There's a letter in the latest Nature from two researchers in Halifax that makes a point which isn't made often enough. Why do so many papers in the literature ignore patent references?
Why are patent citations so conspicuously absent across academic journals, with most even omitting formatting instructions for these in their author guidelines? Patents present novel, rigorously reviewed unpublished work, as well as providing an unmatched resource for detail.
We randomly selected one month (December 2008) and reviewed all citations in the reviews, articles and letters/reports in Nature (1,773 citations) and Science (1,367). These citations included textbooks, http://arXiv.org preprints and abstracts — but no patents.
They go on to point out that searching the patent literature, which traditionally was rather a pain, is much easier now, as is access to the patents themselves. And they have a point. When I was in graduate school in the 1980s, getting a procedure out of a patent was really considered an absolute last resort - it was a special order in the library, and you had this vague feeling that there was some sort of trickiness going on, that none of those syntheses were ever actually supposed to work, anyway.
Not so. While the patent literature is indeed full of junk, the open literature is, too. They're not exactly peer-reviewed, true - but journal papers have a much lower chance of having to stand up in court, so things sort of even out. And as far as organic synthesis is concerned, patents are full of real procedures to make real things (and often enough, with real spectral data to support the claims). Most of the compounds I've made in my career that have seen any light of day at all have done so in patents, and they're real as can be.
I've complained several times when refereeing papers for publication about the lack of relevant patent citations in them. And I'd advise others to do the same - this branch of the scientific literature deserves its due.