Created by Margret J. Geselbracht, Reed College (mgeselbr@reed.edu) and posted on VIPEr on June 9,
2008. Copyright Margret J. Geselbracht 2008. This work is licensed under the Creative Commons
Attribution Non-commercial Share Alike License. To view a copy of this license visit
http://creativecommons.org/about/license/.
WERNER FROM BEYOND THE GRAVE
Read the following paper before class:
“Crystal Structure Determination of a (µ-Amido)(µ-hydroxo)(µ-superoxo)dicobalt(III) Complex from
the Werner Collection” by Bernhard Spingler, Marie Scanavy-Grigorieff, Alfred Werner, Heinz
Berke, and Stephen J. Lippard, Inorg. Chem. 40, 1065-1066 (2001).
To prepare for discussion, write out the answers to these questions and bring them with you to
conference.
1. In 2001, a paper appeared in Inorganic Chemistry with Alfred Werner as a coauthor (82 years after his
death). After reading this short communication, answer the following questions.
(a) In the first paragraph, the authors state “The nature of the O–O bond in the resulting dinuclear
cobalt complexes was the subject of much controversy.” Describe the different options for how
one might think about the “O2” moiety as a ligand in these complexes and the implications for the
nature of the O–O bond. Describe how you think X-ray diffraction and vibrational studies would
allow the distinction between superoxo and peroxo ligands.
There are several “O2” species differing in charge that could be considered as ligands. These
different moieties have different charges, thus affecting the oxidation state of the cobalt
atoms, and different numbers of bonding and antibonding electrons, thus+affecting the nature
of the O–O bond in the coordinated ligand. The oxygenyl cation, O2 , is probably not a
likely ligand as there would be a natural electrostatic repulsion between the positively
charged ligand and the positively charged cobalt atoms. However, we could think about
–
neutral oxygen, O2, as a ligand with a bond order of 22–or the anionic superoxide ligand, O 2 ,
with an O–O bond order of 1.5 or peroxide ligand, O2 , with an O–O bond order of 1. And
coordinating the ligand to the metal may weaken the O–O bond further depending on the
orbital interactions between the metal and the ligand.
–
The2– most likely choices for the nature of the coordinated “O2” are superoxo, O2 , or peroxo,
O2 . The O–O bond in the superoxo ligand should be slightly stronger (higher bond order)
and shorter than in the peroxo ligand. X-ray diffraction studies could provide a measure of
the O–O bond distance in the coordinated ligand and this could be compared to other bond
distances from known superoxide and peroxide species. Vibrational spectroscopy,
specifically measuring the stretching frequency of the O–O bond, provides an indirect
measure of bond strength. The weaker the bond, the easier it is to stretch the bond, and so
the energy (and frequency) of this bond vibration is observed at lower values. One would
expect the stretching vibration for a peroxo ligand would occur at lower energies
(frequencies) for a peroxo ligand than for a superoxo ligand. Again, comparison to other
compounds would be particularly helpful.
(b) Draw structures of the two proposed binuclear complexes 2a and 2b. Assign oxidation states to
the two cobalt atoms and explain how you arrived at these assignments.
Created by Margret J. Geselbracht, Reed College (mgeselbr@reed.edu) and posted on VIPEr on June 9,
2008. Copyright Margret J. Geselbracht 2008. This work is licensed under the Creative Commons
Attribution Non-commercial Share Alike License. To view a copy of this license visit
http://creativecommons.org/about/license/.
2+
NH3
H 3N
O
O
CoIII
H 3N
H 2O
H 2O
CoIII
N
H
NH3
NH3
3+
H3N
H3N
O
CoIV
H 3N
NH3
O
H
O
NH3
CoIII
NH3
NH3
N
H2
Assigning oxidation states to the cobalt atoms
requires taking into consideration the charges
–
on the nitrate counterions (1- for each NO3 counterion) and the charges on the ligands.
Charge on metals + charge on ligands = overall charge on complex
Starting with compound 2b shown on the right, the overall binuclear complex has a +3
charge due to the three nitrate counterions. The partial structure in the paper labels the cobalt
atoms as Co4+ and Co3+. This is –consistent with netural NH3 ligands,
one bridging hydroxide
2–
(OH–), one bridging amide (NH2 ), and one bridging peroxide (O2 ) ligand.
4 + 3 + (-1) + (-1) + (-2) = +3
Considering compound 2a shown on the left, note that the overall charge on the binuclear
complex is only +2 as there are only two nitrate counterions. The NH3 and H2O ligands are
all neutral. The bridging imido ligand (NH2–) has a -2 charge. If we consider the “ozo”
ligand to be akin to peroxide with a -2 charge as in compound 2b, then both cobalt atoms
would be Co3+. If instead, we consider the “ozo” ligand to be superoxide with a -1 charge,
then we would have a mixed Co2+ / Co3+ compound. And if we consider the “ozo” ligand to
be a bridging neutral O2 ligand, then we would have a dinuclear Co2+ compound.
(c) In the discussion, the authors feel confident about assigning this as a dinuclear Co(III) complex.
Why? What evidence do you think they would have looked for to justify a Co(III)Co(IV) complex?
One way to distinguish a dinuclear Co(III) complex from a mixed Co(III)Co(IV) complex
would be to examine the metal-ligand bond lengths in the crystal structure. The Co3+ ion
should larger than the Co4+ ion (higher Zeff for Co4+) and so the mixed complex should have 2
different types of Co–N bond lengths to the terminal amines, for example. The Co–N bond
lengths to the Co4+ ion should be noticeably shorter than the Co–N bond lengths to the Co3+
ion. A dinuclear Co(III) complex should be symmetric, whereas the mixed complex should
have one cobalt that is markedly different from the other. Examining the data presented in
Table 2, the bond lengths are reasonably symmetric to each cobalt atom and the authors state
that the Co–N and Co–O bond distances agree well with other structures of Co(III)
complexes.
As an aside, these two possible assignments of oxidation states would have very different
implications for magnetic properties as well. One could imagine obtaining a diamagnetic
complex if both cobalts are low-spin Co(III). But the Co(III)/Co(IV) complex must result in
unpaired electrons and paramagnetism regardless of the spin states.
(d) Is there a cobalt-cobalt bond in this molecule? How do you decide?
Created by Margret J. Geselbracht, Reed College (mgeselbr@reed.edu) and posted on VIPEr on June 9,
2008. Copyright Margret J. Geselbracht 2008. This work is licensed under the Creative Commons
Attribution Non-commercial Share Alike License. To view a copy of this license visit
http://creativecommons.org/about/license/.
It is interesting to note in Figure 1 that there is no “bond” or connecting line directly linking
the two cobalt atoms. So, one would reasonably conclude that there is no cobalt-cobalt bond.
But, think about the fact that the authors had to decide whether or not to draw that line in the
first place (and they chose not to). This decision is often based on the distance between the
two cobalt atoms, 2.776 Å in this case. To put this value into perspective, the cobalt-cobalt
distance in the crystalline metal is 2.51 Å, and the covalent radius (1/2 of the distance for a
single bond in a neutral molecule) of cobalt is 1.26 Å for a typical Co-Co single bond
distance of 2.52 Å (http://www.webelements.com/webelements/elements/text/Co/radii.html).
Since the metal-metal distance in this case is significantly longer than a typical Co-Co single
bond, it is unlikely that there is any metal-metal bonding in this molecule.