CEO interview: Building a graphene industry, one layer at a time
Thomas provided eeNews Europe with a briefing on progress at the startup he co-founded with Professor Sir Colin Humphreys from the University of Cambridge. Thomas also gave some broader thoughts about the potential of graphene, a single atomic layer material that has been hyped to the skies but has yet to deliver on its promise.
First off it should be understood that Paragraf, since its founding in 2015, has raised nearly £20 million (about $26 million) across seed and Series A equity funding rounds and developed its own R&D laboratory allowing it to produce prototype electronic devices based on its graphene deposition capabilities. Thomas said that the company has reached a head-count of fifty people and with a Hall-Effect sensor in use at CERN and being examined elsewhere, it is starting to gain some traction with industry.
The premise behind the company is that its ability to lay down contamination-free graphene without the use of a copper catalyst could place the company at the focus of a wave of graphene-enabled electronics. Paragraf’s contention is that ICs made using graphene-based transistors could be ten times faster than silicon ICs, and that graphene-based chemical and electrical sensors could have sensitivity increased by a factor of more than 30 from what is currently available.
In addition, the ability to lay that graphene down using chemical vapor deposition (CVD) on conventional substrates and silicon wafers opens up the prospect of the economies of scale afforded by silicon semiconductor manufacturing.
The areas of application for graphene are broad from sensors, through transparent electronics – where it can replace expensive indium in indium tin oxide (ITO) – bioelectronics and on to conventional electronic circuits in analog and digital domains. Graphene has very high levels of electrical and thermal conductivity, exceptional electrochemical stability, large surface area and is also suitable for improving such things as solar photovoltaic cell and rapidly rechargeable batteries.
But is Paragraf going to be able do everything from wafers and materials to packaged components in all application areas?
“Right now we do everything in house apart from some packaging. So, we are starting to out-source,” said Thomas. One thing that Paragraf is finding is that while there is strong interest in the use of graphene magnetic sensors but many applications require additional circuitry or specialized robustness, such as high temperature operation. While graphene is capable of high temperature operation it does require specialized packaging. “We’re never going to be specialist in all the different types of packaging,” Thomas said.
But he added that the goal for the company remains to make graphene fundamental to electronics and this could go as far as adding graphene to transistors to improve performance and certainly includes the monolithic integration of graphene sensors and circuits with conventional silicon circuits.
But meanwhile Paragraf has to build the credibility of graphene gradually, on a case-by-case basis.
One reason that Paragraf needs R&D and manufacturing clean rooms is that graphene is a revolutionary material for which a supportive ecosystem has yet to develop fully. So to show the advantages of graphene the company has to do be also to produce exemplars of the capability such as its high-accuracy Hall-Effect sensor (see CERN measures with graphene-based Hall effect sensor).
A second reason, Thomas said, is that graphene as a material has been overhyped in the past and has failed to deliver, creating a black mark against the technology in the minds of some venture capitalists. “Back in 2009 there was talk of the deposition of large-area graphene but the problem was that the technique left the graphene contaminated so then the problem became how to overcome that. There was a lot of investment in 2009/2010 and those VCs lost their money.”
Paragraf’s Hall-Effect sensor is high value and low volume but as different versions of the same technology are deployed and volumes increase is Paragraf going to become a wafer fab? Surely not?
“As volume increases we will look at the best way forward,” said Thomas. When pushed to be more specific he added: “Well some things may be partially licensed but at very large volume it does become a licensing model.” Thomas added that big sensor manufacturers, often fabless, are very attuned to where they are going to get sensors made and packaged and having security of supply.
However, another short-term challenge for Paragraf is that foundry wafer fabs are wary of putting novel materials down their production lines for fear of contaminating the line and impacting other processes and wafers. So it is up to Paragraf to demonstrate that it has full knowledge and control of all of the issues.
“We are working at running devices on three separate pilot lines right now,” said Thomas.
eeNews Europe is aware of such entities as SkyWater Technology Foundry Inc. and IMEC, that specialise in helping customers develop manufacturing processes but there are other foundries that are willing to be more open to novel processes, such as X-Fab and Tower Semiconductor. Thomas declined to name the pilot lines in use but added that a couple of them are in Europe.
Thomas said that Innovate UK, the UK’s innovation agency, does have a project looking into whether it is feasible to build a graphene manufacturing line in the UK. A result of its deliberations is due in January 2021.
As Paragraf’s volumes of graphene component sales increase Thomas said it is important that its R&D capability doesn’t become bogged down servicing immediate demand. “The H-E sensor is now ‘owned’ by a product manager. It is an enabling technology but lots of different customers want to use it in different ways. We have one potential deployment inside a jet engine.”
Next: Four focus areas
Thomas explained that Paragraf has four parallel areas of focus and that as one comes to product status R&D should be focused on the next developmental work and out-sourcing increased on the first. “We want to move rapidly but the only way to do that is make use of other companies’ expertise,” said Thomas. But he also covered that position with a caveat. “Focus is a real issue. If we don’t focus, we could destroy the business.”
And then there is the idea that few battle plans survive first contact with the enemy: for enemy read market. “Yes, the VCs will tell you it is all about grasping the opportunity,” agreed Thomas.
We checked with Thomas about whether Paragraf has any strategic investors in addition to venture capitalists. “Strategic investment generally comes with strings, and strong opinions about what the startup should be doing. So, we have avoided strategic investors so far. We are planning for another round of venture capital next year and we will consider strategic investors then,” said Thomas.
With Thomas’s observation about the over-hyping of graphene, we asked whether government investment in initiatives such as the UK’s Graphene Institute had been well-placed or wasteful.
“The issue isn’t graphene. It’s the age-old problem. The UK is great at funding R&D but we struggle to fund the commercialization in the UK. The Graphene Institute in Manchester has produced brilliant results but turning that into a business is difficult. We don’t provide the tax holidays and other incentives, nationally or from local authorities, which means entrepreneurs have to go straight to venture capitalists with very little to show.”
Next: Graphene flagship
In Europe one big engine of graphene development is Graphene Flagship; a billion-euro European Commission backed project intended to last from 2013 to 2023 and which has more than 145 academic and industrial research groups in 21 countries. But, with the impact of Brexit, it is not necessarily something that will remain open or advantageous to UK companies.
“The main emphasis of Graphene Flagship is to promote collaboration. I think there have to be at least five entities collaborating on each individual sub-project. We are not involved but we are hearing that European partners are starting to be concerned that UK partners cannot be clear about where their funding will come from.”
At a global and industry-wide scale Thomas concludes that graphene electronics is still very young with a lot of development work to be done. But it shows the 10x and higher levels of improvement that tend to be the threshold of making novel technologies worth the pain of introduction. “We think it will be three, four or five years before volume production of graphene electronics,” said Thomas.
It’s not possible to be certain about the application where the graphene breakthrough will be made, but Thomas is determined to overcome scepticism and get graphene electronics production into the mainstream, preferably with a strong local manufacturing base and ecosystem.
The question remains open as to whether that can be in the UK or in Europe or must inevitably be in the Far-East.
Related links and articles: