Why Manganese-Gallium Is Not The Next Neodymium-Iron-Boron

by Gareth Hatch on December 20, 2011 · 22 comments

in Gallium, News Analysis, Permanent Magnets, Rare Earths

Bookmark and Share Print

I’ve received a number of emails today from people wanting to hear my thoughts on a news release from Northeastern University published earlier today, pertaining to a new magnetic material that researchers at the University have apparently discovered.

According to the announcement, the “super-strong magnetic material” may “revolutionize the production of magnets found in computers, mobile phones, electric cars and wind-powered generators“. According to one of the co-authors of the study, “[s]tate-of-the-art electric motors and generators contain highly coercive magnets that are based on rare-earth elements, but we have developed a new material with similar properties without those exotic elements“.

The material is apparently based on a compound of manganese (Mn) and gallium (Ga), with Northeastern claiming that the material “can be synthesized on the nanoscale to produce a coercive field that rivals materials containing rare-earth elements, which are considerably more expensive to process and mine“.

The message boards are abuzz with this announcement, apparently with many people (i.e. retail investors in the rare-earth sector) now worried that this material is the death knell for permanent magnets based on the rare earths neodymium / praseodymium (Nd / Pr), and thus the hopes and dreams for untold riches from these commodities…

Take a deep breath, folks.  Being a materials scientist by training, I am naturally a big fan of ongoing research & development work on new engineering materials, and I will read with interest more details on this research, in a forthcoming edition of Applied Physics Letters. I am much less of a fan of the now well-worn path of hype disguised as scientific (and more importantly engineering) breakthroughs, which this announcement represents.  Here’s why:

  • While Mn is cheap as chips, Ga is at present 2-3 times more expensive than Nd / Pr;
  • The production of Ga is approximately 200 tpa – of which perhaps 100 tpa comes from recycling – and it is presently all spoken for. Compare this to the more than 20-25 ktpa of Nd + Pr available each year, and the prospects for multiples of this production rate in the near future, from new sources of supply.
  • All new Ga is produced as a byproduct of aluminum and zinc production. The supply dynamics of these two metals alone will determine future availability of Ga – not its potential use in a permanent-magnet material.
  • Given the painfully long road to commercialization for other materials that rely on similar processing routes, it is highly unlikely that synthesis “at the nanoscale” will be less expensive than mining and processing rare earths any time soon.
  • Finally, while we’re at it – a “highly coercive” magnet material, is not the same thing as a “super-strong” magnetic material. The former refers to the ability of a material to resist being demagnetized; the latter to the ability of the magnet to do work.
This new Mn-Ga compound is certainly very interesting scientifically; but unless someone finds a primary Ga deposit, and can perfect inexpensive nanoscale material production, it’s not going to “revolutionize the production of magnets found in computers, mobile phones, electric cars and wind-powered generators” anytime soon.
Bookmark and Share Print
1 hackenzac December 20, 2011 at 1:10 am

It’s never too early to corner the magnetic jet pack technology sector.

2 Antonio Turiel December 20, 2011 at 1:54 am

Very good points, Gareth. I was wondering about Galium supply. Thanks for sharing your thoughts.


3 John Petersen December 20, 2011 at 1:55 am

In 1992 I served as issuer’s counsel for the first carbon nanotechnology IPO. The company was basically a spin out from the carbon nanotechnology lab at Rice University and was using carbon vapor deposition to put diamond films on non-diamond surfaces. Nineteen years and $110 million later, the company is still losing a little money every year.

4 Richard Lee December 20, 2011 at 1:56 am

Gareth, Thanks for sharing the information.
I was asked recently by a LinkedIn Member about the prospect of buying a Manganese Factory in Vietnam but he was not serious in his inquiry up to now.
I have to find out whether Vietnam has any sizable quantity of Ga.
If there is any buyer interested in a Mn production line, kindly let me know.
The owner wants to sell 40% of his factory and the adjacent mine for about US$17 million. Cheers, Richard

5 Vladimir Seredin December 20, 2011 at 3:56 am

Finding of Ga-bearing coal in the northern China and the beginning of Ga production from fly ash as byproduct next year will unvitably lead to sharp increase Ga supply and decrease cost of the rare metal.

6 Claudio Holanda December 20, 2011 at 6:19 am


As usual your viewpoints are well founded.

Follows some information about alloys with Mn. Ni and Ga (from Wikipedia) ” As alloying element in the magnetic shape memory alloy Ni-Mn-Ga.”

So the research from Northeastern University isn’t completely new.

You are right. Beyond the higher price, the global production of gallium is
quite small to threat the production and market of REE.

