>>11439424Because you bought the hype that individual PIs, companies, and funding agencies use to sell the scientific knowledge that they generate. Try not to get too upset when you figure out that a lot of those areas are not economically feasible.
Let's take graphmeme as a case study (1/2):
Graphene has some very interesting electrical properties, it's true. And there was quite interesting work done on the prototypical scale to demonstrate that, in theory, devices made from graphene obey the physical models we use to describe devices made from, e.g., Si, GaAs, etc, that is, standard electronic materials.
The problem is not that such devices do not fundamentally work, it's that they're typically limited by: (i) other physical and material parameters inherent to the material; (ii) their synthesis procedures are highly specialized; (iii) their material and interface properties exhibit deleterious behavior when combined with standard electronic materials; (iv) all of the above lead to inability to integrate these devices with standard electronics and scale their high volume manufacturability from a uniformity and yield standpoint.
Going back to graphmeme:
(i) It cannot provide sufficient charge density to reach the drive currents needed in modern electronic devices due to its limited volume.
(ii) It requires synthesis using metallic substrates (typically copper), and grows in highly fractured (riddled with grain boundaries) sheets that must undergo and lift-off and transfer process to the final substrate. This introduces defects into the graphmeme that utterly devastate its electrical properties. This also means that graphmeme must be placed in position on the final substrate with nanometer-scale accuracy. This is extremely difficult and a huge roadblock. Only IBM has attempted to make significant headway here with CNT patterning, but there is still some difficulty in controlling CNT chirality, and thus its semiconducting behavior.
