Journal of Molecular Structure: THEOCHEM 946 (2010) 33–42

Contents lists available at ScienceDirect

Journal of Molecular Structure: THEOCHEM

journal homepage: www.elsevier.com/locate/theochem

DFT studies on thiophene acetylide Ru(II) complexes for nonlinear optics:

Structure–function relationships and solvent effects
Paulo J. Mendes a,b,*, Tiago J.L. Silva a, A.J. Palace Carvalho a,b, J.P. Prates Ramalho a,b
a
Centro de Química de Évora, Universidade de Évora, Rua Romão Ramalho 59, 7002-554 Évora, Portugal
b
Departamento de Química, Universidade de Évora, Rua Romão Ramalho 59, 7002-554 Évora, Portugal

a r t i c l e i n f o a b s t r a c t

Article history: Density functional theory (DFT) calculations were employed to investigate the second-order nonlinear
Received 20 July 2009 optical (SONLO) properties of g5-monocyclopentadienylruthenium(II) thiophene acetylide complexes.
Received in revised form 11 January 2010 From molecular structure, electronic states, and optical absorption spectra, we have studied the effect
Accepted 15 January 2010
of donor or acceptor substituents in thiophene ligands on their first hyperpolarizabilities in vacuum. Cal-
Available online 22 January 2010
culations in solvated media have also been performed for the complex with the highest first hyperpolar-
izability obtained in vacuum. The results reveal a significant influence of solvation on the first
Keywords:
hyperpolarizability of this compound. The improvement of the second-order nonlinear optical properties
TD-DFT
Quadratic hyperpolarizability
in solvated media is due not only to the change of the excitation energies but also to the increase of
Ruthenium acetylide complexes ground-state molecular polarization and efficiency of metal-to-ligand charge transfer for electronic
Solvent effects excitations.
Ó 2010 Elsevier B.V. All rights reserved.

1. Introduction the second-order NLO response. Our recent experimental results

on thiophene acetylide g5-monocyclopentadienylruthenium(II)
The design of new organometallic materials with large second- [27] derivatives show, in fact, adequate spectroscopic properties
order nonlinear optical properties (SONLO) is currently the subject for second-order NLO purposes (the measurements of the experi-
of extensive research both by theoretical and experimental meth- mental properties are currently in progress). Also, our recent
ods, since they have an important application in the area of inte- time-dependent density functional theory (TD-DFT) studies on
grated optics [1–17]. Usually, compounds possessing large g5-monocyclopentadienyliron(II) complexes with substituted thio-
molecular first hyperpolarizability, b, contain donor (D) and accep- phene-acetylide ligands gave an insight on the electronic factors
tor (A) groups linked through a p-backbone. The NLO properties of that may be responsible for the SONLO properties [28]. Several
such polarizable dipolar compounds are caused by intense, lower- studies have indicated that the calculated hyperpolarizabilities
energy donor-to-acceptor charge-transfer (CT) transitions. In case using the TD-DFT approach match very well with experimental
of metallo-organic compounds, the metal centre can be bound to trends [29–35] and this method has been increasingly used to
a highly polarizable conjugated backbone, acting hence as an elec- accurately calculate first hyperpolarizabilities of organometallic
tron releasing or withdrawing group. Consequently, strong charge- complexes [5,6,10,11,36–41]. Nevertheless, the majority is lacking
transfer transitions can occur, leading to high molecular first in the considerations of the environment effects, namely, the solva-
hyperpolarizabilities (b). This is the case of the general family of tion interactions that in some cases are critical for obtaining quan-
g5-monocyclopentadienylruthenium/iron(II) complexes present- titatively satisfactory results of both the electronic excitations and
ing benzene- or thiophene-based conjugated ligands coordinated the first hyperpolarizabilities in comparison with the experimental
to the metal centre through nitrile or acetylide linkages [18–26], results. In fact, it is well-known that the solvent polarity influences
that showed to be much more efficient donor groups for second- both the structure and optical properties of conjugated organic
order NLO purposes than the traditional organic donor groups. molecules and metal complexes and, therefore, their NLO proper-
The results suggest that combination of acetylide thiophene li- ties. In recent years, however, this subject is attracting increasing
gands with appropriate organometallic fragments would maximize attention and several publications have been devoted to the study
of solvation effects on the hyperpolarizabilities of organic mole-
cules and organometallic complexes [33,41–44].
* Corresponding author. Address: Centro de Química de Évora, Universidade de
In our continuous effort to get a better understanding on the elec-
Évora, Rua Romão Ramalho 59, 7002-554 Évora, Portugal. Tel.: +351266745318;
fax: +351266745303. tronic factors that may dictate the SONLO properties of g5-monocy-
E-mail address: pjgm@uevora.pt (P.J. Mendes). clopentadienylmetal complexes, we report herein a TD-DFT study

