Geometric and electronic properties in a series of phosphorescent heteroleptic Cu(I) complexes: Crystallographic and computational studies

Article abstract of DOI:10.1016/j.poly.2016.12.035

We have investigated the electronic and geometric structures in the lowest excited states of six phosphorescent heteroleptic [Cu^I(NN)(DPEphos)]⁺(DPEphos?=?bis[(2-diphenylphosphino)phenyl]ether) complexes with varying NN?=?diimine ligand structures using density functional theory. In comparison to the ground state, the results show a decrease of the dihedral angle between the N–Cu–N and P–Cu–P planes for these excited states with mixed ligand-to-ligand (DPEphos lone pair?→?π^?(NN)) and metal-to-ligand charge transfer (dπ(Cu)?→?π^?(NN)) character. Sterically less demanding ligands facilitate this process, which is accompanied by a geometric relaxation of the DPEphos ligand and contraction of the Cu–N bonds. The density functional for the excited state calculations has been selected based on ground state validation studies. We evaluated the ability of seven density functionals to reproduce the molecular ground state geometries and absorption spectra obtained by single-crystal X-ray diffraction and solution-phase UV–Vis absorption spectroscopy respectively. Standard methods (PBE and B3LYP), which do not account for dispersion, systematically overestimate internuclear distances. In contrast, approaches including dispersion (B97D3, PBE0-GD3, M06L, M06, ωB97XD) remove this systematic effect and give less expanded molecular structures. We found that only the hybrid functionals (B3LYP, PBE0-GD3, M06), incorporating a portion of exact exchange from Hartree–Fock theory, accurately predict the experimental absorption energies.

Full text of DOI:10.1016/j.poly.2016.12.035

Accepted Manuscript
Geometric and electronic properties in a series of phosphorescent heteroleptic
Cu(I) complexes: Crystallographic and computational studies
Katharina Kubiek, Sreevidya Thekku Veedu, Darina Storozhuk, Reza Kia,
Simone Techert
PII:
S0277-5387(16)30686-6
DOI:
http://dx.doi.org/10.1016/j.poly.2016.12.035
POLY 12393
Reference:
Polyhedron
To appear in:
Received Date:
Revised Date:
Accepted Date:
23 August 2016
11 December 2016
18 December 2016
Please cite this article as: K. Kubiek, S. Thekku Veedu, D. Storozhuk, R. Kia, S. Techert, Geometric and electronic
properties in a series of phosphorescent heteroleptic Cu(I) complexes: Crystallographic and computational studies,
Polyhedron (2016), doi: http://dx.doi.org/10.1016/j.poly.2016.12.035
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Geometric and electronic properties in a series of
phosphorescent heteroleptic Cu(I) complexes:
Crystallographic and computational studies
Katharina Kubi eka,b,∗, Sreevidya Thekku Veedua,b, Darina Storozhuka,b,c, Reza Kiaa,b,1,∗, and
Simone Techerta,b,c,
∗
aFS-SCS, Deutsches Elektronen-Synchrotron (DESY), Notkestraße 85, 22607 Hamburg,
Germany
bMax Planck Institute for Biophysical Chemistry, Am Faßberg 11, 37077 Göttingen,
Germany
cInstitute of X-ray Physics, Georg-August-Universität Göttingen, Friedrich-Hund-Platz 1,
37077 Göttingen, Germany
∗Corresponding authors. E-mail addresses: katharina.kubicek@desy.de (K.K.);
rkia@sharif.edu (R.K.); simone.techert@desy.de (S.T.).
1Present address: Chemistry Department, Sharif University of Technology, P.O. Box 11155-
3516, Tehran, Iran
1

ABSTRACT
We have investigated the electronic and geometric structures in the lowest excited states of
six
phosphorescent
heteroleptic
[CuI(NN)(DPEphos)]+
(DPEphos
=
bis[(2-
diphenylphosphino)phenyl]ether) complexes with varying NN = diimine ligand structures
using density functional theory. In comparison to the ground state, the results show a
decrease of the dihedral angle between the N-Cu-N and P-Cu-P planes for these excited states
with mixed ligand-to-ligand (DPEphos lone pair • *(NN)) and metal-to-ligand charge
transfer (d (Cu) • *(NN)) character. Sterically less demanding ligands facilitate this process,
which is accompanied by a geometric relaxation of the DPEphos ligand and contraction of
the NN ligand. The density functional for the excited state calculations has been selected
based on ground state validation studies. We evaluated the ability of seven density
functionals to reproduce the molecular ground state geometries and absorption spectra
obtained by single-crystal X-ray diffraction and solution-phase UV-VIS absorption
spectroscopy respectively. Standard methods (PBE and B3LYP), which do not account for
dispersion, systematically overestimate internuclear distances. In contrast, approaches
including dispersion (B97D3, PBE0-GD3, M06L, M06, ωB97XD) remove this systematic
effect and give less expanded molecular structures. We found that only the hybrid functionals
(B3LYP, PBE0-GD3, M06), incorporating a portion of exact exchange from Hartree-Fock
theory, accurately predict the experimental absorption energies.
Keywords:
Bisphosphines
Copper(I)
Density functional
Crystal structure
Excited state
2

1. Introduction
Transition metal coordination complexes often show advantageous properties, such as a
strong visible ground state absorption and excited state emission for their application in dye-
sensitized solar cells [1,2], chemo- and biosensors [3-5] as well as organic light emitting
diodes (OLED-s) [6,7]. Consequently, the design and synthesis of transition metal complexes
has become a more and more active field of research [8,9]. As 4d and 5d metals are relatively
expensive and not abundant, 3d transition metal compounds and in particular d10 Cu(I) based
complexes have been discussed as alternatives [10-14]. It is important to note that for Cu(I)
complexes, excited state characteristics such as lifetimes and luminescence quantum yields
are extremely sensitive to variations in their respective ligand structure [15], which in
principle allows tuning for applications [16,17]. However, a detailed understanding of
universal correlations between their ground and excited state properties is necessary for the
development of more efficient sensitizers and phosphorescent materials [18-20].
Therefore, numerous density functional theory (DFT) calculations and experiments have
investigated the excited states of homoleptic prototype [Cu(NN)2]+ complexes with two
diimine ligands (NN), in particular [Cu(dmp)2]+ (dmp = 2,9-dimethyl-1,10-phenanthroline)
[21-35]. The results indicate that photoexcitation of these complexes leads to a singlet metal-
to-ligand charge transfer (1MLCT) excited state and a subsequent ligand-dependent decrease
in the dihedral angle (dha) between the NN ligand planes in the excited states. The geometry
change is accompanied by an increase in the non-radiative decay, in particular for the
energetically lowest triplet charge transfer (3CT) state after intersystem crossing [36-40].
Those findings have stimulated the design and synthesis of Cu(I) complexes with more
constraining ligands, which may hamper a ligand torsional motion and improve the excited
state characteristics for applications. For systems with the phen* ligand (phen* = 1,10-
phenanthroline derivate), substituents in the 2,9-positions of the phen* unit were found to be
most effective [11,41,42]. Moreover, heteroleptic complexes with more bulky ligands show
3

