These clever boffins in Japan have done a detailed analysis of the raw material inputs required to make a typical memory chip. It’s pretty ugly stuff. RAM chips’ eco footprints are horrendously large and toxic, much more so than people suspected hithertofore. They make even cars seem relatively benign. Makes a good argument for re-purposing old machines. And then I was thinking of how the Intel mothership in West Dublin consumes such a huge proportion of the city’s water supply, and I found out that Intel in New Mexico was lambasted by local residents for draining their aquifer.
What are the environmental impacts of producing and using a 32-megabyte DRAM computer chip that weighs a mere 2 grams? The UNU team found that to make every one of the millions manufactured each year requires 32 kg of water, 1.6 kg of fossil fuels, 700 grams of elemental gases (mainly nitrogen), and 72 grams of chemicals (hundreds are used, including lethal arsine gas and corrosive hydrogen fluoride).
The lower bound of fossil fuel and chemical inputs to produce and use one 2-gram microchip are estimated at 1600 g and 72 g, respectively. Secondary materials used in production total 630 times the mass of the final product, indicating that the environmental weight of semiconductors far exceeds their small size. This intensity of use is orders of magnitude larger than that for “traditional” goods. Taking an automobile as an example, estimates of life cycle production energy for one passenger car range from 63 to 119 GJ (42). This corresponds to 1500-3000 kg of fossil fuel used, thus the ratio of embodied fossil fuels in production to the weight of the final product is around two. (text of article)
So I was curious to find out what these chemicals are used for and I found this:
H20: the vast majority of this is used in cleaning baths. It is always deionized water and ususally is operated in a “flow-through” manner such that there is a big tank where they put wafers and water flows into and out of this tank. 32 Kg of water likely accounts for the fact that these baths are probably kept on (because water is cheap) while wafers are not in there. The other use for this is to create steam, which when exposed to Silicon, creates silicon-dioxide (SiO2) which is typically used as an insulator. N2: Okay, this is probably not reused primarily because of the manner it is used. Typically the N2 is used like a hose to dry off wafers (like a gun). This N2 typically is simply added to the 80-some percent of N2 in the ambient air. N2 is used in lesser quantities for replacing bad gasses in vacuum chambers (known as “flushing”), but the fact that this “pure” N2 is mixed with other “bad” gasses, it is probably difficult to use without large amounts of purification. Finally, production facilities probably use this in their storage area (wafer storage) as to avoid unwanted oxides growing on the surface (see below). As: this is really bad (as most of you kiddies know) and is used in doping the Si to make it more conductive, etc (along with other chemicals). This is one of the gasses that N2 is used to flush out of the vacuum system. HF: This is (afaik) the primary technique (as outlined in the RCA cleaning process [mit.edu] to remove native oxides on the surface of the Si. As stated above, when Si comes in contact with water vapor (rich in oxygen), it forms SiO2. Well when Si comes in contact with O2 in ambient air (at a lower concentration), it will also create thinner films of SiO2, and this needs to be removed with something, which HF works very well for. This is typically neutralized and disposed of.
Related
Charles adds:
Hey Mike,
You missed some stuff in your examination of the environmental
impact of semiconductor manufacturing. For example, there’s lots of
plasma processing, and some really nasty flourine chemistries are used
to clean the plasma chambers, most of which escape into the atmosphere.
Fabs are notorious for their water consumption. The notion of fabs
in the southwest is ridiculous. But then, we live in a state that just
lost a big chunk of its water ’cause farmers in the Imperial Valley
think that’s a reasonable place to grow cotton.
I worked in a lab in grad school where we used arsine and phosphine
as precursors for growing compound semiconductors (e.g., gallium
arsenide or indium phosphide). The room was kept at negative pressure
and had its own exhaust system. When the growth chamber broke, we had
to put on self-contained breathing apparatus, crack the chamber, and
then leave the room and wait for 3 days for gas counts to drop to a safe
level. Which for arsine is measured in parts per billion. When we
finally could go back in, we had to wear respirators, ’cause the inside
of the chamber was coated with arsenic. Fun stuff.