Excitons in metals-do they exist? Recently I red an article "Surface Enhanced Fluorescence". It is a topical review by Emmanuel Fort and Samuel Gresillon. 
Here it is:
http://turroserver.chem.columbia.edu/surfaceplasmons/pdf/53_JPD_41_1_2008_SEFSRev.pdf
On the 8th page of pdf there's such thing stated:

The energy is
  transferred to electron–hole pairs of the metal (excitons)
  which act as energy acceptors.

How is it possible? I was searching for knowledge about excitons and it seems that they can't exist in metals. Even princeton says so:
https://www.princeton.edu/~achaney/tmve/wiki100k/docs/Exciton.html
Not a single word about metals.
So... can I treat this article seriously? Or maybe there's something I'm  missing? Can an exciton exist in metal?
 A: Your understanding is correct, speaking of an exciton in a metal does not make sense. As your reference (from princton) states correctly: an exciton is a bound state of an electron and hole. Both bold points cannot be fullfilled in a metal:
1) Any binding energy that results from the Coulomb interaction, has to go to 0 in a metal, because the dielectric constant goes to infinity. 
2) The electronic description of a metal usually only includes one (half-filled) band, so there exist no two different particles in this system (electrons and holes) to form an exciton. 
Note, the more appropriate quasi-particle in metals are plasmons. They come in different flavours, such as surface or bulk plasmons.  
With respect to the energy loss, discussed in your first reference, the dipole-dipole coupling that is assumed is probably correct, just that the energy acceptor in a metal is a plasmon rather than an exciton. Talking about a plasmon in this context is probably not helpful though, because a classical description of this loss mechanism should be sufficient in most situations.
http://en.wikipedia.org/wiki/Plasmon
A: Excitons are a transient response of metals to optical excitation. For early experimental and theoretical discussion see:
http://journals.aps.org/prl/abstract/10.1103/PhysRevLett.18.448 
see also (Phys. Rev., 1967. 157(3): p. 600)
The experimental evidence for the transient excitonic response of silver surface has just been reported by multidimensional multiphoton photoemission spectroscopy:
http://www.nature.com/nphys/journal/vaop/ncurrent/full/nphys2981.html
A: I'm not really sure if this answers your question but I found this website and it's recent (June 1st), I just saw this while looking through the website and I do think it will help you answer your question. 
http://www.sciencedaily.com/releases/2014/06/140601150926.htm
A: The answer from denizb assume the metal is perfect conductor.However for most of "Real metal"(such as Au,Ag et.al), when wavelength go to UV or X ray or even shorter, they become dielectrics(https://refractiveindex.info/). Although the argument from Princeton and some points above still hold, while for the question itself, adding more condition would make it more accurate.
