: CS materials are very efficient at high frequencies. Often, this is colloquially stated as CSs being “faster” than silicon. Technically, this refers to the fact that they have a higher electron mobility. As frequency increases, CS materials are much more efficient than standard silicon, which makes them the materials of choice for high power, high frequency operation like those used in advanced mobile handsets. It is the efficiency of CS materials that gives a smartphone a reasonable battery life – without CS materials, the battery life of next generation handsets would be measured in minutes rather than hours or days.
- OPTICALLY ACTIVE: Three-Fives are extremely efficient emitters and detectors of light. This is something that has been exploited in one of the earliest applications of CS materials, LED lighting. It is the optical efficiency of CS materials that has enabled the replacement of a traditional 100W filament bulb with a 7W, more reliable, more flexible LED bulb. Imagine the reduction in power multiplied by all of the lightbulbs in the world. As such, the efficiency inherent to CS materials in this and many other applications, make them an essential and exciting part of the path to Net Zero.
The efficient emission and detection of light by Three-Fives make them essential for facial/gesture recognition and wearable health monitors. Add 3D mapping for AR, VR and ultra-high-resolution displays using µLEDs (micro-LEDs) for Metaverse applications and it is obvious that the optical capability of Three-Fives will be a big part of the future. Three-Fives can be finely tuned (through control of the composition) to emit or detect a specific wavelength of light. It is this characteristic that is the foundation of the optical networks that enable today’s and tomorrow’s high speed tele/data communications in data centres and over fibre optic links.
- POWER HANDLING CAPABILITY: CS materials like GaN and SiC are well suited for power electronic applications from computer power supplies to electric vehicles (EVs) to grid-based voltage conversion. The path to Net Zero for transport is dependent upon the electrification of vehicles, in which CS materials will be vital (for the vehicles as well as the charging infrastructure). Beyond EV’s consider that fact that the current world-wide energy loss due to voltage conversion is twice the amount of energy generated by renewable sources. Such loss can be greatly reduced by replacing silicon-based components with CS devices.
So, it is clear that CS materials provide the foundation for next-gen mobile communications including 5G, advanced sensing (such as facial recognition and LiDAR systems), power electronics (like those in electric vehicles), and ultra-high resolution, miniaturized displays (such as µLEDs for wearable devices and the Metaverse).
It should not be underestimated how complex the manufacture of these CS materials is. For example, the wafers used for some of the CS optical products described above (and manufactured by IQE) are made of around 400 layers, each the thickness of a single atom, which are overlaid in a process called epitaxy. It is the understanding and leadership in such a sophisticated process, combined with the requirements for the new applications, described above, which mean that the time is ripe for the UK to consolidate its place in the global semiconductor supply chain.
CS SUPPLY CHAIN
Having described the criticality of compound semiconductors, how do we cement their position in the global semiconductor supply chain?
Figure 2 shows the steps involved in the creation of an end device which is sometimes in a single company, or vertically integrated. More typically, these steps are distributed in what is referred to as the foundry model, where components such as design are separated from manufacture. Regardless of the model, all of the steps are required to create a viable product.
Using silicon as an example, the foundry model is widely used where fabless companies (those who don’t have their own manufacturing) engage with large specialised foundries who manufacture to the bespoke, proprietary designs of the fabless company. Companies like TSMC, UMC and GlobalFoundries operate extremely successfully with this model.
For silicon, the foundry supply chain is well established. The capital investment required to enter the silicon supply chain is tens of billions of pounds, and for the most part this prevents new players from entering (although some countries are spending significant sums to attempt to compete in this market). The UK has no current penetration in the silicon market, and this is unlikely to change, given investment costs, labour prices, expertise and supply chains. It would therefore seem wise that we focus on our strengths, which still allow us to be a crucial part of the global supply chain and provide us with the security and resilience we will need for the future.
Figure 2: The Semiconductor Supply Chain