Electronic spectra of TM complexes

Chapter 13, Atkins and Shriver

Electronic spectra
Transition metal complexes often have absorption bands in the visible region leading to their interesting coloration
These bands come from a combination of d-d (excitation involving just the metal dorbitals) and charge transfer transitions Studying spectra provides information about bonding and structure in these species Transition metal spectroscopic transitions are employed in industrial pigments, display devices, lasers etc

The band labeled CT is a charge transfer transition, all the other absorptions are d-d

MO diagram for an octahedral complex

Possible transitions include ones between d orbitals, between ligand based orbital and dorbitals (LMCT) and between the metal dorbitals and ligand orbitals (MLCT)

d-d transitions
The spectra of complexes with more than one d electron usually contain more than one d-d band
This is because the electrons in an atom or ion feel each others presence and interact
They repel each other

For example, a d2 electron configuration ion gives rise to five different energy terms (five energetically distinct ways of arranging these electrons) even in the case of a free ion/atom
In the presence of a ligand field even more distinct terms are possible leading to many possible transitions

Coupling of angular momenta

For light(ish) elements, the energetically distinct ways of arranging electrons in a free ion are characterized by different orbital (L), spin (S) and total angular momenta (J)
The orbital angular momenta of the individual electrons (l) couple with one another to produce a state with well defined total orbital angular momentum L
For two electrons with angular momentum quantum numbers l1 and l2.
L = l1+l2, l1+l2-1, .., |l1-l2|

The spin angular momenta of the individual electrons (s) couple with one another to produce a state with well defined total spin angular momentum S

Spin orbit coupling

The orbital and spin angular momenta L and S couple together in a vector fashion to give a total angular momentum J
Different values of J have different energies

For L =1 and S =1 we can achieve J values of 2, 1 and 0

For L =2 and S =1 we can achieve J values of 3, 2 and 1

Free ion terms

For a free ion we can label all the energetically distinct ways of arranging electrons using Term Symbols (2S+1)LJ, where S is the total spin angular momentum for the term
(or state), L is the total orbital angular momentum of the state and J is the total angular momentum for the state
Note L is usually denoted by a letter where S, P, D, F etc. correspond to L = 0, 1, 2, 3 etc. 2S+1 is called the multiplicity, singlet, doublet, triplet etc.

Several different terms can arise from a single electron configuration Each term has a different energy

Hunds rules
Hund proposed a set of rules to help decide which term is lowest in energy The ground state term always has the highest possible multiplicity If there are several terms with the same multiplicity then pick the one with the largest value of L

Deriving free ion term symbols

If the electrons are in different type of atomic orbitals all possible combinations of l and ms are possible and life is easy If more than one electron occurs in a given type of orbital (2p2, 3d2, 3d3, 4f4 etc.) then care must be taken as some terms will be forbidden by the Pauli principle

Deriving term symbols 2

If the electrons are different orbitals, all possible terms can be arrived at by considering all possible combinations of l and s for the available electrons
Consider s1d1
Spin S can be + + or +
i.e. S =1 or 0

L can be 2+0 or 2-0 Possible terms are 3D and 1D (neglecting different values of J)

Deriving term symbols 3

Consider p1d1
Spin S can be + + or +
i.e. S =1 or 0

L can be 2+1, 2 or 2-1 (all possibilities between the two extremes derived by adding or subtracting the individual ls) Possible terms are 3F, 3D, 3P, 1F, 1D and 1P

Consider p1d1f1
Spin S can be + + + or + + -
i.e. S = 3/2 or

L can be 3+2+1, 5, 4, 3, 2 ,1, 3-2-1 (all possibilities between the two extremes derived by adding or subtracting the individual ls) Possible terms are 4I, 4H, 4G, 4F, 4D, 4P, 4S and 2I, 2H, 2G, 2F, 2D, 2P, 2S

Microstates
To work out the possible terms for a case where more than one electron is in a given set of orbitals we need to set up a table showing the allowed microstates A microstate is a specified set of values of ml and ms for each electron We then associate the possible microstates with terms

Microstates for atomic carbon

Atomic carbon has a 1s22s22p2 electron configuration. We can set up a chart showing all of the possible ways of arranging the pair of p electrons (showing all the microstates)

Microstates for atomic carbon 2

We can collect these possible microstates into a chart set up according to total MS and ML for each microstate (left below)

From 3P

We now have to figure out which terms these microstates correspond to. We start to do this by considering the microstate with highest MS and highest ML with this MS (in this case MS = +1 and ML = +1). This microstate must belong to a 3P term. We now eliminate from the table all of the other microstates that would come from this term and work on what is left