extraordinary properties. Extreme examples are [Cu(NN)(PP)]+ (PP = diphosphine) and in
particular [Cu(phen*)(DPEphos)]+ (DPEphos = bis[(2-diphenylphosphino)phenyl]ether)
complexes [17,43-50]. In comparison to analogous homoleptic complexes, the radiative
lifetimes and luminescence quantum yields of their lowest excited states are enhanced by
orders of magnitude and the complexes have been already successfully tested for OLED and
light-emitting electrochemical cell applications [17,43-50]. Also for this class of compounds,
a minor variation in the substituent of the 1,10-phenanthroline (phen) moiety, in particular at
the 2,9-positions, can change the excited state parameters by factors of up to one hundred
[43,44].
Based on this information, it is important to study the electronic and geometric structures of
heteroleptic [Cu(phen*)(DPEphos)]+ complexes in their excited states, in particular for
different substituents at the phen unit. For [Cu(phen*)(DPEphos)]+ or related complexes, so
far the corresponding work [48-52], and in particular also quantum chemistry computations,
are comparatively sparse, but such calculations are the first important step in the
investigations of excited states. They provide the background for further (spectroscopic)
studies and can complement and support the interpretation and analysis of experimental data.
This requires, however, a reliable and accurate computation of molecular structures, in
particular, as the results from geometry optimizations can dramatically influence subsequent
calculations, e.g. determinations of absolute energies [53,54].
Nowadays, DFT is the most advanced theory which can be used to calculate the electronic
and geometric structures of [Cu(phen*)(DPEphos)]+ complexes. Nevertheless, DFT still faces
a number of challenges, which may lead to less accurate results [55-58]. Validation studies
have shown that the calculated geometric and electronic parameters can dramatically depend
on the choice of the density functional and for each type of calculation some functionals are
more fitting than others [53-54,59-70]. Moreover, for any given property and molecular
system, suitable functionals can be different and it is highly advisable to first perform
4

benchmark studies for the respective target system [53-54,59-70]. For the
[Cu(phen*)(DPEphos)]+ complexes, the relative positioning of the large phen* and DPEphos
ligands, and thus the geometry of the molecule, will be affected by dispersion, which is not
included in standard DFT approaches [71,72]. Moreover, the calculated absorption spectra
and highest occupied molecular orbital (HOMO) to lowest unoccupied molecular orbital
(LUMO) energy gaps are expected to depend significantly on the treatment of the exchange
part of the energy in the DFT calculations [54]. In this context, ground state validation studies
are important as they can evaluate the ability of functionals to accurately predict the
structures of [Cu(phen*)(DPEphos)]+ complexes. Based on these studies, an appropriate
functional can be selected and then be used in subsequent calculations, in particular for the
excited states. For validation, usually the results from DFT are compared to a reference, and
for the [Cu(phen*)(DPEphos)]+ complexes ground state structures from X-ray diffraction
(XRD) and experimental absorption spectra are the only available standard for evaluation
[44,45,48-50,73]. DFT geometry optimizations, however, are usually done on single
molecules, while XRD measurements are performed on single-crystals. In contrast to single
molecules, the structure of molecules in crystals is influenced by surrounding molecules.
Consequently, a comparison of the structures as calculated by DFT and measured by XRD
suffers from intrinsic uncertainties. The magnitude of this uncertainty needs to be estimated
in order to identify when a deviation between molecular structures from DFT and XRD may
originate simply from crystal packing effects and is not significant for validation.
In this paper, we present both experimental and theoretical studies of six phosphorescent
heteroleptic [Cu(NN)(DPEphos)]+ complexes (1-6) (Fig. 1) with varying NN ligand
structures. We have investigated the electronic and geometric structures of the complexes in
their lowest excited states using DFT. To obtain accurate results, the density functional for
the excited state calculations has been selected based on ground state validation studies. We
evaluated the ability of seven density functionals to reproduce the ground state molecular
5

structures and absorption spectra derived from single-crystal XRD and solution-phase UV-
VIS absorption spectroscopy respectively, both from our own experiments as well as from
previously reported measurements by other groups [44,45,48-50,73,74]. Both functionals not
considering dispersion as well as functionals accounting for dispersion were tested. The
magnitude of the uncertainty in those validation studies due to crystal packing effects was
estimated in an approximate fashion. We determined the influence of surrounding molecules
on the structure of molecules in single-crystals by comparing the molecular structures of
complexes 1, 2 and 6 in different crystalline environments, using both data from own XRD
measurements and from previously published experiments by other groups [44,48-50,74].
2. Materials and methods
2.1. Selection of the [Cu(NN)(DPEphos)]+ complexes
Compounds 1-5 were selected to be representative of the family of [Cu(phen*)(DPEphos)]+
complexes with systematic variations in their phen* substituent structure, in particular for the
prominent 2,9-positions (1-4). An additional criterion was the availability of their ground
state crystal structures as a reference for density functional validation (1-5; Cambridge
Structural Database [44,45,48-50,73,75]). To complement the studies, the influence of a
substantial chemical variation of the NN ligand structure was investigated for complex 6 [74].
Complex 1, 2 and 6 were chosen for our XRD measurements and a representative comparison
of molecular structures in different crystal environments.
2.2. Synthesis and chemical analysis
All reagents were purchased from Aldrich and used as received. Solvents were spectral grade.
[Cu(phen)(DPEphos)]BF4 (1), [Cu(dmp)(DPEphos)]BF4 (2) and [Cu(bq)(DPEphos)]PF6 (6)
were synthesized according to the reported literature procedures [43,44,48-50,74]. The
6

products were analyzed by Electrospray Ionization Mass Spectrometry (ESI-MS, Varian MS-
1
500), Fourier-transform infrared spectroscopy (FT-IR, Bruker IFD 25) and H nuclear
magnetic resonance spectroscopy (NMR, Agilent Oxford 400 MHz), confirming the
successful synthesis of complexes 1, 2 and 6. Details can be found in the Appendix A,
Supplementary data (SD).
2.3. Single crystal X-ray diffraction
Single crystals of complexes 1, 2 and 6 suitable for XRD analysis were grown from a
solution of the synthesized powders in methanol, acetonitrile and dichloromethane (DCM)
respectively by slow evaporation of the solvent. For complexes 1 and 6, the crystals were
found to be unstable under ambient conditions, while crystals of complex 2 were observed to
be stable. The crystallographic experiments were performed on a Bruker SMART APEX II
system based on a D8 three-circle diffractometer with an Incotac microfocus X-ray source.
The theX-ray diffraction data were collected using graphite-monochromated and 0.5 mm-
MonoCap-collimated Mo-Kα radiation (λ = 0.71073 Å) at 100 K with an Oxford Cryojet low-
temperature cooling device. Details on data integration and refinement can be found in Ref.
[76] and the SD.
2.4. Photophysical measurements
A CARY 5E spectrophotometer was utilized to measure the electronic absorption spectra of
complexes 1, 2 and 6 in DCM. Similar to Hallmann et al. [77], electronic emission spectra in
DCM were recorded with a Horiba Fluorolog spectrophotometer. Complexes 1 and 2 were
excited at 400 nm and complex 6 at 450 nm while recording emissions in the wavelength
region 500-800 nm.
2.5. Theoretical calculations
7

