Added proxies to emissions of elementary metals

[fchichorro]

in short, what this PR does is to expand recipe metals in ionic form to their elementary corresponding metal in fedefl. For example:

ReCiPe flowable Sb(V) (Antimony (V)) to be expanded to fedefl flowable Antimony. Before this PR, Sb(V) is being expanded to Antimony (V) only.

Why do this? To have FEDEFL impacts consistent with ecoinvent impacts.

Here's all of the proposed flowable expansions:
Sb(V) => Antimony
As(V) => Arsenic
Cd(II) => Cadmium
Cr(III) => Chromium
Co(II) => Cobalt
Cu(II) => Copper
Cu(II) => Copper(I)
Pb(II) => Lead
Hg(II) => Mercury
Mo(VI) => Molybdenum
Ni(II) => Nickel
Ag(I) => Silver
Tl(I) => Thallium
Zn(II) => Zinc

This results in 800+ row mappings.

A bit more background, the way I understand it
For a lot of metals in their elementary state, ReCiPe does not have emission impacts (only impacts as a resource, in impact category resource use). As elementary metals I mean flowables like Iron, or Chromium, or Cobalt, or Lead, etc. Typically these do not occur in their elementary state in nature, for example iron occurs much more frequently in a mineral state such as hematite (Fe2O3). Therefore, what does emission to air / water / ground for e.g. Arsenic mean? Well, it's hard to know, most emissions of elementary metals make no sense, and maybe that is why ReCiPe does not provide emissions to water/air/ground for these elementary metals. However, through a user's perspective it may make sense to add emission of Arsenic to e.g. water, especially if we expect it to come in a certain flavour, such as its ionic state Arsenic(V). That's what ecoinvent does for a lot of metals: they reuse CFs from their ionic state. So here, I propose that we follow Ecoinvent and do the same.

@ErCollao @ganorris