Really? I mean, economics of industrial chemistry is quite complicated, but it seems utterly trivial. If the price goes high enough what's the problem? With a moment's Googling, I promptly found some lab articles about synthesizing nickel sulfide from nickel to get *really* high grade nickel sulfide.
In fact, it appears from what I can find that processing of the (apparently fairly common -- 40% of all primary nickel deposits, and 58% of world nickel production) nickel sufide ores first separates out the nickel sulfides, and then converts them to elemental nickel.
They could just, well, stop doing that. If battery nickel sulfide prices exceed the prices of metallurgical nickel by enough to cover the difference between the nickel sulfide purification costs and the costs of converting nickel sulfide into metallurgical nickel, they will stop.
The primary problem with producing nickel sulfides appears to be iron removal, which is a pain.
But in the short term of a few years it seems like just redirecting existing nickel sulfide streams away from making elemental nickel would do the trick.
Are they particularly pure or something? Lacking in cobalt sulfide and copper sulfide and iron sulfide and similar expected contaminants?
Regardless of what Wikipedia suggests, nobody makes nickel sulfate from nickel metal at scale; it's far too expensive. Battery-grade nickel sulfate is primarily produced directly from high-grade sulfide ores, while metallurgical nickel (ferronickel, nickel pig iron, NPI etc) is produced mainly from low-grade sulfide ores and laterites. The key difference in sulfide ores is the nickel concentration and form, which affects how difficult it is to get a high purity product.
Like sulfides, laterites can also be graded. High-grade laterites are more common than sulfides, while low-grade laterites are
extremely common. The main difference is the geologic history. Laterites (your typical reddish-yellow tropical soils) are a byproduct of decomposition of olivine and serpentine, followed by subsequent oxidation. The iron tends to oxidize first and precipitate, while nickel, magnesium, and some other metals stay in solution and tend to precipitate downward until the solution is neutralized, wherein they precipitate out. But the degree of separation from the iron is never complete, usually quite limited, and sometimes doesn't even happen at all. This is limonitic ore, which is harder to process to a high-quality product (but on the upside, contains more cobalt and chromium). A well separated "silicate laterite" has the nickel mixed in with silicates and has a high magnesia content. Silicate laterates are suitable to pyrometalluric processes (smelting), while limonitic needs hydrometalluric processes (such as HPAL).
The 40% / 58% production does not correspond to reserves. Sulfides are preferred for production because they're cheaper to process, for a variety of reasons (even still, they're a minority of production). Reserves are a 20 / 80% split. And remember that all "reserves" figures you see for any resource are relative to a fixed price, tech, and exploration level. Sulfide deposits are highly geographically limited, while laterites can be found almost everywhere, and the only constraint is their production costs.
A key distinguishing factor is that oxide ores (laterites) are homogenous (nickel evenly mixed) but laterites are heterogenous (nickel minerals occur as distinct grains). So when processing sulfide ores, you use liberate the individual grains and then use physical processes (such as froth or magnetic separation) to concentrate the nickel minerals. Consequently, the purity of the nickel mineral grains themselves is a key factor, not just the overall nickel concentration of the ore. And it's not as simple as it sounds - in theory you can produce separate pentlandite (nickel), chalcopyrite (copper) and pyrrhotite (iron) concentrates - but part of your nickel exists as pentlandite inclusions and in solid solution in the pyrrhotite. So most producers outside of Canada don't even bother trying to separate them before smelting to ferronickel. The pyrrhotite fraction is however problematic because it contains most of your sulfur but only a minority of your nickel.
Where separated, nickel concentrates are usually only 5-15% nickel. 28% is considered "exceptional".
Nickel sulfate only makes up about 10% of the nickel market, and EV batteries in turn only consume a fraction of that. Neither ferronickel nor nickel sulfate ever exist as pure nickel metal during their production; ferronickel is a nickel-iron alloy (about 2/3rds nickel), and is produced that way. It's far too expensive to convert it to battery-grade sulfate. I forget the cutoff on battery grade, but I'm pretty sure it's at least 95% pure. While ferronickel is made via pyrometallurgical (smelting) processes, nickel sulfate is made from leaching of ores or concentrates (such as with sulfuric acid or ammoniacal ammonium sulfate) to dissolve the nickel (and other metals) into solution. Unwanted metals (such as copper) are then precipitated out of the solution via cementation or hydrogen sulfide (the latter also removes arsenic), whie residual iron can be removed with chlorine and nickel carbonate (this also coprecipitates lead and arsenic). Cobalt - which remains in the stream - can be separated by various methods, such as mixing nickel hydroxide into the solution (created by neutralizing nickel sulfate with sodium hydroxide) to precipitate cobalt hydroxide, or solvent extraction with a tertiary amine, such as D2EHPA. I would expect that most battery producers have no interest in reducing the cobalt content of their nickel sulfate, and would instead prefer to maximize it.
There are a wide range of electrolytic and carbonyl processes for producing high-grade nickel metal from nickel sulfate, but they do not apply when the desired feedstock is the nickel sulfate itself. A company like Tesla will process the sulfate on their own as they see fit to produce their cathodes.
As mentioned, the "dream scenario" for battery grade nickel production is if you can produce it from (common, cheap) laterites, and in particularly, the (really common, really cheap) low-grade limonitic laterites. One annoying thing about them is their high moisture content, due to the hydroxide minerals they contain (incl. limonite itself, iron hydroxide). If you could effect the results of drying and reduction roasting simultaneously with direct leaching, via acid dissolution, this would be hugely advantageous - except that under ambient conditions, almost everything (incl. the iron, which you don't want in solution) will dissolve. But under high temperatures and pressures the iron will precipitate out as hydrolated iron(III) oxide. This is HPAL - high pressure acid leach. Which is awesome, except for the fact that now your plant has to be able to withstand high-temperature high-pressure sulfuric acid! Being able to make a plant that can operate reliably under these conditions has been a big challenge, but it looks like there's been major progress on this front in recent years.
If you want to learn more on any refining process, I strongly recommend Ullmann's Encyclopedia of Industrial Chemistry. Great for scratching any chemical engineering itch
Discovered it back when I was writing
a technical analysis on the colonization of Venus.