All calculations were performed for the cations of complexes 1-6 (Fig. 1) and used the
Gaussian09 (G09) [78] program with the DFT method. Density functionals for the calculation
of the singlet ground state structures of 1-6 spanned from the standard GGA functional PBE
[79] and the hybrid-GGA functional B3LYP [80,81], which do not account for dispersion to
functionals which consider dispersion. The latter include the GGA functional B97D3 [82],
the hybrid-GGA PBE0 with the GD3 correction [83-85], the hybrid-meta-GGA ωB97XD
[86,87], which also includes long-range corrections as well as semi-empirical functionals, the
local meta-GGA M06L and hybrid-meta M06. The latter include non-covalent interactions
and were constructed with dispersive interactions in mind using parameters from extensive
validation studies [88,89]. In all the calculations, the Stuttgart-Dresden (SDD) [90,91] basis
set and effective core potential (ECP) were utilized for the Cu atom and the 6-31G* basis set
was used for all other atoms. For the description of the complexes in solution we used the
conductor-like polarizable continuum model (CPCM) [92-94] with DCM as a solvent. The
structures of the singlet ground and lowest lying triplet excited states were optimized to their
lowest minimum energy point on their potential energy surfaces and a full analysis of the
molecular orbitals (MO-s) was performed. Based on the ground state validation studies, the
PBE0-GD3 functional was selected for calculations on triplet excited states. These
calculations use the unrestricted wave function formalism and were carefully checked for
spin contamination. The stability of all geometries was tested by a frequency analysis. TD-
DFT calculations of the vertical excitation, at the singlet ground state equilibrium geometry,
with linear response, non-equilibrium solvation provided 40 singlet and triplet transition
energies as well as their corresponding oscillator strengths and character upon electronic
ground state excitation. For complex 2, the relaxation of the singlet excited state geometry
after vertical excitation was investigated by TD-DFT geometry optimizations with
equilibrium, linear response solvation providing the minimum energy point on the excited
state potential energy surface. To break the symmetry at the start of the optimization, ground
8

state geometries were perturbed slightly changing the dihedral angle (dha) between the N-Cu-
N and P-Cu-P planes. For the generation of (partial) density of states and to display the
electronic spectra GaussSum 2.2 [95] was utilized. For molecular graphics, the programs
MERCURY [96], Avogadro [97] and ChemCraft [98] were employed.
2.6 Comparison of the structures
Following the approach by Minenkov et al. [53], we investigated deviations between
geometries for two structures by evaluating the mean signed error (MSE) and mean unsigned
error (MUE) for changes in all interatomic distances and in bonding distances to the Cu(I)
coordination center respectively. The MSE and MUE were calculated according to MSE =
2/(N(N-1))* _i^N-1 _j>i^N (|Rij(DFT)-Rij(X-ray)|) and MUE = 2/(N(N-1))* _i^N-1 _j>i^N
(Rij(DFT)-Rij(X-ray)) [53]. Here, Rij is the interatomic distance between atoms i and j and N
is the number of atoms. This approach allows the study of systematic variations in bond
lengths that are not easy to access by a simple comparison of molecular Cartesian coordinates
[53]. Large distances which are influenced by Van der Waals interactions dominate the MSE
and MUE for the all distances approach [53] and a corresponding evaluation tests the ability
of the density functionals to describe those interactions. For details see Ref. [53] and the SD.
3. Results and Discussion
3.1. Molecular ground state structures from X-ray crystallography
To estimate the deviations in molecular structural parameters for different crystal packing,
MSE and MUE-s for changes in the interatomic distances of the molecular structures of
complexes 1, 2 and 6 were calculated, comparing results from our XRD experiments and
9

previously reported crystallographic measurements for the complexes by other research
groups [44,48-50,74]. For a maximum variation in crystal packing, in comparison to previous
crystallization processes [44,48-50,74], we utilized different solvents and counterions.
Consequently, we observed significant differences in the crystal systems, space groups,
asymmetric unit and unit cell structural parameters (SD). Key crystallographic data from our
measurements are presented in Table 1.
In Fig. 2 we have superposed the molecular structures for complex 1 as derived by XRD
within this work and by Kuang et al.[44]. The corresponding figures for complexes 2 and 6
are presented in the SD. Results by Zhang et al. [50], Yang et al. [49] and Costa et al. [48] are
similar to the results by Kuang et al. [44] and our results, and thus are omitted here. Both
common motifs and significant structural differences are found. All the XRD experiments for
complexes 1, 2 and 6 show that the Cu(I) center is in a distorted tetrahedral coordination
environment and that the NN ligand is tilted towards one of the two P atoms. Also, the O
atom of the DPEphos ligand is at a non-bonding distance (> 3.0 Å) and for complex 1 we
found an intramolecular -stacking interaction between the phen ligand and a phenyl group of
the DPEphos ligand, which has been previously reported by Costa et al. [48]. The dha-s
between the average N-Cu-N and P-Cu-P planes are nearly identical for the two different
crystal structures of complex 1 and 2, ~89 and ~82° respectively, but for complex 6 a
difference of 6° is observed (this work: 88.6°; 82.6° [74]). Furthermore, in comparison to
Refs. [44,49,50,74], we found displacements, rotations and tiltings of the phenyl groups of
the DPEphos ligand as well as tiltings of the average Cu-phen (1) or Cu-bq (6) planes nearly
perpendicular to the P-Cu-P plane by angles of around 7° and 32° respectively (Fig. 2 and
SD). An analysis, including all interatomic distances of the three complexes, reveals that
those differences in molecular structures correspond to an average MSE of ~0.025 Å and
MUE of ~0.25 Å (Fig. 3, upper graph, “crystal packing”). In other words, the average change
10

in interatomic distance for the same molecule in a different crystal lattice is of the order of
~0.025 Å and the absolute value of all changes in distances is ~0.25 Å.
In Table A1 (SD) we compare important bond lengths (Å) and valence angles (°) of
complexes 1, 2 and 6, as derived by XRD within this work (Exp.) and by Kuang et al. [44]
and Zhang et al. [74] respectively. An analysis including all the valence angles at the Cu atom
of the three complexes gives an average standard deviation of ~ 2° for differences in the
valence angles between two different structures of the same complex. An evaluation of all the
Cu-N and Cu-P bond lengths of the three complexes provides an average MSE of ~0.01 Å
and MUE of ~0.02 Å (Fig. 3, lower graph, “crystal packing”). Both the standard deviation of
the valence angles and the MSE are of the order of the typical standard deviations of 1-2° or
0.01-0.02 Å respectively reported by Martin et al. [99] for a variety of transition metal
complexes in different crystal environments.
3.2. Ground state geometric structures: X-ray crystallography and density functional theory
The ground state structures of complexes 1-6 have been calculated using seven different
density functionals and the calculated structures have been compared to molecular ground
state geometries derived from XRD experiments, using both data from own measurements
and previously published results from other research groups [44,45,48-50,73,74]. The MSE-s
and MUE-s for an analysis including all the interatomic distances of complexes 1-4 and 6 are
shown in Fig. 3 (upper graph). In addition, the lower part of Fig. 3 depicts the MSE-s and
MUE-s for an evaluation of all the Cu-N and Cu-P bond lengths of the complexes. Complex
5 is excluded in this analysis as our calculations indicate that the molecular structure of this
complex, as measured by XRD, does not correspond to the calculated structure for its ground
state (for more details refer to the following sections). Fig. 3 shows that the standard
functionals B3LYP and PBE substantially overestimate the interatomic distances and predict
expanded molecular structures, providing MSE-s of 0.1-0.15 Å for the all interatomic
11