0166-1280/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.theochem.2010.01.029
34 P.J. Mendes et al. / Journal of Molecular Structure: THEOCHEM 946 (2010) 33–42

on the model complexes [RuCp(H2PCH2CH2PH2)(CC{SC4H2}Y)] 12 lowest excitation energies were computed. The simulation of
(Y = NMe2, NH2, OMe, H, CHO, CN, NO2). Both electron-acceptor the absorption spectra was performed by a convolution of gaussian
and electron-donor groups in the acetylide thiophene moiety were functions centred at the calculated excitation energies. Ground and
considered in order to verify the ability of these ligands to act as excited state dipole moments were obtained by DFT/TD-DFT using
hyperpolarizable chromophores when interacting with a good elec- the Gamess-US software [60] with the same basis sets and func-
tron-donor organometallic fragment such as the g5-monoc- tional as in the previous calculations. At present, the TD-DFT is
yclopentadienylruthenium(II) group. We also studied the solvation the most successful and extensively used method to calculate exci-
effects, using the self-consistent reaction field approach (SCRF), in tation energies and to produce electronic spectra. It is known that
particular using the self-consistent isodensity defined cavity within this method gives accurate results for low-lying excited states, well
the polarizable continuum model (SCI-PCM), on the electronic prop- below the ionization limit. For such states, the typical error of TD-
erties and first hyperpolarizabilities of the complex for which the DFT lies within the range of 0.1–0.5 eV, which is almost compara-
higher value of static quadratic hyperpolarizability was obtained. ble with the error of high-level correlated approaches such as
The knowledge of how the solvent polarity affects the structure EOM-CCSD or CASPT2 [61]. Although TD-DFT gives good results
and NLO properties of these complexes will give us important infor- for low-lying excited states, it is now well-known that TD-DFT pro-
mation in order to compare the calculated first hyperpolarizabilities vides unreliable treatment of molecular processes involving
in solution with those to be obtained in the near future through the molecular charge transfer such as Rydberg transitions, valence
planned experimental measurements. states of molecules exhibiting extended p-systems, doubly excited
states and charge-transfer excited states [61]. Also, in many cases,
2. Computational details results obtained with TD-DFT are sensitive to the choice of the ex-
change–correlation functionals [62]. Thus, the reliability of TD-DFT
The energy calculations and geometry optimizations of ruthe- results should be verified by comparison with experimental data,
nium complexes were carried out at DFT level using the Gaussian for which the TD-DFT calculations in solvated media becomes
03 software package [45]. Becke’s three parameter hybrid func- essential.
tional [46] with the LYP correlation functional [47] (B3LYP) were For the solvation effects study we have used a self-consistent
employed in all calculations. This is the most widely used func- reaction field (SCRF) approach, in particular using the self-consis-
tional since it provides superior performance in numerous energy tent isodensity defined cavity within the polarizable continuum
assessments of small molecules and reproduces the geometries model (SCI-PCM), as implemented in Gaussian 03 [45]. This meth-
of smaller and larger molecules very well. However, it is well- od has been used to study solvation effects on the hyperpolariz-
known that it is better for main-group chemistry than for transi- abilities of different chromophores [42,43,63–65] and ruthenium
tion metals, it underestimates reactions barriers, is inadequate complexes [44]. Also, the PCM approach was recently used,
for several types of nonbonded interactions and fails in prediction namely, for the mechanistic studies of Ru-catalyzed processes
of thermodynamic quantities [48,49]. In spite of these shortcom- [66] and thermodynamic properties of ruthenium complexes
ings, this functional is widely used in calculations of hyperpolariz- [67,68]. Using this method, the geometry of the studied complex
abilities of organic and organometallic systems [10,34,37,40] and was re-optimized and the UV/Vis spectra and hyperpolarizabilities
was shown to reproduce reasonably well some experimental were calculated by DFT/TD-DFT with the same functionals and ba-
trends [33,34,50]. As a compromise between calculation quality sis sets.
and computational cost we have adopted the 6–31G basis set We used the GaussSum 2.1.6 software [69] for the detailed
(for geometry optimizations) and 6–31 + G basis set (for the cal- analysis of the atomic orbitals contributions to the ground-state
culation of hyperpolarizabilities) for C, H, N, O and H and the molecular orbitals and of the electronic transitions in terms of
LANL2DZ effective core potential basis set for S, P and Ru [51– the contributions of groups of atoms involved. The analysis of the
54]. In the case of the hyperpolarizability calculations the LANL2DZ electronic transitions on the basis of electronic densities of the
basis set was also augmented with a polarization function (expo- orbitals involved is widely used. In particular, the approach of
nents of 0.496 and 0.364) and a diffuse function (exponents of using contributions of groups of atoms involved in electronic tran-
0.0347 and 0.0298) for elements S and P, respectively [55–57]. sitions is also reported [28,70,71]. The eigenvalues connected to
Geometries were optimized at this level of theory without any the Kohn–Sham (KS) orbitals do not have a strict physical meaning,
symmetry constraints followed by the calculations of the first or- especially in a mixed model. However, the literature indicates that
der hyperpolarizabilities. The total static first hyperpolarizability KS orbitals can be used with confidence, at least qualitatively and
b was obtained from the relation: some interpretation of the KS orbital energies is possible. In fact,
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi the reason for the accuracy of TD-DFT excitation energies is that
btot ¼ b2x þ b2y þ b2z ð1Þ the differences of the Kohn–Sham orbital energies are usually
excellent approximations for excitation energies [72]. Chemcraft
upon calculating the individual static components 1.5 program [73] was used for the visualization of computed re-
1X sults, including the representation of the geometries of the com-
bi ¼ biii þ ðb þ bjij þ bjji Þ ð2Þ plexes and orbitals.
3 i–j ijj

