[ghostly_s]

Not sure, I read an article some years ago that said as much. I expect you could identify natural diamonds by impurities and subtle imperfections in the cut, but I don't know how you could conclusively prove a too-perfect diamond was synthetic.

[gambiting]

According to this article[0] it is possible, but it's not like there's one 100% accurate method - but a skilled gemologist should be able to identify a synthetic diamond:

https://www.gia.edu/sites/Satellite?c=Page&childpagename=GIA...

[ridgeguy]

Yes, it's possible with a high degree of confidence to distinguish natural from synthetic diamonds and for synthetics, to determine which method (HPHT or CVD) was used to make them.

The IIDGR [1] company makes a series of devices (DiamondSure, DiamondView and DiamondPlus) that will indicate whether a diamond is a natural (and if so, of what type), or flag it for further characterization. One system will flag type IIa naturals for further analysis, as they resemble the best of CVD synthetic diamonds. Microscopy + (IR, fluorescence, and photoluminescence) spectroscopy distinguish a stone's origin to high confidence. Even though the best CVD stones have far lower impurities than the best naturals, they have internal strain patterns (optical birefringence) that can be diagnostic.

Most jewelers don't have these machines, but they are available as a service.

Faceted CVD stones have been seen up to 5 carats [2] and 10 carat stones made by HPHT are known [3]

[1] https://www.iidgr.com [2] https://www.gia.edu/gems-gemology/winter-2016-labnotes-CVD-s... [3] https://www.gia.edu/gems-gemology/winter-2016-labnotes-blue-...