coordination compounds - Why are negatively charged ligands behind the neutral ligands in spectrochemical series?

I had doubt that shouldn't negatively charged ligands be better electron pair donors than neutral ligands like $\ce{CO}$ or $\ce{PPh3}$. But its not so. Why?

Answer

I think you are confused by what the spectrochemical series actually shows. It is not a rating of stronger or less strong ligands to mean they can bind more or less strongly to the metal. It is only a rating of which ligands induce a high crystal field split in the resulting complex.

The splitting derives from the metal $\mathrm{e_g}$ orbitals — $\mathrm{d}_{x^2 - y^2}$ and $\mathrm{d}_{z^2}$ — actually being antibonding to the metal-ligand interaction and thus higher in energy than the (unmodified in a complex with only σ type interactions) $\mathrm{t_{2g}}$ orbitals.

For systems more complex, i.e. including π interactions, the $\mathrm{t_{2g}}$ orbitals (the other three d-orbitals) interact in a π manner with ligand p-orbitals. If the ligand has populated p-orbitals such as halides have, the $\mathrm{t_{2g}}$ energy is raised as it is turned into a slightly antibonding interaction (the bonding one being on the ligands’ side). Since this brings $\mathrm{t_{2g}}$ closer to $\mathrm{e_g}$, the field split is reduced and the ligands are termed weak-field ligands. In simplified Lewis schemes, one could represent this by a double forwards bond from the ligands to the metal.

In the case of empty p-type orbitals such as in $\ce{CO}$, metal to ligand back-bonding can occur by populated metal orbitals interacting with unfilled ligand orbitals; now it is the metal’s side of things that is stabilised. Since this π backbonding stabilises the $\mathrm{t_{2g}}$ orbitals, their energy is lowered and the field split becomes greater. Hence they are termed strong-field ligands. In simplified Lewis structures one could represent that by drawing a donor-acceptor arrow from the ligand to the metal (σ forwards bond) and one the other way (π back bond).

This all has nothing to do with the overall strength of metal-ligand interaction, as I mentioned earlier. In fact, strong complexes are typically obtained for certain metal-ligand combinations and the same ligand can be very bad in interacting with other metals. Classic example: $\ce{CO}$ ligands are extremely good at stabilising low metal oxidation states such as in $\ce{[Fe(CO)4]^2-}$, while they are rather lousy in stabilising high ones: $\ce{[Fe(CO)6]^2+}$ can only be synthesised in the extreme absence of all other ligands, i.e. in $\ce{SbF5}$ from $\ce{FeF2}$ where the fluorides will be drawn towards antimony to generate the even more stable $\ce{[SbF6]-}$ complex.