distances method. For the functionals which consider dispersion, B97D3, PBE0G-D3, M06L,
M06 and ωB97XD, the MSE-s nearly vanish (< 0.05 Å), indicating that the overestimation of
bond distances is compensated, and similar results have been reported for the molecular
structures of ruthenium complexes [53]. It should be noted here that the average change in an
interatomic distance for the same molecule in a different crystal lattice is ~0.025 Å (“crystal
lattice”, Fig. 3, upper graph), and thus is substantially smaller than the differences between
MSE-s derived for the different groups of density functionals, i.e. MSE-s are a reliable
quantity for density functional validation. An evaluation of the MUE-s shows that functionals
which consider dispersion give on average lower absolute errors, MUE-s < 0.020 Å, than
functionals not accounting for dispersion, MUE-s > 0.022 Å. However, MUE-s for
differences in all interatomic distances of the same molecule in different crystal environments
(“crystal lattice”, Fig. 3, upper graph) show average uncertainties of up to ~0.25 Å, i.e. the
MUE-s are not a reliable quantity for validation. An analogous comparison of bonding
distances to the Cu coordination center (Fig.3, lower graph) indicates that the B3LYP
functional clearly overestimates the metal-ligand bond lengths. In agreement with previous
reports for transition metal complexes [53,99] the standard PBE functional predicts bond
lengths for the Cu center much more accurately, but as discussed above is not well suited to
calculate accurate overall geometries for the molecules.
A qualitative comparison of the ground state structures as measured by XRD and calculated
by DFT for the different functionals supports the finding that functionals which account for
dispersion offer an improved description of the ground state molecular structures and this is
depicted in Fig. 4 for complex 1. In particular, for the phen moiety the PBE0-GD3 functional
reproduces the experimental structure more closely.
Based on the validation studies, for a further analysis we have selected calculations at the
PBE0-GD3 level. In Table 2 we present the corresponding bond lengths and valence angles
12

for the Cu coordination center. An evaluation of the dha-s between the average N-Cu-N and
the P-Cu-P planes for complexes 1-4 and 6 gives good agreement between the experimentally
measured and calculated values, 88.9°/84.8° (1), 82.4°/82.5° (2), 78.8° [44]/79.9°(3),
86.7°[45]/88.4° (4) and 88.6°/89.3° (6), respectively. The dha-s in the ground state of
complexes 2 and 3 are substantially smaller than for complexes 1, 4 and 6. In other words,
complexes 2 and 3 show a larger geometrical distortion in their ground state and in particular
a decrease in the dha is observed for an increase in the alkyl chain length of the substituent at
the 2,9-position of the phen* unit. This distortion is accompanied by a tilting of the phen*
unit in direction of one of the P atoms of the DPEphos ligand (SD) and results in an angle of
~ 11° between the Cu-NN and N-Cu-N planes for complexes 2 and 3. In contrast, these
angles are around 0-3° for complexes 1 and 4. It should be noted that for complex 4 with a
bulky phenethyl ligand in the 2,9-position, in principle, a substantial distortion in the dha for
the ground state is expected. However, our calculations show that an intramolecular -
stacking interaction between a phenyl unit of the phenethyl ligand and one phenyl moiety of
the DPEphos ligand effectively constrains the dihedral angle of the molecule to close to
around 88°. In addition, we want to mention here that a comparison of ground state structures
derived from XRD and DFT for complex 5 shows dramatic differences in the structures (SD)
and an explanation will be given in the following sections.
3.3. Electronic ground state structures from density functional theory
We used seven different density functionals to calculate the electronic structures in the
ground states of complexes 1-6 in DCM. For calculations with different density functionals,
strong variations in the electronic structures and HOMO-LUMO gap energies have been
reported in previous work by other research groups [54,59,60]. We evaluated those variations
in our calculations (see SD) and found that only the hybrid functionals (B3LYP, PBE0-GD3,
M06) incorporating a portion of exact exchange from Hartree-Fock (HF) theory predict
13

HOMO-LUMO energies which are in agreement with the distinct absorption edges observed
in the UV-VIS spectra of the complexes (SD and [44-50,73,74]). Based on these findings, in
the following we discuss the results from calculations employing such a hybrid functional,
namely PBE0-GD3.
In Fig. 5 we display the calculated electronic structure of complex 1 in DCM. Similar figures
for complexes 2-6 can be found in the SD. To calculate the partial density of states (PDOS),
every complex was divided into three parts: the Cu center, the DPEphos ligand and the NN
ligand, and their percentage contributions were obtained as the sum of the atomic orbital
coefficient squares. The nature of each molecular orbital was then assigned based on its
PDOS.
We first compared results for complexes 1-4 and 6, as the results for complex 5 differ
significantly. For complexes 1-4 and 6, we observed that the highest occupied molecular
orbital (HOMO) has predominant DPEphos lone pair (~55 %) and Cu dyz (~ 35 %) character
(axis definition SI), while the lowest unoccupied molecular orbitals (LUMO) are delocalized
over the NN ligands (~97%, * orbitals). The HOMO-1 orbitals show Cu d contributions of
35-60% and smaller * DPEphos, and lone pair DPEphos and NN character. This is in
agreement with the DFT results by Zhang et al. [50] for complexes 1 and 2 in the gas-phase.
Similar to the DFT results by Yang et al. [49], for simplified versions of complexes 1 and 2 in
the gas-phase, small ligand-dependent differences in the MO composition are observed, with
a slightly increasing Cu d orbital character in the HOMO and HOMO-1 in the sequence
complex 1 > 2 > 3. More dramatic variations are found for complex 4 and 6 (SD), and this is
also expected considering the larger differences in their NN ligand structure. Their Cu d
orbital character in the HOMO is similar to that of complex 1, but for the HOMO-1, on the
other hand, a 10% decrease of the Cu contribution is found.
14

In general, as expected MOs with high contributions from the NN ligands show strong NN
ligand-dependent energy changes in comparison to those with smaller NN character. In
particular, for complexes 1-4 and 6, the HOMOs (NN * character ~ 8%) show smaller
variations in their energy (complex 1: -6.21 eV; complex 2: -6.18 eV; complex 3: -6.13 eV;
complex 4: -6.25 eV; complex 6: -6.25 eV) than the LUMOs (NN * character ~ 100%;
complex 1: -2.24 eV; complex 2: -2.17 eV; complex 3: -2.1 eV; complex 4: -2.15 eV;
complex 6: -2.75 eV). The LUMO energies for complexes 2, 3 and 4 are 0.07, 0.014 and 0.09
eV higher in comparison to complex 1, respectively, and the effect is not compensated by the
relative change in the HOMO energies of + 0.03, +0.08and -0.04 eV. We explain the higher
LUMO energy in complexes 2, 3 and 4 by their lower electronegativity for the N atoms of the
NN ligand in comparison to complex 1, as the electron-donating methyl, phenethyl and butyl
groups in dmp, dpep and dbp push charge into the phen moiety. As a consequence, more
energy is required to transfer an additional electron into the dmp, dpep and dbp ligand and the
LUMO energy is enhanced. Furthermore, a combination of effects is also possible and Kuang
and co-workers [44] have argued that interligand steric repulsions lead to an elongation of the
Cu-P bonds in complex 2 and possibly also destabilize its CT excited state. For complex 6,
we suggest that its significantly lower LUMO orbital energy is a consequence of the larger
conjugation and increase in delocalization for the bq ligand with respect to phen, which goes
along with a smaller reduction potential for bq [100]. This behavior is in agreement with
previous reports on the photophysical properties for the complexes in solution [43-45,48-
50,73,74].
Results for complex 5 differ significantly with respect to complexes 1-4 and 6 (see also SD).
The LUMO and HOMO of complex 5 show nearly identical character to the LUMO-s and
HOMO-1 for complexes 1-4 and 6, namely NN * and Cu d (35%)/DPEphos lone pair (55%)
respectively. However, the HOMO of complex 5 exhibits predominately NH2 lone pair
15

