Why do aromatic compounds have an upfield shift upon coordination to metals e.g. ferrocene? The 1H NMR shows a single environment at ~4.1 ppm. This is in contrast to the usual aromatic shift of between 6-9 ppm. At the moment, I have the following rationale, although I am not convinced about 2/3:

1) Coordination of the aromatic group to the Fe essentially localises the double bonds, as Fe will backbond into the aromatic $\pi$*-orbitals.

2) Bonding of the aromatic group to the Fe will disrupt the ring current which would also cause an upfield shift. However, there seem to be several papers online which suggest the presence of ring currents even in sandwich complexes?

3) There are 6 $\pi$-electrons for only 5 carbons in Cp, therefore there are essentially 1.2 $\pi$-electrons per Cp carbon whereas there is 1 $\pi$-electron for each carbon in Cp. Therefore Cp is more shielded than benzene naturally and will have a lower shift.

  • 1
    $\begingroup$ Your "3)" goes in more or less right direction, but what you wrote is rather dire simplification of what needs MO theory to be explained properly, I think. Definitely not 1) or 2) $\endgroup$
    – Mithoron
    Commented Feb 16, 2019 at 23:10
  • $\begingroup$ Hi, thanks for the reply. How could you rationalise the fact that aromatic groups such as benzene also go to a lower shift upon coordination to metal ions then? $\endgroup$
    – K.P.
    Commented Feb 17, 2019 at 11:29
  • $\begingroup$ There are bonds between HOMO and LUMO of the aromatic ligand and metal. Metal and ligand donate electron density to each other and ligand may wind up with less ten it started. $\endgroup$
    – Mithoron
    Commented Feb 17, 2019 at 17:26
  • 2
    $\begingroup$ You should compare apples and apples. The $\delta$ ($\ce{CDCl3}$) for benzene is 7.36 ppm and that for bis(benzene)chromium is 7.16 ppm. reference $\endgroup$
    – Zhe
    Commented Feb 22, 2019 at 12:55

1 Answer 1


Chemical shift data dug up for a number of iron compounds in the website of Hans Reich reveals that the shift is not unique to aromatic compounds, but rather appears due to Fe(2+):

enter image description here

The base shift for a vinyl proton (C=C-H) is 5.25 ppm.

The following is a series of spectra from magritek, illustrating the effect of Fe(2+) complexation on the shift of the cyclopentadienyl ligand.

enter image description here

Sandwich compounds based on magnesium do not show such large upfield shifts:

enter image description here

This seems to eliminate your conjecture 3 as an explanation, since cyclopentadienyl is aromatic and addition of Fe(2+) causes an upfield shift, whereas Mg(2+) does not cause such a large one.

Now Fe(2+) complexes may be paramagnetic or diamagnetic depending on the the ligand (low/high field). Paramagnetic metals have unpaired electrons that can induce strong dipolar magnetic fields. In porphin, for instance, paramagnetic Fe(3+) results in strong shielding in the aromatic ligand, with one proton shifted upfield ~20 ppm. A diamagnetic Zn(porphin) complex does not show such strong shielding. Ferrocene is diamagnetic.

The quick and sloppy answer therefore appears to be that the shift observed in ferrocene is mainly due to Fe(2+), even though the compound as a whole is diamagnetic.

Regarding the bonding arrangement, the wikipedia has this to say:

In terms of bonding, the iron center in ferrocene is usually assigned to the +2 oxidation state, consistent with measurements using Mössbauer spectroscopy. Each cyclopentadienyl (Cp) ring is then allocated a single negative charge, bringing the number of π-electrons on each ring to six, and thus making them aromatic. These twelve electrons (six from each ring) are then shared with the metal via covalent bonding. When combined with the six d-electrons on Fe2+, the complex attains an 18-electron configuration.

To add a little more detail, Mulay and Fox hypothesize on the nature of metal-ligand bonding in ferrocene based on the magnetic anisotropy:

The anisotropy between K3 and the average of K1 and K2 was found to be $\pu{49.5×10—6 cgs units}$, and is evidence in favor of a single dπ—pπ bond between each of the rings and the iron atom. Pauling's "resonance structures" also seem plausible; the studies do not support Fischer's "complete π‐electron donation" structure. Theoretical calculations based on Langevin's formula for diamagnetism show a close agreement with the experimental value.

A more detailed account of bonding in ferrocene can be found in Miessler and Tarr's textbook.

[6]: Magnetic Anisotropy of Ferrocene and Chemical Bonding, J. Chem. Phys. 38, 760 (1963); https://doi.org/10.1063/1.1733734, L. N. Mulay and S.N.D. Mary Eleanor Fox


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.