7 Igor December 20, 2011 at 7:36 am

Gareth Clever! Excellent! I agree 100% between the appearance of scientific development and the product is no road for one year. And also we do not know what the properties materials in the finished product and how it will behave over time.

8 G H December 20, 2011 at 8:06 am


The page at the link you provided contains no information, only a come-on to sell a $42 document. Posting here is a creative way to market it, I’ll give you that.

Apparently you’re claiming that this coal will be segregated to go only to consumers that will capture the gallium byproduct? How is this different than the relatively high amount of mercury in many coal sources that is currently being lost as air pollution? I grant you gallium is more valuable than mercury, but until you post the commercial economics of the process, we have no way to judge your prediction.

9 Eamon Keane December 20, 2011 at 10:18 am

I did my undergrad thesis on Al-Ga alloys after my supervisor’s interest was caught by one such press release which claimed it would be cheaper than gasoline and could run all the world’s cars! Needless to say, it couldn’t, with primary Ga production a prominent obstacle.

The alloy of interest seems to be Ni2MnGa, which has been known about since at least 1980. Gallium thus makes up 31% of the alloy – the same as Nd/Pr in NdFeB. Assuming the same weight of magnets as NdFeB (500 kg/MW), current primary gallium production (106 tpa) would cover 685 MW – about 1.6% of 2011 wind installations.

Vladimir’s article above mentions recovering gallium from Chinese fly-ash, with slightly higher gallium concentrations than typical – 92 ppm vs. 80 ppm, but with very high gallium recovery – 90%. On foot of this a pilot plant has been built with processing capacity of 800,000 tpa Al and about 150 tpa Ga and will commence production in 2012. Four more such plants are proposed out to 2015, potentially bringing in 700 t Ga. No mention of costs.

Even at 1 ktpa gallium, it still pales in comparison with the 25 ktpa Gareth mentions. Also, I wonder did they authors bother to check who the largest gallium producer is when using this soundbite:

“The government would be in a bind if it had to rely on China to produce hybrid cars and wind generators.”

10 Dr. Jim Wolter December 20, 2011 at 11:38 am

Gareth I am no rare-earth expert but do have a MS in Physics and follow soft/permanent magnetic materials development. Sorry for a “beef” on your excellent site but I wish someone would tell the truth about “wind power”. As you may know, Utilities have flocked to wind farms because they can all recover capital costs (plus a margin for their stockholders) and wind machines are sold according to their nameplate ratings (e.g. 5 megawatt) NOT BY THE POWER THEY PRODUCE ANNUALLY. The marketing forces behind wind have Michigan Utilities quoting a 32% availability factor – but even if that were true more than half of their power is generated off peak – in the night when it is not needed. Face the facts – in this summer’s 35-day Texas heat wave – the New York Times stated that the entire Texas wind capacity produced 8% of its aggregate nameplate rating. A 42-year observation published in a peer-reviewed scientific study done in Europe recently concluded that in all of Western Europe the combined, year-around wind availability was in aggregate 13%. So before your investors count their future returns – watch the money – Warren Buffet just invested over a $billion from his “wind machine business” into a solar-PV project with First Solar. Read the tea-leaves …

11 Ferdy Pacheco December 20, 2011 at 11:57 am

Gareth:”This new Mn-Ga compound is certainly very interesting scientifically; but unless someone finds a primary Ga deposit”…

Check out Gold Canyon GCU.V). They are a gold exploration company, but they have been sitting on a huge primary gallium deposit for years.


12 Paul Gubbens December 20, 2011 at 2:35 pm

Dear Gareth
I am very doubtful about this FeGa nano compound. I can not exclude it. In Fe and Co the spin-orbit coupling is mainly broken up due the high electric fields. Therefore, the magnetic anisotropy (comes from the orbit) in metallic Fe is relatively low. For high magnetic anistropy you need Rare Earth.
Greetings Paul Gubbens

13 gobucks December 20, 2011 at 2:35 pm


Two things.

1. The first thing you hear about a new material is always the best thing you hear about it.
2. Nanoscale manufacturing is great as long as you need nanograms of material.

OK, three things.

3. Manganese may be cheap, but there are no domestic sources.

Merry Christmas to all!!!

14 Ferdy Pacheco December 20, 2011 at 3:30 pm

{3. Manganese may be cheap, but there are no domestic sources. }

American manganese (AMY.V) (actually a Canadian company), has a very nice project in Arizona. Should be coming online in couple of years.

15 Anon December 20, 2011 at 6:01 pm

American Manganese (AMY.V) may also become the lowest cost producer based on a new proprietary process.

Happy Holidays to everyone!

16 Mycroft Risby-Saxham December 21, 2011 at 10:59 am

Dear Dr Hatch

Good piece.