character (95%). In other words, in comparison to complexes 1-4 and 6, the introduction of
an NH2 group in 5 results in an additional MO in the HOMO-LUMO energy gap region. As a
consequence of this orbital ordering, the ground and excited state properties for 5 are
significantly different than those for complexes 1-4 and 6 (see also excited states).
3.4. Ground state absorption spectra
We have calculated the energies of the first 40 lowest singlet and triplet transitions (SD) for
complexes 1-6 using TD-DFT. The results for the selected singlet transitions are given in
Table 3 and the results for the triplet transitions can be found in the SD. The character of each
transition was assigned based on the character of the MO-s participating in the excitation.
Transitions within a ligand are denoted as intraligand (IL) excitations, transitions from one
ligand to another as ligand-to-ligand charge transfer (LLCT) excitations and transitions from
the Cu center to one of the ligands, MLCT transitions.
In Fig. 6 we compare our experimental and TD-DFT calculated ground state absorption
spectra for complex 1 in DCM. Analogue graphs for complexes 2-6 can be found in the SD.
There is excellent agreement between the spectra, in particular as TD-DFT tends to
frequently overestimate transition wavelengths of excitations, accompanied by a substantial
charge density redistribution [101-104]. All the spectra show two main peaks in the UV-VIS
region. The bands in the UV region of the experimental absorption spectra of all th
complexes stem from transitions with mixed MLCT, LLCT and IL character. The
experimental absorption band maxima in the blue region originate mainly from transitions at
396 (1), 392 (2), 391 (3), 383 (4), 386 (5) and 463 nm (6), with oscillator strengths of 0.105
(1), 0.084 (2), 0.074 (3), 0.074 (4), 0.070 (5) and 0.064 (6), and MLCT/LLCT character
(HOMO•LUMO 93% (1), 95% (2), 95% (3), 97% (4), and 92% (6) and H-1•LUMMO 91%
(5)). Here an electron is essentially transferred from the dyz orbitals at the Cu center and the
lone pair orbitals of the DPEphos ligand to the * orbitals of the NN ligand. For complexes 1-
16

4 and 6, this excitation corresponds to the energetically lowest singlet (HOMO•LLUMO) but
for complex 5, it corresponds to the second lowest singlet excitation (H-1•LUMMO). The
energetically lowest excitation in 5 corresponds to a HOMO•LUMMO excittationn ((99%), where
an electron is excited from the lone pair orbitals of the donating NH2 group into the *
orbitals of the phen moiety. This excitation is shown as a low energy tail in the absorption
spectrum of 5 [73] and is not seen for complexes 1-4 and 6.
3.4. Relaxation of singlet excited states and triplet excited state geometric structures
We have optimized the geometric structures in the lowest triplet excited states of complexes
1-6 in DCM. In Fig. 7 we compare those structures with the calculated structures for the
ground state, but only for complexes 1-4 and 6, as complex 5 shows significantly different
behavior for its lowest triplet state. The main difference is a substantial decrease in the dha
between the N-Cu-N and P-Cu-P planes from 88.9 (1), 82.5 (2), 79.9 (3), 88.4 (4) and 89.3°
(6) in the ground state to 58.0 (1), 69.7(2), 67.1 (3), 73.0 (4) and 65.8° (6) respectively in the
3CT state. This change corresponds to a ligand torisional motion, is in agreement with results
for other Cu(I) complexes [21-35] and shows that the dha is an important parameter in the
excited state dynamics of complexes 1-4 and 6. The largest changes are observed for
complexes 1 and 6, 30 and 23° respectively, with smaller changes for complexes 4, 2 and 3 of
15, 13 and 13° respectively. A sterically less demanding ligand facilitates a ligand torsional
motion and decrease of the dha in the excited states. This change in dha is accompanied by
corresponding rotations and displacements of the phenyl groups of the DPEphos ligand as
well as changes in important bond lengths and angles (Table 4). In comparison to the ground
state, the N(1)-Cu-P(1) and N(2)-Cu-P(2) angles decrease by 1-15° and N(1)-Cu-P(2) and
3
N(2)-Cu-P(1) increase by 3-30° in the CT states. Again, the largest differences are observed
for complex 1, the smallest for complex 3. At the same time a shortening of all the Cu-N
bonds and an elongation of all the Cu-P bonds by ~ 0.11 Å and an increase in the Cu···O
17

3
distance by on average 0.05 Å is observed in the CT state, similar to the trend reported for
other Cu(I) complexes with phosphine ligands [105]. The N-Cu-N angles increase from ~79°
to ~84°, while the P-Cu-P angles decrease from ~ 115° to ~ 105°. The natural bite angle of
3
DPEphos is around ~102°, with values as high as ~120° [106]. In the CT state the structure
of the DPEphos ligand relaxes in the direction of its natural bite angle, while the Cu-NN
distance becomes shorter, leading to steric stress and a displacement of the NN ligand
structure from its natural bite angle around ~ 80°. This is reflected in the C(N)-C(N) bond
lengths which are shortened by ~0.04 Å and the N-C bonds which are elongated by 0.01-0.05
Å in the triplet excited state.
3
In addition to the molecular structure for the CT state, for comparison, the structure in the
1CT state after initial phototexcitation and relaxation was also calculated, but only for
complex 2 because TD-DFT calculations of excited state properties typically require
significantly more computational effort than DFT calculations. The structural parameters
such as the dha between the N-Cu-N and P-Cu-P planes and molecular geometries (SD)
including bond lengths and angles are nearly identical for the two states. We conclude that a
1
ligand torsional motion in the CT state of complex 2 results in a molecular geometry in the
minimum of the 1CT state potential energy surface which is nearly identical to that in the 3CT
state.
As mentioned at the beginning of this section, the geometrical structure of complex 5 in its
lowest triplet excited state is significantly different than that of complexes 1-4 and 6, and this
is a consequence of the IL nature of the lowest singlet excitation in 5. This difference is also
reflected in deviations of the bonding distances and valence angles between the ground and
lowest triplet excited state of 5 (Table 4). Moreover, it should be noted that the calculated
molecular structure in the lowest triplet excited state of complex 5 shows excellent agreement
with the molecular structure derived from ground state XRD [73] (SD). The molecular
18

structure in the crystal is best represented by the geometry of an IL state where lone pair
electrons of the NH2 group are transferred to the phen moeity.
3.5. Electronic structure of the triplet excited state
We analyzed the electronic structures for complexes 1-6 in their lowest triplet excited states.
Mulliken charges for complexes 1-4 and 6 point at the difference between the ground state
and the lowest triplet excited state being of MLCT/LLCT nature. The Cu atom and the
DPEphos ligand become more positive in the triplet state by ~ 0.20-0.25 e and ~ 0.30-0.35 e
respectively and the NN ligands gain in total around 0.55 e for all the complexes. In contrast,
the Mulliken charges for the ground and lowest triplet excited state of complex 5 indicate no
significant changes for the Cu atom and the DPEphos ligand. However, the NH2 and phen
moeity become more positive and negative respectively by around 0.17 e in the triplet excited
3
state, again showing the phen* IL character of this state. Energies for the CT states of
complexes 1-4 and 6 in DCM are 2.15, 2.36, 2.39, 2.48 and 1.81 eV higher than for their
ground state respectively. The trend of decreasing energy differences in the order complex 4
> 3 > 2 > 1 maps the trend for the experimental emission energies (SI and [44,45,48-50];
emission maxima: 544 (4), 560 (3), 565 (2) and 700 nm (1)). A larger geometrical relaxation
in the excited states of complex 1 results in a larger decrease in energy for these states in
comparison to complexes 2-4. Following the energy gap law, this larger decrease in energy
3
should correspond to a significantly lower lifetime of the CT state of 1 and indeed this has
been observed previously ( = 0.2 (1), 14.3 (2) and 16.1 µs (3)) [44]. A comparison between
complexes 4 and 6 does not provide significant insight as the energy gap law works best for
complexes with a good similarity of structure [107].
In Fig. 8 we compare the results for the energy-level diagrams, MOs and orbital compositions
from our DFT calculations for the triplet excited and the ground state of complex 1 in DCM.
19