Microstates for atomic carbon 3

We now examine the microstates that are left after eliminating the first term and try and figure out the next term and so on.
MS +1 +2 +1 ML 0 -1 -2 0 X X XX X X ML -1 +2 +1 0 -1 -2 +1 MS 0 X X X X X ML -1 +2 +1 0 -1 -2 X +1 MS 0 -1

The remaining microstates (above left) after eliminating those from the 3P term can be broken down into a set that must come from a 1D term (middle) and another set that must belong to a 1S term (right)

Energies of different carbon terms

Note the large energy differences that can arise between terms just due to electronelectron repulsion. Differences due to changes in J are small compared to those due to changes in S and L. - This is generally true for light atoms

Microstates for a d2 configuration

We can go through a similar process of examining microstates for a d2 electron configuration

Microstates in () come from 3F term, microstates in [] come from a 3P term, microstates in a box come from a 1G term and microstates in a circle come from 1P term. There is one other microstate that comes from 1S term

The Racah repulsion parameters

The different free ion terms for an electron configuration have different energies due to variations in electron-electron repulsion The different energies can be expressed using a small number of electrostatic parameters, A, B and C
These parameters are integrals related to the extent of electron-electron repulsion. The larger they are, the greater the repulsion is

A, B and C are called Racah parameters

Energies of d2 free ion terms

E(1S) = A + 14B + 7C E(1G) = A + 4B + 2C E(1D) = A - 3B + 2C E(3P) = A + 7B E(3F) = A - 8B Note that the difference between any pair of these terms is purely a function of B and C, not A If we can measure the energies of two properly chosen spectroscopic transitions between these terms we can calculate B and C

Values for Racah parameters

The Racah parameters depend on ion size
The smaller the ion, the larger B and C get. Essentially, as you confine the electrons to a smaller volume they repel each other more

The ratio C/B is almost constant and close to 4.

Ligand field transitions

When a spherical ion is placed in a lower symmetry environment the degeneracy of some of the terms is broken For an octahedral field
S ---> A1g P ---> T1g D ---> Eg + T2g F ----> T1g + T2g + A2g

Transitions from the ground state to all of these new states are possible
This leads to quite complicated spectra

Predicting the loss of degeneracy

The way in which the degeneracy of a state is broken by lowering the symmetry of a species can be predicted using group theory
Going from a spherical ion (point group R3) to an octahedral complex is just a symmetry lowering as far as group theory is concerned

The results from these predictions are tabulated in tables for descent of symmetry

Descent of symmetry tables

d1 correlation diagram for Oh

The energetic effects of the loss of degeneracy generated by symmetry lowering can be illustrated on a correlation or Orgel diagram
Here the ground state, 2D term, of a d1 isolated ion is split into two different states by the octahedral field

d2 correlation diagram for Oh

The free ion terms for a d2 metal are also split by insertion into an octahedral field
3F into 3A2g, 3T2g and 3T1g 3P into 3T1g (no split)

We can work out the electron configurations corresponding to these states using group theory
Note that on the left hand side of the diagram (weak field limit, is extremely small relative to B and C) the splitting is expressed purely in terms on B and C. On the right hand side of the diagram (strong field limit, is very big compared to B and C) the splittings can be expressed purely in terms of . In the middle the diagram, the relative energies of the states depend on B, C and .

Deriving the possible states from the strong field limit

For two parallel spin d electrons in an octahedral complex we can have three different electron configurations
t2g2 eg2 ------> 3T1g
3T 2g 3A 2g

t2geg ------> ------>

+ 3T1g

Possible states are derived using group theory using direct products or determined by examination of a direct product table

Direct product table

The possible states that you can get by having electrons in two different orbitals can be read off from a direct product table
To use this table select the orbital symmetry of one electron across the top and orbital symmetry of the second electron down the side. At the intersection of the row and column you are given the allowed symmetries of the states involving the pair of electrons. For example, if you have two t2 electrons, you can get a 3T1 state and 1A1, 1E and 1T2 states. If you have an e and a t2 electron you can 1T1, 1T2, 3T1 and 3T2 states.

Tanabe -Sugano diagrams

Correlation diagrams are not quantitative. They just show in a general sense what happens to the energies of terms as a function of Tanabe-Sugano diagrams illustrate the relative energies of states in a quantitative fashion
They are drawn for a fixed ratio of B/C (~4) The ground state always lies on the x-axis They can be used along with spectra to estimate both and B for a given complex.

