Evidence: Well-supported
The experiment, without the myth layer
Andre Geim, Konstantin Novoselov, and collaborators isolate few-layer and monolayer graphene by mechanical cleavage from graphite — the “Scotch tape” method — transfer flakes onto oxidized silicon, and measure an electric-field effect in atomically thin carbon films.
Landmark: Novoselov et al., Science 306, 666 (2004), “Electric field effect in atomically thin carbon films.”
What made 2004 the inflection point was the combination:
- Flakes you could see (contrast on SiO₂ of the right thickness).
- A monolayer assignment that the community accepted.
- A device geometry (gate through the oxide) that produced a field effect.
- Numbers that lined up with the 2D honeycomb / Dirac picture Wallace had sketched in 1947.
Chemists had already made very thin carbon (1962). Surface scientists had already grown monolayer carbon on metals and SiC (1963-2003). 2004 is when a freely handled sheet on an insulator became an electronic material other people could copy next week.
Culture note, kept in its box
Manchester lore ties this to “Friday night experiments” — side bets next to official projects. Useful color. Not a substitute for the Science paper.
What 2004 did not do
- Did not invent carbon sheets.
- Did not produce square metres of CVD graphene (that scale-up is later).
- Did not settle quality standards for powders sold as graphene in 2026.
- Did not imply graphene would replace silicon CMOS. Band gap, contacts, and wafer uniformity showed up immediately as engineering walls (2011-2012).
Monday-post kit
A historical post can stand on three links: Wallace 1947 (theory), Boehm 1962 (oxide monolayers), Novoselov Science 2004 (transport on oxide). Then point readers at 2010 for the Nobel and at 2026-W38 if you need a “what the field is actually shipping now” contrast.
Next
Source: timeline/2004.md