February 06, 2004

[A bit of research, which I should have done before posting, reveals that the foam that damaged the Columbia's wing was the old, Freon-based BX-250 foam. This means, of course, that a significant portion of this post is factually incorrect and rhetorically overblown. The IC resin story is, as far as I know, accurate.--IAW]

In 1997, NASA decided to replace the sprayed-on Freon-based insulating foam that was used on the space shuttles external tanks with a new, "environmentally friendly" formulation. The result? The number of damaged tiles on the shuttles increased 11 times because the new foam broke free during flight with such ease. And, on February 1 of last year, a piece of the new foam caused the deaths of seven astronauts, the loss of an orbiter, and the grounding of the entire shuttle fleet.

Now, via Slashdot, we have this cheery news:

In about June 2001 a rumor began circulating through the industry that certain Fujitsu HDDs were failing at an unusually high rate. One after another, 3.5-inch HDDs mounted internally in desktop and other PCs were failing to spin up. The cause was a failure in the drive controller, the CL-SH8671 (codename: Himalaya 2.0) from Cirrus Logic, Inc of the US. The failure was caused by a short between pins within the integrated circuit (IC) package.

Originally the issue was thought to be affecting only HDDs, but more recently similar defects have begun appearing in a range of other equipment, including set-top boxes, PC main boards, IC test systems and industrial machinery. The issue is developing into a major problem, and has rapidly come to involve a host of equipment and IC manufacturers.

Why is this happening? Because the IC semiconducter industry changed the flame retardant in the plastic resins used to encapsulate IC chips and other components from a bromine-based formulation to a red phosphorous-based formulation:

The prevailing industry position is that the primary cause of the IC failure is the EME-U series of encapsulation resins containing red phosphorus, developed as part the halogen-free environmental product program at Sumitomo Bakelite Co, Ltd of Japan.

And why did they make the switch?

The answer lies with the equipment and materials manufacturers, who must stress environmental considerations.

The flame retardant most commonly used in encapsulation resins is a combination of Br-based compounds with additive Sb 2 O 3 (antimony trichloride). This mixture is extremely effective, and an encapsulation resin with 2 to 3% content will clear the American UL94-V0 standard for flame retarding performance. It also has a long record of successful performance in the field.

Br-based compounds, however, have been cited as potential sources of dioxins and other toxic gases when combusted, and this eventually led to restrictions on their use from about 1990, primarily in Europe. This accelerated the trend toward halogen-free material development, not only in encapsulation resins but in all types of applications.

Think about that: Europe was concerned about the toxic releases from encapsulation resin with, at most, a 3% Br-based content...while it was on fire. Not while the chip was directing the illumination of phosphors on your television screen, or helping to start your car's ignition, or keeping your heart monitor going while you're in the hospital. While it was on fire.

Shuttles are falling from the sky and devices trivial and critical alike are failing at twice their previous rates...but if your house burns down, your computer will not release any dioxins.

I cannot think of a more perfect pair of examples to illustrate the dangers of dogmatic environmentalism. We use all sorts of toxic compounds in an incredible variety of manufacturing processes because they're effective. If they're toxic, they should be sensibly regulated or replaced with safer materials that are just as effective and don't cost eight times as much.

Do you feel safer knowing that the electronic flight-control components of the next airliner you take may be subject to a higher rate of failure? Are you pleased that your television, computer, or automobile might crap out years earlier?

Outside of fires at manufacturing facilities and the possible incineration of retired equipment, the risk of dioxin release is virtually nil. The way to deal with such problems is at the source of the danger: improved fire control standards in factories that handle the material, and improved emission and toxic disposal procedures at plants that regularly destroy or recycle large quantitites of electronics.

But why do that when you can issue broad prohibitions that will cost hundreds of millions of dollars, create poor-quality products, and kill astronauts?