Tanabe Sugano diagrams

Tanabe-Sugano diagrams include the effect of mixing states with the same symmetry
Electronic states with the same symmetry can not cross they always mix This introduces the curvature seen for many of the lines on the plot

Using a TS diagram
We can use a TS diagram by first trying to assign the observed spectrum (decide which transitions are responsible for the bands)

First two bands in V3+ spectrum probably involve transitions to the 3T2 and 3T1 states (3A2 also seems possible, but does not work out in the later part of the analysis!). 1E and 1T2 are not likely as the final states, as transitions involving a spin state change are extremely weak (as in the Mn2+ spectrum)

Using a TS diagram 2
Having identified two bands take the ratio of their transition energies (~25500/17000 =1.5) and find the point along the /B axis on the TS diagram that corresponds to this ratio (/B~27 gives the right ratio) Using this /B value, for one of the transitions read off across onto E/B axis. This gives you E/B for this transition. As you know E you can calculate B for the complex and you can then use this to calculate
E/B ~25.5 for the first band, suggesting that B is ~665 cm-1. This would give ~ 17950 cm-1

Tanabe Sugano diagrams

d2 and d3

Tanabe Sugano diagrams

d4 and d5

Tanabe Sugano diagrams

d6 and d7

Tanabe Sugano diagrams

d8

High-spin low-spin transitions and Tanabe Sugano diagrams

Note that the abrupt change in slope at about /B ~ 22 occurs because the ground state changes from 4T1 to 2E and everything is plotted relative to the energy of the ground state
This is a high-spin to lowspin transition occurring as the ligand field strength is increased!

The nephelauxetic effect

The Racah repulsion parameters for a metal complex vary as the ligand is changed
As the complex becomes more covalent the electrons are to some extent spread over the ligands so the electron-electron repulsion is reduced This reduction is repulsion as covalency increases is called the nephelauxetic effect (literally cloud expanding)

A nephelauxetic series can be set up based on , the nephelauxetic parameter = B(Complex) / B(free ion)

F- > H2O > NH3 > CN- > Br This series is consistent with fluoride complexes being the most ionic and bromide complexes the most covalent

Distortions
If a complex distorts from regular octahedral geometry this will show up as extra bands or shoulders on bands
Perfectly octahedral [Ti(H2O)6]3+ should only give one d-d transition Band has structure indicating distortion due to the JahnTeller effect. Geometry changes to eliminate the degeneracy of the 2T2g ground state

Charge transfer transitions

In addition to seeing transitions between states that are essentially based on metal d-orbitals you can get transitions that involve ligand based orbitals and metal d-orbitals
Called charge transfer (CT) transitions as an electron is transferred from the metal to the ligand or vice-versa Charge transfer transitions often change in energy as the solvent polarity is varied (solvatochromic) as there is a change in polarity of the complex associated with the charge transfer transition
This can be used to distinguish between d-d transitions and charge transfer bands

LMCT
Ligand to metal charge transfer (LMCT) bands are responsible for the colors of species like CdS and MnO4The energy of a charge transfer band depends upon the energy gap between the metal (acceptor) and ligand (donor) orbitals
this is a function of the metal the oxidation state and the ligand

Energies MnO4- < TcO4- < ReO4-

Optical electronegativities
The energies of the ligand and metal orbitals involved in a CT transition depend on the electronegativity of the ligands and metal
So the energy of a charge transfer band depends upon this difference in electronegativity For a species such as [CrX(NH3)5]2+ (see below). The energy of the LMCT band will change as the halide X is changed

Optical electronegativity 2
As the energy of a charge transfer band depends upon the difference in electronegativity between ligand and metal we can define optical electronegtivities based on
~ C (ligand - metal)
C is a constant chosen to give values for that are approximately the same as the Pauling electronegativities

Note we have separate values for and . This is because the symmetry ligand based orbitals are lower in energy than the symmetry lone pairs on these ligands (see Fig 13.8)

MLCT transitions
Complexes that have ligands with low lying * orbitals often show MLCT transitions They are very common with complexes involving bipy and phen

Resonant Raman
The nature of the ligands involved in a charge transfer transition can be elucidated using resonant raman spectroscopy
If the incident laser frequency is chosen to be coincident with the charge transfer band energy (resonant), the intensity of some of the raman active vibrational modes is enhanced
The vibrations that have an enhanced intensity are ones that involve motion of the ligands associated with the charge transfer transition. The above resonant Raman data shows that both the phenanthroline and the CO ligands are involved in a MLCT transition for the above complex as vibrations for both of these ligands gain intensity when the Raman spectrum is collected on resonance

Raman spectroscopy
Laser excitation source is used. Intensity of inelastically scattered photons is measured