Due to the Kleinman symmetry [58]

3. Results and discussion
ðbxyy ¼ byxy ¼ byyx ; byyz ¼ byzy ¼ bzyy ; . . .Þ

one finally obtains the equation that has been employed: The ruthenium(II) complexes (1–7) studied in this work with
atom labelling as well as the optimized geometry for complex 7
btot ¼ ½ðbxxx þ bxyy þ bxzz Þ2 þ ðbyyy þ byzz þ byxx Þ2 are depicted in Fig. 1. Table 1 shows selected structural data for
þ ðbzzz þ bzxx þ bzyy Þ2 1=2 ð3Þ the optimized structures. All complexes have a pseudo-octahedral
three-legged piano stool arrangement of the acetylide and phos-
The electronic spectra for the studied compounds were calcu- phine ligand around the metal atom, on the assumption that the
lated by TD-DFT [59] using the same hybrid functionals and basis cyclopentadienyl (Cp) group occupies three coordination sites.
sets as used for the calculation of the hyperpolarizabilities. The The angles and bond lengths around the ruthenium atom are
 P.J. Mendes et al. / Journal of Molecular Structure: THEOCHEM 946 (2010) 33–42 35

electron-acceptor character of Y seems to be observed. This behav-

iour is accompanied by a shortening of the Ru–C1 bond length as
the electron-acceptor character of Y increases, which is typical for
an enhanced metal-to-ligand p-backdonation interaction. This me-
tal-to-ligand p-backdonation interaction is supported by a decrease
of the C1–C2 bond order of the acetylide ligands upon coordination
(see C1–C2 bond lengths for HAC„C{SC4H2}Y in Ref. [28]) and a
better linearity of the Ru–C1–C2 and C1–C2–C3 geometries, with
angles close to 180°, for electron-acceptor groups. For bond dis-
tances within the thiophene ring, it is observed that C4–C5 bond
length decrease with the electron-acceptor character of Y, whereas
for C3–C4 an opposite trend was found.
The concept of bond-length-alternation (BLA), defined as the
difference between the average carbon–carbon adjacent bond
lengths along a conjugated backbone, was found to be an impor-
tant parameter that correlates with the optical nonlinearities of or-
ganic and organometallic/coordination complexes [75–79]. The
BLA parameter for the complexes 1–7, as well as for the
HAC„C{SC4H2}Y ligands (for comparison), is given in Table 1.
Fig. 1. Structure, atom labelling of the Ru(II) complexes and optimized structure The negative sign of BLA indicates that both in the complexes
for 7. and in the thiophene ligands the neutral resonance form (aromatic)
is the dominant contributor to the ground-state (the neutral and
charge-separated resonance forms for the complexes are depicted
consistent with experimental crystal data for parent ruthenium(II) in (Fig. 2)). However, for electron-acceptor substituents, a less neg-
r-arylacetylides [74] and thiophene nitrile [26] derivatives. Also, ative BLA is found for both complexes and acetylide ligands, which
the bond lengths and angles within the thiophene ligand moiety indicates a relative enhanced contribution of the quinoidal reso-
are typical for ruthenium(II) thiophene derivatives [26]. The Ru– nance form in these series. This trend is more pronounced for the
Cp bond length is almost independent of the presence of different ruthenium complexes for which a major difference between the
Y-substitutents, but a trend of a slight increase with the better BLAs was found (0.027 Å to 0.063 Å vs 0.038 Å to 0.052 Å