Analogous graphs for complexes 2-6 can be found in the SD. As for any unrestricted
calculations, there are two complete sets of orbitals, one for the α electrons (α spin) and one
for the β electrons (β spin) which use the same set of basis functions but different molecular
3
orbital coefficients. For complexes 1-4 and 6 in the flattened CT state the former ground
state HOMO orbitals and orbitals just below gain significant DPEphos and NN character,
both from lone pair and * orbitals. In particular, the NN * orbital character is enhanced
while the contribution of the Cu center decreases significantly. In contrast, the former ground
state LUMO and LUMO+1 orbitals still show predominant NN * character in the 3CT state.
For complex 5 in the lowest triplet excited state, the former ground state HOMO orbitals gain
significant Cu d, DPEphos (lone pair) and NN * character, but the former ground state
LUMO orbitals exhibit still mainly NN * character.
4. Conclusions
3
We have studied the electronic and geometric structures in the ground and CT states of six
phosphorescent heteroleptic [Cu(NN)(DPEphos)]+ complexes with varying NN ligand
structures using DFT. The soundness of the calculations has been verified by ground state
validation studies. We compared the ability of seven density functionals to predict the
molecular ground state geometries and absorption spectra derived from single-crystal XRD
and solution-phase UV-VIS absorption spectroscopy respectively. Our analysis shows that
the methods which account for dispersion predict the geometrical structures of the molecules
more accurately than standard approaches which do not consider dispersion. The latter
systematically overestimate the internuclear distances and give expanded molecular
structures. Moreover, the experimental absorption energies are only modelled accurately in
calculations with hybrid functionals which incorporate a portion of exact exchange from HF
theory.
20

Results from the DFT calculations at the PBE0-GD3 level are in good agreement with the
experimental data. They explain deviations in the molecular geometries, ground state
absorption spectra as well as excited state emission quantum yields and lifetimes for
complexes with different NN ligand structures through differences in the electronegativity,
conjugation and bulkiness of those ligands. In particular, in comparison to the ground state,
the results show a decrease of the dihedral angle between the N-Cu-N and P-Cu-P planes for
3
the CT excited states with MCLT/LLCT character. Sterically more demanding ligands
impede this change in geometry. In this context, it would be important to investigate the
dynamics in the excited states of [Cu(NN)(DPEphos)]+ complexes, for example in time-
resolved experiments, and here we expect that the current results will significantly facilitate
the interpretation of data.
Acknowledgements
We thank R.W. Hartsock and K.J. Gaffney for discussion and for access to the Gaffney group
chemistry lab at the PULSE Institute of the SLAC National Accelerator Laboratory (Stanford
University, Stanford, California 94305, USA). Moreover, we thank V. Mäckel and L. Glaser
for detailed discussion of the manuscript. K.K. acknowledges financial support from the
Volkswagen Foundation under the Peter Paul Ewald fellowship program (87008) and from
SFB 755 (BO3). S.T. acknowledges additional support from SFB 1073 (BO6).
Appendix A. Supplementary data
CCDC 1498327-1498329 contain the supplementary crystallographic data for complexes 1, 2
and
6.
These
data
can
be
obtained
free
of
charge
via
http://www.ccdc.cam.ac.uk/conts/retrieving.html, or from the Cambridge Crystallographic
Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: (+44) 1223-336-033; or e-mail:
deposit@ccdc.cam.ac.uk. Input and output files for the (TD-)DFT calculations are available
21

upon request. Additional supplementary information and data can be found in Appendix A in
the online version of this article.
References
[1] O’Regan, B.; Grätzel, M. Nature 1991, 353, 737-740.
[2] Nazeeruddin, M.K.; Grätzel, M. Struct. Bond 2007, 123, 113-175.
[3] Rakow, N.A.; Suslick, K.S. Nature 2000, 406, 710-713.
[4] DeGraff, B.A.; Demas, J.N. Coord. Chem. Rev. 2001, 211, 317-351.
[5] Yam, V.W.-W.; Wong, K.M.-C. Chem. Commun. 2011, 47, 11579-11592.
[6] Elliott, M.; Pichot, F.; Bloom, C.J.; Rider, L.S. J. Am. Chem. Soc. 1998, 120, 6781-6784.
[7] Yersin, H.; Rausch, A.F.; Czerwieniec, R.; Hofbeck, T.; Fischer, T. Coord. Chem. Rev.
2011, 255, 2622-2652.
[8] Balzani, V.; Campagna, S.; Eds., Photochemistry and Photophysics of Coordination
Compounds I, Springer Berlin Heidelberg, Berlin, Heidelberg, 2007.
[9] Balzani, V.; Campagna, S.; Eds., Photochemistry and Photophysics of Coordination
Compounds II, Springer Berlin Heidelberg, Berlin, Heidelberg, 2007.
[10] Zhang, W.; Alonso-Mori, R.; Bergmann U.; Bressler, C.; Chollet, M.; Galler, A.;
Gawelda, W.; Hast, R.G.; Hartsock, R.W.; Kroll, T.; Kjær, K.S.; Kubi ek, K.; Lemke, H.T.;
Liang, H.W.; Meyer, D.A.; Nielsen, M.M.; Purser, C.; Robinson, J.S.; Solomon, E.I.; Sun, Z.;
Sokaras, D.; van Driel, T.B.; Vankó, G.; Weng, T.-C.; Zhu, D.; Gaffney, K.J. Nature 2014,
509, 345–348.
[11] Armaroli, N. Chem. Soc. Rev. 2001, 30, 113-124.
[12] Armaroli, N.; Accorsi G.; Cardinali, F.; Listorti, A. Top Curr Chem 2007, 280, 69-115.
[13] Kia, R.; Mirkhani, V.; Harkema, S.; van Hummel, G. J. Inorg. Chim. Acta 2007, 360,
3369-3375.
[14] Kia, R.; Scholz, M.; Raithby, P. R.; Techert, S. Inorg. Chim. Acta 2014, 423, 348-357.
22

