If the commercial world is Darwinian, it's not in the sense of being a battle for supremacy. Companies, like species, coexist precisely because they don't always simply compete to be the biggest, but instead often content themselves with finding a niche. So just as the first advice for any entrepreneur would be “don't try to beat Google/Apple/Amazon”, so the equivalent message for advocates of new electronic materials is “don't pretend you're the new silicon.” That lesson has been amply learnt in organic electronics, where semiconducting materials initially predicted to rival the central fabric of microelectronics now instead play to their strengths for applications in optoelectronics, display technologies and interfacing with biology.

Credit: PHILIP BALL

So it is with graphene. The story in which silicon is at the limits of its potential and this new contender waits to be (literally) rolled out as a cheap, robust and superior replacement is all too easily written. But no one in the field now doubts that, at least in the foreseeable future (and its fascinating fundamental physics aside), graphene must find niche applications if it is to justify all the excitement, not to mention all the funding.

But 'niche' doesn't mean small or trivial. A common consensus is that the possibilities of graphene in microelectronics are probably most profound in the burgeoning area of radiofrequency (rf) devices1 — which means mobile devices such as smart phones, tablets and wearable electronics, as well as smart sensors and tags. Here the advantage of graphene as the substrate for charge conduction is its very high carrier mobility, which can be two orders of magnitude greater than that of silicon. This means that graphene field-effect transistors (GFETs) can be switched very fast, operating at the gigahertz speeds demanded in rf applications. Graphene's lack of a bandgap, which makes GFETs poor prospects for digital electronics because they can't be fully switched off2, is no longer a hindrance here: rf electronics, for example to make radio receivers and amplifiers, can be an analog technology for which complete switch-off is not essential.

Much of the pioneering work in developing graphene rf circuits has happened in the IBM laboratories at Yorktown Heights in New York3,4,5. A team there has previously made GFETs from epitaxial graphene on silicon carbide that can operate at up to 100 GHz (ref. 4), and has fashioned such transistors into an integrated circuit that operates as a rf mixer5.

That, however, was merely a proof of concept. In particular, the processing needed to complete the circuit after putting the GFETs in place degraded their performance. That's why an improvement in the manufacturing process now reported by the IBM team should be significant6. By making the fabrication of the GFETs the last stage in the process, they can use conventional methods without harming the delicate carbon films. That not only gives an improvement in gain performance of about four orders of magnitude but makes the whole process fully compatible with industry standards. The circuit acts as a rf receiver, capable of wireless communication at 4.3 GHz. Inevitably, the digital text used to demonstrate successful reception encoded a three-word message: 'I-B-M'.