That page have invaluable articles. I have saved most of them in pdf format ,edited with Open Office.
You just need to install Open Office (free at openoffice.org), its like the Window's Office, but better and free (and open code). When installing it, select integrate it with your browser, so when you are in a web page, you can import and edit it. It allows you to export the page in many differents formats, pdf included.
Although the articles in that page are great, they are from 1994. Most of the article's authors continue working in that field, so searching for them points to more recent articles.
Coming back to the original topic, i think plants mainly needs photons between 420 and 685nm, as more as possible.
Only spectral requeriment is some blue light, in wide meaning (from 400 to 500-510nm), and not for all plants species: wheat, for example, dont require it at all. Experiments carried to determine the quantity of blue required shows that most species that require it have enough for health grow with 30-40 uE/m2, and many of them, far lower, while some requires a percentage of blue light (respect the whole irradiance). But required or not, blue light promotes smaller internodal distances, wich often is desiderable, and maximum effect is done at the same wl than phototropic effects (~450nm+-30nm).
For the remaining wl, there is little knowledge about sinergies between differents wl, apart of the famous Emerson Effect (giving 660 and 700nm light at same time produces more photosyntesis than giving it the same quantity but separatelly). So any advice about best spectral mixes are theoretical, and must be proved yet.
So to do a LED array for most plants only requires some percentage of the blue light (or an absolute emission per sq meter), and the rest, of any other wl, being the best the red leds, because same energy emited at longer wl carry more photons, and because currently red leds are more energy efficients (optical watts /input watts), although this is changing, and maybe in some years blue leds will be more efficients, so it would be necessary to study carefully the best option (more photons per input watt).
Its possible that some wl mixes work better than a 2 light band source (blue*red), but it must be proved.
Although plants grown under sun have less efficiency both absorbing and proccesing green photons, plants have a great capacity of adaptation, so plants growing under dominant green light finally have similar photosynthetic capacity processing green photons tan red ones, and higher than blue ones (see fig1 and explanation of
"SPECTRAL COMPOSITION OF LIGHT AND GROWING OF PLANTS IN CONTROLLED ENVIRONMENTS")
So with enough blue light, any LED array work growing plants if enough irradiance are provided. Sure some mixes will work better (more growth for same input watts), but most botanical studies states that total growth are directly related to total irradiation in photons. I believe main concern when building a LED array for plants is select the LEDs wich emits more photons (per input watt and/or per cost).
Author