The Dangerously Clean Water Used To Make Your iPhone

The ultra-pure water used to clean semiconductors and make microchips would suck vital minerals right out of your body. Plus it tastes really nasty.

iPhone water

FACT: Water can be too clean to drink—so clean that it’s actually not safe to drink.

That’s the kind of claim about water that people scoff at—it seems ridiculous on the face of it.

Water too clean to drink?

Give me a break. It’s water. Cleaner is better.

But this is one wild water story that’s true.

Every day, around the world, tens of millions of gallons of the cleanest water possible are created, water so clean that it is regarded as an industrial solvent, absolutely central to high-tech manufacturing but not safe for human consumption.

The clean water—it’s called ultra-pure water (UPW)—is a central part of making semiconductors, the wafers from which computer microchips are cut for everything from MRI scanners to greeting cards.

Chips and their pathways are built up in layers, and between manufacturing steps, they need to be washed clean of the solvents and debris from the layer just completed.

But the electronic pathways on microchips are now so fine they can’t be seen even with ordinary microscopes. The pathways are narrower than the wavelengths of visible light. They can only be seen with electron microscopes. And so even the absolute tiniest of debris can be like a boulder on a semiconductor—so the chips have to be washed, but with water that is itself absolutely clean.

The water must have nothing in it except water molecules—not only no specks of dirt or random ions, no salts or minerals, it can’t have any particles of any kind, not even minuscule parts of cells or viruses.

And so every microchip factory has a smaller factory inside that manufactures ultra-pure water. The ordinary person thinks of reverse-osmosis as taking "everything" out of water. RO is the process you use to turn ocean water into crystalline drinking water. And in human terms, RO does take most everything out of the water.

But for semiconductors, RO water isn’t even close. Ultra-pure water requires 12 filtration steps beyond RO. (For those of a technical bent, the final filter in making UPW has pores that are 20 nanometers wide. At the IBM semiconductor plant I visited, they send the 20 nm filters out to be inspected by a private company, using a scanning electron microscope. They want that company to find filters with nothing in them.)

Just the one IBM microchip plant in Burlington, Vermont, makes 2 million gallons of UPW a day for use in manufacturing semiconductors, and there are dozens of chip plants around the world. UPW is also used in pharmaceutical manufacturing, but it is a purely human form of water—water that is literally nothing like the stuff that exists naturally on Earth.

Water is a good cleaner because it is a good solvent—the so-called "universal solvent," excellent at dissolving all kinds of things. UPW is particularly "hungry," in solvent terms, because it starts so clean. That’s why it is so valuable for washing semiconductors.

It’s also why it’s not safe to drink. A single glass of UPW wouldn’t hurt you. But even that one glass of water would instantly start leeching valuable minerals back out of your body.

At the chip plants, the staff comes to regard UPW as just another part of a high-tech manufacturing process. One senior IBM official was stunned when I asked her what UPW tasted like. Despite presiding for years over the water purification process, she not only had never tasted it, it has never even occurred to her to taste it. One of her deputies had, though, and he piped right up. "I stuck my tongue in it," he said. "It was horrid."

In fact, super-clean water tastes flat, heavy, and bitter. The opposite of what we like. The appealing freshness in water comes not just from it’s temperature and its appearance, but from a sprinkling of salts and minerals that give it a crisp taste.

So there it is: Not only is it possible for water to be too clean to drink—it’s exactly that kind of water that makes your iPhone possible.

Adapted from The Big Thirst: The Secret Life and Turbulent Future of Water, to be published in April by Free Press / Simon & Schuster. © 2011, Charles Fishman.

Read the feature from Fast Company's April issue.

Read more from The Big Thirst on FastCompany.com.

[Image: Flickr user phalkunz]

Add New Comment

15 Comments

  • Tanjila Rahman

    change your ‪#‎Water‬ change your ‪#‎Life‬ Kangen Water® is a trademark of Enagic® Inc. of Japan. Kangen has become a household word in Japan and is now used mostly to describe healthy alkaline drinking water which is rich in minerals, purged of impurities, and ionized through electrolysis. Alkaline Kangen Water® is an excellent choice for healthy hydration. Not only does it taste great and refreshing, but it is also highly "drinkable" www.kangenwaterint.info

  • Cory Salvesen

    Ultrapure water does NOT taste bad and does not leech minerals from your body. This is a common myth. There are absolutely no studies that show this effect in humans or animals. Ultrapure water has a PH of 7. If it's acidic, it's not ultrapure.  Imagine you mix some ultrapure water with a cracker and grind it up.  Stick an electronic TDS meter in it and you'll find it's at least 50ppm (as most of it won't dissolve) , which is no more pure than your average bottled water. This always happens in your stomach.  Any water you drink is immediately NOT ultrapure.. its not even especially pure.  You can drink ultrapure water all day.