[15] Ford, P. C. Rev. Chem. Intermed. 1979, 2, 267–296.
[16] Juris, A.; Balzani, V.; Barigelletti, F.; Campagna, S.; Belser, P.; von Zelewsky, A.
Coord. Chem. Rev. 1988, 84, 85–277.
[17] Costa, R.D.; Ort, E.; Bolink, H.J.; Monti, F.; Accorsi, G.; Armaroli, N. Angew. Chem.
Int. Ed. 2012, 51, 8178-8211.
[18] Mara, M. W.; Fransted, K. A.; Chen, L. X. Coord. Chem. Rev. 2015, 282, 2–18.
[19] Zhang, W.; Kjær, K. S.; Alonso-Mori, R.; Bergmann, U.; Chollet, M.; Fredin, L. A.;
Hadt, R. G.; Hartsock, R. W.; Harlang, T.; Kroll, T. et al., Chem. Sci. 2016, DOI
10.1039/c6sc03070j.
[20] Jäger, M.; Freitag, L.; González, L. Coord. Chem Rev. 2015, 304-305, 146-165.
[21] Chen, L.X.; Jennings, G.; Liu, T.; Gosztola, D.J.; Hessler, J.P.; Scaltrito, D.V.; Meyer,
G.J. J. Am. Chem. Soc. 2002, 124, 10861-10867.
[22] Siddique, Z. A.; Yamamoto, Y.; Ohno, T.; Nozaki, K. Inorg. Chem. 2003, 42, 6366–
6378.
[23] Zgierski, M. Z. J. Chem. Phys. 2003, 118, 4045–4051.
[24] Chen, L.X.; Shaw, G.B.; Novozhilova, I.; Liu, T.; Jennings, G.; Attenkofer, K.; Meyer,
G.J.; Coppens, P. J. Am. Chem. Soc. 2003, 125, 7022-7034.
[25] Shaw, G.B.; Grant, C.D; Shirota, H.; Castner Jr., E.W.; Meyer, G.J.; Chen, L.X. J. Am.
Chem. Soc. 2007, 129, 2147-2160.
[26] Iwamura, M.; Takeuchi, S.; Tahara, T. J. Am. Chem. Soc. 2007, 129, 5248–5256.
[27] Lockard, J. V.; Kabehie, S.; Zink, J. I.; Smolentsev, G.; Soldatov, A.; Chen, L. X. J.
Phys. Chem. B 2010, 114, 14521–14527.
[28] Iwamura, M.; Watanabe, H.; Ishii, K.; Takeuchi, S.; Tahara, T. J. Am. Chem. Soc. 2011,
133, 7728-7736.
23

[29] Gothard, N.A.; Mara, M.W.; Huang, J.; Szarko, J.M.; Rolczynski, B.; Lockhard, J.V.;
Chen, L.X. J. Phys. Chem. A 2012, 116, 1984-1992.
[30] Penfold, T.J.; Karlsson, S.; Capano, G.; Lima, F.A.; Rittmann, J.; Reinhard, M.;
Rittmann-Frank, M.H.; Braem, O.; Baranoff, E.; Abela, R.; Tavernelli, I.; Rothlisberger, U.;
Milne, C.J.; Chergui, M. J. Phys. Chem. A 2013, 117, 4591–4601.
[31] Mara, M. W.; Jackson, N.E.; Huang, J.; Stickrath, A.B.; Zhang, X.; Gothard, N.A.;
Ratner, M.A.; Chen, L.X. J. Phys. Chem. B 2013, 117, 1921–1931.
[32] Tromp, M.; Dent, A.J.; Headspith, J.; Easun, T.L.; Sun, X. Z.; George, M.W.; Mathon,
O.; Smolentsev, G.; Hamilton, M.L.; Evans, J. J. Phys. Chem. B 2013, 117, 7381–7387.
[33] Iwamura, M.; Takeuchi, S.; Tahara, T. Phys. Chem. Chem. Phys. 2014, 16, 4143-4154.
[34] Hua, L.; Iwamura, M.; Takeuchi, S.; Tahara, T. Phys. Chem. Chem. Phys. 2015, 17,
2067-2077.
[35] Garakyaraghi, S.; Danilov, E.O.; McCusker, C.E.; Castellano, F.N. J. Phys. Chem. A
2015, 119, 3181–3193.
[36] Blaskie, M.W.; McMillin, D.R. Inorg. Chem. 1980, 19, 3519–3522.
[37] Kirchhoff, J.R.; McMillin, D.R.; Robinson, W.R.; Powell, D.R.; McKenzie, A.T.; Chen,
S. Inorg. Chem. 1985, 24, 3928–3933.
[38] Palmer, C.E.A.; McMillin, D.R.; Kirmaier, C.; Holten, D. Inorg. Chem. 1987, 26, 3167–
3170.
[39] McMillin, D.R.; McNett, K.M. Chem. Rev. 1998, 98, 1201–1220.
[40] Scaltrito, D.V; Thompson, D.W.; O’Callaghan, J.A.; Meyer, G. J. Coord. Chem. Rev.
2000, 208, 243–266.
[41] Eggleston, M.K.; McMillin, D.R.; Koenig, K.S.; Pallenberg, A.J. Inorg. Chem. 1997, 36,
172-176.
24

[42] Cunningham, C.T.; Cunningham, K.L.H.; Michalec, J.F.; McMillin, D.R. Inorg. Chem.
1999, 38, 4388-4392.
[43] Cuttell, D.G.; Kuang, S.-M.; Fanwick, P.E.; McMillin, D.R.; Walton, R.A. J. Am. Chem.
Soc. 2002, 124, 6-7.
[44] Kuang, S.-M.; Cuttell, D.G.; McMillin, D.R.; Fanwick, P.E.; Walton, R.A. Inorg. Chem.
2002, 41, 3313-3322.
[45] Armaroli, N; Accorsi, G.; Holler, M.; Moudam, O.; Nierengarten, J.F.; Zhou, Z.; Wegh,
R.T.; Welter, R. Adv. Mater. 2006, 18, 1313–1316.
[46] Zhang, Q.; Zhou, Q.; Cheng, Y.; Wang, L.; Ma, D.; Jing, X.; Wang, F. Adv. Funct.
Mater. 2006, 16, 1203–1208.
[47] Dumur, F. Organic Electronics 2015, 21, 27-39.
[48] Costa, R.D.; Tordera, D.; Ortí, E.; Bolink, H. J.; Schönle, J.; Graber, S.; Housecroft,
C.E.; Constable, E.C.; Zampese, J.A. J. Mater. Chem. 2011, 21, 16108–16118.
[49] Yang, L.; Feng, J.-K.; Ren, A.-M.; Zhang, M.; Ma, Y.-G.; Liu, X.-D. Eur. J. Inorg.
Chem. 2005, 10, 1867-1879.
[50] Zhang, K.; Zhang, D. Spec. Acta Part A 2014, 124, 341-348.
[51] Vorontsov, I.I.; Graber, T.; Kovalevsky, A.J.; Novozhilova, I.V.; Gembicky, M.; Chen,
Y.-S.; Coppens P. J. Am. Chem. Soc. 2009, 131, 6566–6573.
[52] Tschierlei, S.; Karnahl, M.; Rockstroh, N.; Junge, H.; Beller, M.; Lochbrunner, S.
ChemPhysChem 2014, 15, 3709–3713.
[53] Minenkov, Y.; Singstad, A.; Occhipinti, G.; Jensen, V.R. Dalton Trans. 2012, 41, 5526–
5541.
[54] Kreutzer, J.; Blaha, P.; Schubert, U. Comput. Theor. Chem. 2016, 1084, 162-168.
[55] Neese, F.; Ames, W.; Christian, G.; Kampa, M.; Liakos, D.G.; Pantazis, D. A.; Roemelt,
M.; Surawatanawong, P.; Ye, S.F. Adv. Inorg. Chem. 2010, 62, 301-349.
25