Transition intensity
Not all electronic transitions in coordination complexes have equal intensity
Charge transfer bands are often very strong
~ 1000 - 50000 L mol-1 cm-1

Transitions involving final and initial states with different spins (e.g. Mn2+) are very weak ( < 1 L mol-1 cm-1) for first row transitions metals and weak for heavier transition metals d-d transitions are intermediate in intensity

How do we understand these differences?

centrosymmetric complexes ~ 20 - 100 L mol-1 cm-1 others, ~ 250 L mol-1 cm-1

Selection rules - spin

There are quantum mechanical selection rules governing how probable (intense) a transition is going to be As photons do not have spin they can not in general induce transitions that involve a change in spin (S). Hence, transitions that involve a change in S are very weak spin forbidden
However, this rule can broken if there is strong coupling between orbital and spin angular momentum (spin-orbit coupling) and for heavy elements this coupling is large leading to some intensity for spin forbidden transitions

Selection rules - Laporte

The intensity of a transition depends on the symmetry of the states involved. For centrosymmetric species (free ions and octahedral complexes)
g g transitions are forbidden u u transitions are forbidden u g and g u transitions are allowed The above rules are the Laporte selection rules This means that all d-d transitions in octahedral complexes are formally forbidden as they are g g !
Hence all d-d transitions in octahedral complexes are weak

For tetrahedral complexes there is no center of symmetry and this selection rule does not strictly apply
d-d transitions in tetrahedral complexes are typically stronger than those in octahedral complexes

Why does symmetry mater?

For a transition involving the absorption of a photon (IR, optical etc.) there must be a change in dipole moment on going from the ground to the excited state. This occurs when the following integral has a none zero value:

1 2 d

1 and 2 are the ground and excited state wavefunctions, is the dipole moment operator and the integral is over all space

This integral can only be none zero if the function being integrated has a component with symmetry like that of the highest symmetry irreducible representation of the molecules point group As transforms like the functions x, y and z, (u symmetry for a centrosymmetric species) this in practice means that you only observe a transition when 1 and 2 differ in symmetry, with respect to being centrosymmetric (g) or anticentrosymmetric (u), as the overall product 12 must have g symmetry.

Selection rules CT bands

As long as a charge transfer band does not violate the no change in spin rule it is likely to be very strong as there no other demanding restrictions on its occurrence

Vibronic coupling
If d-d transitions in centrosymmetric complexes are forbidden by the Laporte selection rule why do we see them?
Complexes undergo vibrational motion. Some of these vibrational modes can distort the complex so as to remove its center of symmetry Coupling between modes that break the center of symmetry and the electronic transition leads to the transitions being weakly allowed (essentially we excite the complex to a final state that is both electronically excited and vibrationally excited in such as way that the vibration removes the center of symmetry)
This interaction is called vibronic coupling Vibronic coupling leads to the rather broad nature of most d-d transitions
You excite into a range of different vibrational levels and this range of levels adds breadth to the transition

Luminescence
When a species is excited by absorption of a photon it can loose energy by
nonradiative decay radiative decay (luminescence)

There are two types of radiative decay

fluorescence, spin allowed emission phosphorescence, spin forbidden emission

Luminescent behavior can lead to technological applications

Lasers and phosphors

Ruby Laser
The first optical laser to be developed was based on luminescence from Ruby crystals
Ruby is Al2O3 doped with Cr3+. Transitions associated with the octahedrally coordinated Cr3+ are responsible for the laser action Poplulation inversion is achieved by exciting to 4T2g or 4T1g. Internal conversion and intersystem crossing leads the population of 2Eg states. These are long lived as radiative decay to the ground state is spin forbidden. Putting the crystal in a laser cavity can lead to lasing from the 2Eg to 4A2g transition.

Lanthanide compounds
Lanthanide compounds are widely used as phosphors (long lifetime luminescent material) The spectra of lanthanide compounds are very similar to those of the free ions as the ligands interact very weakly with the f-electrons
Very small ligand field splittings

This leads to weak transitions and bands that are very narrow (weak vibronic coupling)
Get rather pure colors along with pastel shades

The Cotton effect

The Cotton effect includes Circular Dichroism (CD) and Optical Rotatory Dispersion (ORD) ORD is the change in optical rotation from a chiral species as function of wavelength A CD spectrum gives the difference between the absorption coefficients for left and right circularly polarized light as a function of wavelength
CD spectrum only exists for chiral species