However, I’m afraid your figure for annual gallium production is way out.

In 2010 alone, primary gallium production was estimated to have been north of 200 tonnes, excluding any recycling.

Have a fine New Year.

17 Gareth Hatch December 21, 2011 at 11:19 am

@Vladimir Seredin: thanks for the link – we’ll take a look at recent work on this topic.

@Claudio Holanda: it now appears that this is a thin-film material, doubly unsuited for bulk permanent magnets.

@Eamon Keane: the paper is presently available at http://arxiv.org/ftp/arxiv/papers/1111/1111.6267.pdf – per Chuanbing Rong at Ford: “The composition is Mn2Ga~Mn3Ga, indicating 30~38wt%Ga. The coercive field is up to 2.5 T. However, the saturation magnetization (potential remanence) is only 130 emu/cc = 0.16T, compared with NdFeB’s 1.6T. Actually, all Mn-based magnets have low Ms, which is not strange.”

Alos, there appear to be clnflicting data for Ga production, even within a single paragraph of the US Geological Survey’s summary on gallium – either 106 t or 184 t depending on which sentence we look at…

@Dr. Jim Wolter: your appoint about nameplate versus actual production rates is acknowledged :-)

@Ferdy Pacheco: thanks for the link. I guess we could call that a Ga deposit, though the concentrations of Ga are not exactly high…

@ Paul Gubbens: indeed – we use rare earths for a reason!

@gobucks: thanks for the chuckle :-)

@ Mycroft Risby-Saxham: as mentioned above, I’ve seen two different numbers from the US Geological Survey’s own publication on Ga (106 t and 184 t respectively); I’m sure there are other estimates out there too.

18 Peter Vesborg December 22, 2011 at 5:40 am

@ Gareth Hatch
Good points.
I’d love to have your reference for the 200 tpa production figure. My research indicates a primary supply of Ga in 2010 of only 125 t.
Also – do you have any reference for Ga-supply from Zinc? I’ve only seen data for extraction from Bauxite.

@ Vladimir Seredin
The potential Ga supply from coal is huge. Always was. However, can you point to any producer which has produced meaningful amounts of Ga from coal ash in the last 30 years anywhere in the world?

To me gallium from coal is a typical pie-in-the-sky resource.

19 Dave Higgen January 2, 2012 at 5:31 pm

Hi Gareth,

As a scientifically literate (but uneducated in this area) bystander, I’ve always wondered how the newer magnetic materials were discovered?

First Samarium cobalt, then NIB… was there any (maybe quantum-mechanical) theory behind this, or was it just accidental serendipity?

I’ve never seen any history of this… it would be very interesting to learn more!

Dave Higgen

20 anon2 January 3, 2012 at 11:37 am

Hi Dave Higgen,

NdFeB was discovered in 1984 by General Motors and Sumitomo Special Metals in a joint research effort to replace Samarium Cobalt after the cobalt shortage of 1978 in Zaire.

The US Dept of Energy’s ARPA-E program funds a number of universities, Northeastern University included, to investigate “rare-earth free” magnetic materials.

21 Gareth Hatch January 5, 2012 at 6:46 am

@ Peter Vesborg: I rounded up the USGS numbers published in their 2011 Mineral Commodity Summary for gallium (the numbers haven’t changed in the past few years). My understanding is that Dowa Mining of Japan (and others) recovers gallium as a byproduct of zinc refining.

@Dave Higgen: probably a combination of both… YCo5 was discovered at US Air Force Materials Lab at Wright-Patterson AFB in the 1960s, with further work leading to the discovery of SmCo5 and later Sm2Co17 type compounds. The independent discovery of NdFeB-based magnet materials was announced at the Pittsburg MMM conference in 1983, by Croat et al of the General Motors Research Lab (using a rapid solidification process) and Sagawa et al from Sumitomo Special Metals (using a sintered powder method). Alloy systems based on Fe were of significant interest because of the lower costs than Co, and greater availability.

There was also important earlier work done by Koon and his team at the US Naval Research Lab, which was arguably essential to the subsequent success of the other two groups. To my knowledge there was no collaboration between any of these groups.

Interestingly there was at least one group, working in Russia, that had published work on the NdFe(B) alloy system in the late 1970s, but did not, to my knowledge, look at the magnetic properties.

22 Dave Higgen January 10, 2012 at 6:38 pm

Thanks Gareth,

I guess what puzzles me is how the heck did they know *where* to look? Was there any theoretical work suggesting that the Rare Earth elements might lead to strong magnets? Or were they just throwing stuff at the wall, trying random combinations?

Of course, the follow-on question is: what other marvels might lurk somewhere in the periodic table?

Comments on this entry are closed.

Previous post:

Next post: