
Farrier at work forging a horseshoe. By Stefan Ott, Shuttlestock
In this article for The Crucible, analysts Andrew Matheson and Patrick Stratton of CPM Group home in on Hafnium, that most minor of minor metals. Don’t judge a market by it size. “For want of a horseshoe nail, a battle was lost”, as they say.
Most people have heard this saying. It’s been around for a long time and variations exist in numerous languages. It boils down to paying attention to the little things, which are things large and small industry, big and small mining, and big and small economists sometimes fail to do.
As an example, a friend of one of the authors (going back a few years) ran a starch business for a private equity owner. Some of this starch found its way into tyre manufacturing, and the company was the only actual supplier of the product, via various channels, to some of the major global tyre makers. So, when the plant accidentally burnt down, several large OEMs were more than somewhat inconvenienced. They had been buying from several intermediaries and had never troubled to understand the whole upstream end of their supply chains.
The world has woken up in the past three years to the perils of opaque or concentrated supply chains, and of taking for granted that raw materials, components, labor, and other units of production would ‘always’ be there. Amongst other things, the triple assault of Covid, international political tensions, and maritime mishaps clarified for many firms and governments the need to understand in far greater detail the supply chains they rely on, and then to look for ways to diversify risk across a range of factors beyond just cost.
There is still much work to be done, and there are still industries where upstream inputs are frequently seen as a given and not as a strategic risk. The poster child for this kind of thinking, exemplifying the decoupling of application and supply awareness, is the electronics industry.
There may be some logic to not caring about material inputs, in that the electronics industry as a rule can pay premium prices for any commodity it needs. At the same time, however, designing-in truly scarce or hard-to-source materials is a dangerous gambit. Witness the case of hafnium.
Hafnium is among the most minor of minor metals. What sets hafnium apart is that historically the only reason to produce it has been in order to supply zirconium to the nuclear industry (see box). Most zirconium ores contain about 2% hafnium, hence low hafnium (“de-hafniated”) zirconium used in nuclear fuel rods provides about 20 kg of hafnium metal per tonne of nuclear-grade zirconium. In practice, “de-hafniated” means less than 100 ppm hafnium in the zirconium.
Hafnium is supplied to the electronics industry as a chloride and there are today only four real sources: Russia, China, France, and the United States. The first of these is currently off-limits and the second is politically problematic in the eyes of many other countries’ industrial hafnium users and governments. This leaves supply from Framatome, in France, and ATI, in the Unites States. Clearly, there is a concentration issue in the supply of hafnium.
Hafnium has been used for a couple of decades as a high-k dielectric in semiconductor fabrication. “High-k” means a high dielectric constant, that in turn means a material is able, in a short distance, to attenuate an electrical signal. As feature sizes have continued to shrink in semiconductors, high-k dielectrics have become of increasing importance in memory and logic chips. Hafnium, amongst the most stable oxides (on a par with yttrium and the most-stable rare earths), has been used since the early 2000s by companies such as Intel.
Of late, however, hafnium demand in semiconductors has taken off. Reportedly as much as 30 t/yr of hafnium is currently being consumed in semiconductors across a number of nodes, and in memory as well as logic devices. This has been a huge and destabilising demand increase that the market has struggled to manage, with prices in late December 2022 close to US$4,000/kg for metal, versus around US$400/kg a year ago.
Other major uses for hafnium include (mainly land-based) gas turbines – hafnium enables directionally solidified alloy grains to adhere at the grain boundary – as well as nuclear reactors (as control rods, where its high neutron capture cross-section is an asset), Ziegler-Natta polymer catalysts (where hafnium competes, curiously enough, with zirconium), and in welding tips. None of these segments is especially price sensitive, yet few areas of demand can sustain a 10-fold increase in critical material cost without some blow-back. We anticipate superalloy producers are already hard at work on alternatives.
At the moment strange things are (reportedly) happening. We are told ATI is running its zirconium-hafnium circuit to extract hafnium chloride (and is stockpiling the zirconium, likely as a chloride) while Framatome reportedly is struggling to produce an adequately pure hafnium chloride to meet electronics specifications. The fact that there are no obvious large uses outside nuclear fuel rods for zirconium is at once inconvenient, and an opportunity. How long this can continue is hard to know, although one can reasonably judge that as zirconium demand returns, ATI should be well-placed.
There is a broader message at work here. It is impossible to assess today where the global electronics industry will end up having made a bet on hafnium. Bets on other minor metals (think tungsten and tantalum) have worked well so far. That being said, hafnium is an unusual challenge in that it is solely a by-product (of the nuclear power industry). In fact, it is striking (and surprising) that the electronics industry could have staked its collective future on such a marginal metal.
The moral of this tale is that industrial commodities, especially minor metals, do not grow on trees, so to speak. OEMs would be well-advised to work with their suppliers of critical materials, before committing to choices whose risks are not understood.
By Andrew Matheson
and Patrick Stratton (contributing author)
About the authors
CPM Group is an independent commodities research, consulting, and investment banking advisory company headquartered in New York. The company is considered the foremost authority on markets for precious metals, along with manganese and molybdenum. Its entry into tantalum research came through a new collaboration with Andrew Matheson and Patrick Stratton, who are both recognised experts in the tantalum market.
Andrew Matheson, the founder and principal of OnG Commodities LLC, has 25 years of experience in the tantalum industry, leading Cabot Corporation’s tantalum ore procurement and mineral development activities, as general manager of Cabot’s sputtering target business and serving as director of R&D. His experience includes a range of other specialty materials including niobium, scandium and rare earth metals.
Patrick Stratton spent 16 years with Roskill, where he led the tantalum research. His experience also covers niobium, gallium, magnesium metal and titanium. In addition to being the lead author of published research reports on these commodities, he has also undertaken many consulting assignments for producers, project developers, financial institutions and government bodies.