Table 1
Selected calculated structural data for the compounds [Ru(g5-Cp)(H2P{CH2}2PH2)(CC{SC4H2}Y)] in vacuum.

Compound
1 2 3 4 5 6 7
Bond lengths (Å)
Ru–Cpa 1.925 1.924 1.924 1.925 1.928 1.929 1.930
Ru–C1 2.018 2.020 2.018 2.015 2.004 2.003 1.998
C1–C2 1.234 1.233 1.233 1.233 1.235 1.235 1.236
C2–C3 1.402 1.401 1.401 1.402 1.395 1.395 1.392
C3–C4 1.370 1.371 1.372 1.376 1.384 1.389 1.390
C4–C5 1.430 1.431 1.430 1.429 1.415 1.411 1.410
C5-C6 1.372 1.366 1.362 1.362 1.377 1.379 1.373
C6-Y 1.390 1.395 1.351 1.080 1.409 1.445 1.418
Ru–Pb 2.385 2.386 2.387 2.388 2.389 2.390 2.391

Bond angles (°)

Ru–C1–C2 178.6 178.3 179.2 179.3 179.8 179.2 179.6
C1–C2–C3 179.1 179.1 179.2 179.8 179.8 179.4 179.4
P1–Ru–P2 83.6 83.6 83.6 83.6 83.5 83.5 83.5
P1–Ru–C1 83.8 83.9 84.0 84.5 85.2 84.9 85.4
P2–Ru–C1 82.5 82.4 82.8 82.9 83.8 83.8 84.1
P–Ru–Cpa,c 132.1 132.0 131.9 131.8 131.5 131.4 131.2
BLAd,e (Å) 0.059 (0.047) 0.062 (0.049) 0.063 (0.052) 0.060 (0.041) 0.034 (0.041) 0.027 (0.038) 0.028 (0.042)
a
Cp centroid.
b
Average Ru–P1/P2 bond lengths (no significant changes was observed for Ru–P1 and Ru–P2 bond distances).
c
Average P1/P2–Ru–Cp bond angles (no significant changes was observed for P1–Ru–Cp and P2–Ru–Cp bond angles).
d
BLA=(dC3–C4 + dC5–C6)/2- dC4–C5.
e
Values for HACC{SC4H2}Y ligands [29] in parentheses for comparison.

Fig. 2. Resonance forms for the [RuCp(H2P{CH2}2PH2)(CC{SC4H2}Y)] complexes.