The polarization of light

Electromagnetic radiation can be polarized
either in a plane
Used for measuring ORD

or elliptically/circularly
Used for measuring CD spectra

Use of CD and ORD spectra

CD and ORD spectra can be used to identify which enantiomer of a chiral species you have
Compounds with similar structures give similar ORD/CD spectra if they have the same absolute configuration

CD and ORD spectra can provide slightly better effective spectral resolution than an absorption spectrum and can be used to aid band assignment

The CD spectrum of Co(en)33+

The CD spectra suggest the presence of another absorption band at ~24,000 cm-1 that is not obvious from the absorption spectrum

Assigning absolute configuration

Tris chelate complexes prepared using the amino acid alanine exist as two geometrical isomers both of which are chiral The absolute configuration of the optically pure alanine complexes can be assigned by comparing their CD spectra to those for optically pure [Co(en)3]3+
Note the absolute configuration of [Co(en)3]3+ had to be determined first for this to work!

CD spectrum for (-)[Co(edta)]-

Assign as based on comparison with [Co(en)3]3+ CD spectra

Relationship between ORD and CD

ORD and CD spectra are related to one another mathematically
Maxima in the CD spectra correspond to first derivative like shapes in an ORD spectrum of the same species

ORD spectra for cobalt complexes

ORD spectra can also be used to assign absolute configuration by comparison with reference compounds
Note that these spectra emphasize the importance of specifying your wavelength when you talk about a simple optical rotation measurement!

EPR or ESR spectroscopy

Electron spins behave in a similar way to nuclear spins when in a magnetic field
You can flip the electron spin magnetization (if there is any S>0) using radiation and do EPR (electron paramagnetic resonance) which is the electron equivalent of NMR

Typical frequencies of ESR transitions are much higher than those for NMR transitions (9 GHZ for a 0.3 T magnetic field) Transition energy is given by E = g BB
g = 2.0023 for a free electron, it is a kind of chemical shift for the electron
B is the applied magnetic field, B is a fundamental constant

Hyperfine structure
Nuclear spins in a compound can couple to the electron spin
Coupling between electron and nuclear spins leads to splittings effectively because the nuclear spin changes the magnetic field experienced by the electron
h = gB(B + AmI) A coupling constant, measures strength of coupling

Coupling only occurs if the electron spin is delocalized into the part of the molecule where the nuclear spin is located Simple example: for SiH3 radical a electron on the silicon can couple to all three protons leading to a quartet
4% 29Si (I = ) also couple giving pairs of satellites around the main peaks

The strength of the coupling (value of A) can be used as measure of how delocalized the electron is
For Cu(II) complexes the following values of A (coupling between Cu and the electron spin) were determined for different ligands Ligand A/mT F9.8 H2O 9.8 O28.5 S26.9

Coupling constant drops as the electron is better delocalized off the Cu(II)

Complex example of hyperfine splitting

63Cu,

I = 3/2, 14N, I = 1 Basic spectrum has 4 lines due to coupling with Cu. Each line is further split by coupling to 14N and H Substitution of NH with D does not change spectrum
No delocalization onto N-H!

Substitution of CH to give CMe does change the spectrum

Collapses to 5 lines in each component due to coupling with two equivalent I = 1 14N Indicates delocalization onto CH groups

Bonding and spectra in ML5 species

An understanding of the bonding and spectra of a complex with ML5 stoichiometry can be gained by thinking about what happens when a ligand is removed from an octahedral species

Shape of ML5 species

The shapes of ML5 complexes can sometimes be predicted by considering an orbital correlation diagram
Note that we would predict a flat base square pyramidal geometry for species with 6 d electrons or less, but some distortion once the dz2 orbital becomes occupied

Mixed valence compounds

Mixed valence species are often encountered in both the solid state molecules in solution
Fe3O4, Mn3O4, GaCl2, CsAuCl3, AgO, Ag2F, [Ta6Cl12]3+

Mixed valence II/III

Nature of mixed valence compounds

A interesting question with mixed valence species is should be describe the compound as Mn+Mm+ (two distinct oxidation states) or M[(n+m)/2]+ M[(n+m)/2]+ (each metal has the average oxidation state)?
Both possibilities occur in real compounds and they lead to different physical properties Compounds with two distinct metal environments often show pronounced intravalence charge transfer bands that can lead to intense coloration
In Prussian blue (~Fe4[Fe(CN)6]3.xH2O), the blue color arises from transitions involving the transfer of an electron from Fe(II) to Fe(III) These kinds of charge transfer bands can only occur when there is electronic coupling between the two metal sites (sites must be near to one another and linked by something)

Robin - Day classification

Mixed valence compounds were broken into three classes with different properties depending upon the local environments of the element with mixed valence