6 Conclusion and Outlook
The idea of inventing a simple spectroscopic means in a form of the symmetric CO
stretching frequency (TEP) to describe the metal–ligand bond was appealing as it
could be used to single out those ligands that can be easily replaced in the course of a
catalytic reaction. This was an important step forward at a time where only CO but
not ML stretching frequencies were accessible to vibrational spectroscopy. However, there are three fundamental problems that make the TEP questionable, which
may be even a misleading parameter: (1) As shown in this article, the A 1 symmetric
CO stretching frequency does not describe all aspects of the complex ML interactions, reflected by a redshift which is too low. (2) There is mode–mode coupling
between CO and MC stretching vibrations that leads to coupling errors in the range
of (20–200 cm
À1 ) [131, 274], which can be larger than the variation in the TEPs
caused by different ligands L. (3) The relationship between ML bond strength and
TEP, considered by Tolman as a key electronic feature of transition metal–carbonyl
complexes (CO) n ML m , is generally neither quantitatively nor qualitatively fulfilled,
because the electronic ML bonding mechanism is much more complex than
described by the TEP [130]. This leads to the following conclusions:
1. The basic idea of Tolman was and still is a valuable one. Vibrational spectroscopy, in the form of infrared, Raman, or the more modern terahertz spectroscopy,
provides sensitive tools to describe the electronic structure of any transition metal
complex or any catalyst in general. However, there is no longer a need to refer to
the CO stretching vibration, since one can access nowadays the ML stretching
vibration directly. There is only the necessity to convert measured normal mode
frequencies into local mode frequencies and then derive all other local mode
properties needed, especially the local stretching force constants that reflect in a
universal way the intrinsic strength of any bond.
2. The local stretching force constants k
a (ML) can be easily converted into bond
strength orders (BSO) n, which provide a useful ordering of chemical bonds
according to their strength. If the BSO n is used as MLEP instead of k
a (ML),
MLEP values can be easily compared from one ML bond type to the other. This
procedure can be carried out with measured or calculated data, where in the
former case a conversion from normal mode into local mode frequencies can be
carried out using the procedure of Cremer and co-workers [76]. In addition,
chemically meaningful reference molecules with known BSO n have to be
utilized in order to determine the BSO n values from a power relationship. As
shown in this work, Mayer bond orders are the best choice for metals and
transition metals.
3. The MLEP covers all electronic effects of ML bonding including steric effects.
Therefore, no Tolman cone angle is needed. However, steric effects can be
isolated via local bending force constants and corresponding bending orders.
We are currently compiling a library of MLEP values for ML bonds across the
periodic table, some of which are summarized in the Appendix. Work is also in
progress to release of first open-source version of the local mode analysis program.
Characterizing the Metal–Ligand Bond Strength via Vibrational Spectroscopy:. . .
259
The idea of inventing a simple spectroscopic means in a form of the symmetric CO
stretching frequency (TEP) to describe the metal–ligand bond was appealing as it
could be used to single out those ligands that can be easily replaced in the course of a
catalytic reaction. This was an important step forward at a time where only CO but
not ML stretching frequencies were accessible to vibrational spectroscopy. However, there are three fundamental problems that make the TEP questionable, which
may be even a misleading parameter: (1) As shown in this article, the A 1 symmetric
CO stretching frequency does not describe all aspects of the complex ML interactions, reflected by a redshift which is too low. (2) There is mode–mode coupling
between CO and MC stretching vibrations that leads to coupling errors in the range
of (20–200 cm
À1 ) [131, 274], which can be larger than the variation in the TEPs
caused by different ligands L. (3) The relationship between ML bond strength and
TEP, considered by Tolman as a key electronic feature of transition metal–carbonyl
complexes (CO) n ML m , is generally neither quantitatively nor qualitatively fulfilled,
because the electronic ML bonding mechanism is much more complex than
described by the TEP [130]. This leads to the following conclusions:
1. The basic idea of Tolman was and still is a valuable one. Vibrational spectroscopy, in the form of infrared, Raman, or the more modern terahertz spectroscopy,
provides sensitive tools to describe the electronic structure of any transition metal
complex or any catalyst in general. However, there is no longer a need to refer to
the CO stretching vibration, since one can access nowadays the ML stretching
vibration directly. There is only the necessity to convert measured normal mode
frequencies into local mode frequencies and then derive all other local mode
properties needed, especially the local stretching force constants that reflect in a
universal way the intrinsic strength of any bond.
2. The local stretching force constants k
a (ML) can be easily converted into bond
strength orders (BSO) n, which provide a useful ordering of chemical bonds
according to their strength. If the BSO n is used as MLEP instead of k
a (ML),
MLEP values can be easily compared from one ML bond type to the other. This
procedure can be carried out with measured or calculated data, where in the
former case a conversion from normal mode into local mode frequencies can be
carried out using the procedure of Cremer and co-workers [76]. In addition,
chemically meaningful reference molecules with known BSO n have to be
utilized in order to determine the BSO n values from a power relationship. As
shown in this work, Mayer bond orders are the best choice for metals and
transition metals.
3. The MLEP covers all electronic effects of ML bonding including steric effects.
Therefore, no Tolman cone angle is needed. However, steric effects can be
isolated via local bending force constants and corresponding bending orders.
We are currently compiling a library of MLEP values for ML bonds across the
periodic table, some of which are summarized in the Appendix. Work is also in
progress to release of first open-source version of the local mode analysis program.
Characterizing the Metal–Ligand Bond Strength via Vibrational Spectroscopy:. . .
259