[56] Schwarz, H. Angew. Chem. Int. Ed. 2003, 42, 4442-4454.
[57] Cohen, A.J.; Mori-Sanchez, P.; Yang, W.T. Science 2008, 321, 792-794.
[58] Cramer C.J.; Truhlar, D.G. Phys. Chem. Chem. Phys. 2009, 11, 10757-10816.
[59] Muscat J.; Wander , A.; Harrison, N.M. Chem. Phys. Lett. 2001, 342, 397-401.
[60] Tretiak, S.; Igumenschev, K.; Chernyak, V. Phys. Rev. B 2005, 71, 33201.
[61] Bühl, M.; Kabrede, H.J. Chem. Theor. Comput. 2006, 2, 1282-1290.
[62] Riley, K.E.; Merz, K.M. J. Phys. Chem. A 2007, 111, 6044-6053.
[63] Bühl, M.; Reimann, C.; Pantazis, D.A.; Bredow, T.; Neese, F. J. Chem. Theor. Comput.
2008, 4, 1449-1459.
[64] Sieffert, N.; Bühl, M. Inorg. Chem. 2009, 48, 4622-4624.
[65] Rydberg, P.; Olsen, L. J. Phys. Chem. A 2009, 113, 11949-11953.
[66] Yuan, J.; Hughes, R.P.; Golen, J.A.; Rheingold, A.L. Organometallics,
2010, 29, 1942-1947.
[67] Marom, N.; Tkatchenko, A.; Scheffler, M.; Kronik, L. J. Chem. Theory Comput. 2010,
6, 81-90.
[68] Jimenez-Hoyos, C.A.; Janesko, B.G.; Scuseria, G.E. J. Phys. Chem. A 2009, 113, 11742-
11749.
[69] Mohamed, T.A.; Shaaban, I.A.; Farag, R.S.; Zoghaib, W.M.; Afifi M.S. Spec. Acta Part
A 2015, 135, 417-427.
[70] Sardrooni, J. J.; Rastkar, A.; Yousefina, N. R.; Azamat, J. J. Theor. Comput. Chem.
2013, 12, 1350066.
[71] Kristyan, S.; Pulay, P. Chem. Phys. Lett. 1994, 229, 175-180.
[72] Reimers, J.R.; Cai, Z.-L.; Bilic, A.; Hush, N.S. Ann. N. Y. Acad. Sci. 2003, 1006, 235-
251.
26

[73] Yao, X.-X.; Guo, Y.-M.; Liu, R.; Feng, X.-Y.; Li, H.-H.; Liu, N.; Yang, F.-L.; Li, X.-L.
Polyhedron 2015, 92, 84-92.
[74] Zhang, Q.; Ding, J.; Cheng, Y.; Wang, L.; Xie, Z.; Jing, X.; Wang, F. Adv. Funct. Mater.
2007, 17, 2983-2990.
[75] Allen, F.H. Acta Crystallogr. Sect. B: Struct. Sci. 2002, 58, 380.
[76] Thekku-Veedu, S.; Raiser, D.; Kia, R.; Scholz, M; Techert, S. J. Phys. Chem. B 2014,
118, 3291–3297.
[77] Hallmann, J.; Techert, S. J. Phys. Chem. Lett. 2010, 1, 959–961.
[78] Gaussian 09, Revision E.01, Frisch, M.J.; Trucks, G.W.; Schlegel, H.B.; Scuseria, G.E.;
Robb, M.A.; Cheeseman, J.R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G.A.;
Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.P.; Izmaylov, A.F.; Bloino, J.; Zheng, G.;
Sonnenberg, J.L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.;
Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery Jr., J.A.; Peralta,
J.E.; Ogliaro, F.; Bearpark, M.; Heyd, J.J.; Brothers, E.; Kudin, K.N.; Staroverov, V.N.;
Keith, T.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J.C.; Iyengar,
S.S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J.M.; Klene, M.; Knox, J.E.; Cross, J.B.;
Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R.E.; Yazyev, O.; Austin,
A.J.; Cammi, R.; Pomelli, C.; Ochterski, J.W.; Martin, R.L.; Morukuma, K.; Zakrzewski,
V.G.; Voth, G.A.; Salvador, P.; Dannenberg, J.J.; Dapprich, S.; Daniels, A.D.; Farkas, O.;
Foresman, J.B.; Ortiz, J.V.; Cioslowski, J.; Fox, D.J. Gaussian, Inc., Wallingford CT, 2013.
[79] Perdew, J.P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77, 3865-3868.
[80] Becke, A.D. J. Chem. Phys. 1993, 98, 5648-5652.
[81] Lee, C.; Yang, W.; Parr, R.G. Phys. Rev. B 1988, 37, 785-789.
[82] Grimme, S.; Ehrlich, S.; Goerigk, L. J. Comp. Chem. 2011, 32, 1456-1465.
[83] Perdew, J.P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77, 3865-3868.
27

[84] Perdew, J.P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett., 1997, 78, 1396.
[85] Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys. 2010, 132, 154104.
[86] Chai, J.D.; Head-Gordon, M. Phys. Chem. Chem. Phys. 2008, 10, 6615-6620.
[87] Wu, Q.; Yang, W.T. J. Chem. Phys. 2002, 116, 515-524.
[88] Y. Zhao, Y.; Truhlar, D.G. Theor. Chem. Acc. 2008, 120, 215.
[89] Y. Zhao, Y.; Truhlar, D.G. Acc. Chem. Res. 2008, 41, 157.
[90] Andrae, D.; Haußermann, U.; Dolg, M.; Stoll, H.; Preuß, H. Theor. Chim. Acta 1990, 77,
123-141.
[91] Stratmann, R.E.; Scuseria, G.E.; Frisch, M.J. J. Chem. Phys. 1998, 109, 8218-8224.
[92] Barone, V.; Cossi, M. J. Phys. Chem. 1998, A102, 1995-2001.
[93] Cossi, M.; Barone, V. J. Chem. Phys. 2001, 115, 4708-4717.
[94] Cossi, M.; Rega, N.; Scalmani, G.; Barone, V. J. Comput. Chem. 2003, 24, 669-681.
[95] O'Boyle, N.M.; Tenderholt, A.L.; Langner, K.M. J. Comp. Chem. 2008, 29, 839-845.
[96] Macrae, C.F.; Edgington, P.R.; McCabe, P.; Pidcock, E.; Shields, G.P.; Taylor, R.;
Towler, M.; van de Streek, J. J. Appl. Cryst. 2006, 39, 453-457.
[97] Hanwell, M.D.; Curtis, D.E.; Lonie, D.C.; Vandermeersch, T.; Zurek, E.; Hutchison,
G.R. Journal of Cheminformatics 2012, 4:17.
[98] http://www.chemcraftprog.com
[99] Martin, A.; Orpen, A. G. J. Am. Chem. Soc. 1996, 118, 1464-1470.
[100] Yang, PhD thesis, University of North Texas, Denton, Texas, 1995.
[101] Dreuw, A.; Weisman, J.L.; Head-Gordon, M. J. Chem. Phys. 2003, 119, 2943-2946.
[102] Niehaus, T.A.; Hofbeck, T.; Yersin, H. RSC Adv. 2015, 5, 63318-63329.
[103] Záliš, S.; Amor, N.B.; Daniel, C. Inorg. Chem. 2004, 43, 7978-7985.
[104] Turki, M.; Daniel, C.; Záliš, S.; Vl ek, Jr., A.; van Slageren, J.; Stufkens, D.J. J. Am.
Chem. Soc. 2001, 123, 11431-11440.
28

[105] Vorontsov, I.I.; Graber, T.; Kovalevsky, A.J.; Novozhilova, I.V.; Gembicky, M.; Chen,
Y.-S.; Coppens P. J. Am. Chem. Soc. 2009, 131, 6566–6573.
[106] Dierkes, P.; van Leeuwen, P.W.N.M. J. Chem. Soc., Dalton Trans. 1999, 1519–1529.
[107] Lakowicz, J.R. Springer, New York, 2006, Principles of fluorescence spectroscopy.
29