  • D Olin

     You are clueless. I work with 2-18 Meg DI water on a weekly basis. It dries your hands out like crazy by pulling all of the minerals and oils from your skin. When drank it pulls the electrolytes from your system and dries you out.

    Ultra pure water has no TDS(total dissolved solids) in fact you can't even measure its conductivity. You have to measure ultrapure water in RESISTIVITY. Store bought DI or distilled water is not as pure as what is used by industrial/pharmaceutical manufacturers. As soon as it is exposed to open air

  • Ryan Takahashi

    You clearly have no idea what you are talking about. Do you even understand the process of osmosis? Or the concept of water intoxication? Please don't spread false information like that. You'll get someone hurt or killed.

  • bob

    As soon as the ultra pure water interacts with the gastric juices, it's no longer ultra pure, fact.
    I'd wager drinking ultra pure water would be an identical experience to drinking distilled, as long as the pH is the same.

    Cory is 100% right.

  • carydecamp

    Bob,

    Cory's payoff statement is that "Any water you drink is immediately NOT ultra pure." Then you say that the ultra pure water interacts with the gastric juices rendering it no longer pure. You are both right.

    But...duh!!!!!! That's precisely the point - and the problem for humans.

    Ultra-pure water is used for the same reason any solvent it used - because it "draws," "binds" or "seeks to mix" with other solutes to create a new solution (no longer ultra-pure water). It works on your body just like it does on a silicon wafer or a printed circuit board. It will do that to anything it comes it contact with provided relevant solutes are present. PH and relative conductivity are irrelevant.

    Obviously it's no longer pure after you drink it. It's gathering to itself tasty things it wants, removing them from your body and ultimately carrying them out of it.

    Don't drink it in quantity unless you want to appear on http://www.darwinawards.com.

  • carydecamp

    Bob,

    Cory's payoff statement is that "Any water you drink is immediately NOT ultra pure." Then you say that the ultra pure water interacts with the gastric juices rendering it no longer pure. You are both right.

    But...duh!!!!!! That's precisely the point - and the problem for humans.

    Ultra-pure water is used for the same reason any solvent it used - because it "draws," "binds" or "seeks to mix" with other solutes to create a new solution (no longer ultra-pure water). It works on your body just like it does on a silicon wafer or a printed circuit board. It will do that to anything it comes it contact with provided relevant solutes are present. PH and relative conductivity are irrelevant.

    Obviously it's no longer pure after you drink it. It's gathering to itself tasty things it wants, removing them from your body and ultimately carrying them out of it.

    Don't drink it in quantity unless you want to appear on http://www.darwinawards.com.

  • DudeMan

    I don't think so. Dip your arm in the stuff and your skin drys out in minutes. 

    Sure when you drink it it'll mix with body fluids and it's no longer ultra pure. That doesn't mean it can't hurt you. It's still still leaching minerals from your body.

  • Andrew Berry

    I used to work as a technician for a company called Millipore-Waters, and I was warned about tasting the "polished" water, as it was corrosive...this stuff leaches EVERYTHING out of your body

  • louise brake

    I use ultrapure water at work and we dump it down the drain.  Once it is done cleaning etc. its not ultrapure anymore.  Ultrapure is expensive to make so I would assume a large production facilities would recycle this water. Remember when making RO water just as much goes down the drain as goes into your tank.  So if I use RO to feed my upw polish loop and produce and use 8,000 gal/day, then I'm dumping 16,000 gal/day down the drain.

  • Claudio Silva

    I believe that UPW is not disposed of. It's reasonable to assume that after washing the semiconductors, it's just refiltered and made into UPW again. I don't any reason why that water would be wasted to filter a brand new "dirty" water. A "dirty" UPW is probably way purer than regular water....

  • Buzn8tor

    I work for a semiconductor manufacturer and we were told by the local water board/state environmental agency that the water we were putting back into the system was "too clean"... so we have to add back in trace amounts of salts, minerals, etc. 

  • Grammar

    Thanks for the article! Note: "just from it’s temperature" should be "just from its temperature". This is the internet.