36 P.J. Mendes et al. / Journal of Molecular Structure: THEOCHEM 946 (2010) 33–42

for HAC„C{SC4H2}Y). The overall data suggests an enhanced has also significant Ru-d character (37%). For the complex 7 this
ground-state polarization due to an increasing of the donor–accep- orbital is mostly localized on the [Ru–Cp–P] moiety, as observed
tor strength, i.e., the better [RuCp(H2P{CH2}2PH2)] electron-donor for LUMO + 1. Compared to this orbital, LUMO + 2 has a higher con-
organometallic moiety (compared to H in acetylide ligands) can tribution of the acetylide group at the detriment of the Cp ring
form a very effective push–pull system in combination with accep- orbitals. Concerning the energies, an overall stabilizing effect is ob-
tor substituents in the thiophene acetylide moiety. Thus, higher served for all the studied orbitals from the complex with the better
first hyperpolarizabilities for Ru complexes with acceptor groups electron-donor (1) to the one with the better electron-acceptor
are expected, according to the behaviour found in related iron(II) substituent (7), in particular for the complexes 5–7. Within the
compounds [28]. set of complexes with electron-donor substituents, as the elec-
For what concerns the electronic structure of the complexes, the tron-donor character decreases (from 1 to 3), the HOMOs are more
orbital energies and composition in terms of the atom groups’ con- stabilized than the LUMOs. For the complexes 5–7, as the electron-
tributions for selected sets of HOMOs and LUMOs for the com- acceptor character increases, an opposite relative stabilization is
plexes 1, 4 and 7 are reported in Table 2 (complexes 1 and 7 observed, i.e., LUMOs are further stabilized with respect to the
represent the group of complexes with electron-donor and elec- HOMOs. This effect is significant on the HOMO and LUMO frontier
tron-acceptor substituents, respectively, whereas 4, with Y @ H, orbitals; in particular for LUMO in the complex 7 (the strong elec-
is included for comparison). The selected orbitals and correspond- tron-acceptor NO2 substituent in the thiophene ligand leads to an
ing energy levels are depicted in Figs. 3 and 4, respectively. The enhanced stabilization of the LUMO). As a result, the HOMO–LUMO
largest contributions of HOMO-1, HOMO-2 and HOMO-3 arise gap decreases in the order: 5 (3.49 eV) > 6 (3.26 eV) >> 7 (2.86 eV).
from the [Ru–Cp–P] moiety, in particular from metal orbitals Since these frontier orbitals are involved in the main electronic
whose participation ranges from 37% to 53% for all the complexes. transition for this set of complexes (see optical data below), the ob-
Also important contributions of the thiophene ligand groups are served trend should favour the second-order nonlinear optical
found for HOMO-3 and HOMO-1, in particular the thiophene ring properties of the complex 7.
group for HOMO-3 and the acetylide group for HOMO-1. The par- The calculated first static hyperpolarizabilities for the com-
ticipation of the substituents (Y) is only sizable for the complex 1, plexes are shown in Table 3. For 1–3, btot values are rather modest
in particular for HOMO-3. The HOMO is mostly spread over the Ru– and comparable with the corresponding uncoordinated acetylide
CC–T–Y axis for the studied complexes. The largest contribution for ligands, whereas for the remaining complexes the first hyperpolar-
this orbital arises from the thiophene ligand groups, in particular izability is significantly higher and comparable with that found in
for 1 (83%). The relative contribution of these groups to HOMO analogue iron derivatives [28]. According to the structural results
lowers from 1 to 7, which is accompanied by an increasing partic- discussed above, the increasing contribution of the quinoidal
ipation of the Ru-d orbital, from 14% (1) to 28% (7). The shapes of resonance form for the complexes with acceptor substituents
the sets of LUMOs are strongly dependent on the substituent in would result in an effective electronic p-coupling between the
the thiophene ligand. For 1 and 4 the LUMO and LUMO + 1 are cyclopentadienylruthenium(II) moiety and those substituents
clearly centred in the [Ru–Cp–P] moiety whereas for 7, considering through the p-system of the thiophene path. For typical push–pull
these two orbitals, only the LUMO + 1 maintains the same charac- molecules, this can give an adequate degree of mixing between
ter. The largest contribution for LUMO in this complex arises from neutral and charge-separated resonance forms [75–77] and, conse-
the thiophene ligand orbitals, in particular from the nitro and thi- quently, enhanced first hyperpolarizabilities since that will be the
ophene ring groups. The LUMO + 2 also show important differences key to intramolecular charge transfer. In order to study the influ-
within the complexes. This orbital is mostly spread over the Ru– ence of the Ru–C1 bond length, and hence the interaction between
CC–T–Y axis for 1 and 4. In the complex 4, LUMO + 2 is mainly the cyclopentadienylruthenium(II) moiety and the thiophene li-
localized in the thiophene ring group whereas for 1 this orbital gand, on the value of the first hyperpolarizability for compound
7 (compound with higher b), this property was also calculated
for different Ru-C1 bond distances, in the range usually reported
Table 2 for Ru–C r-bound ligands. Fig. 5 shows that a maximum of b is ob-
Energies and percentage composition in terms of the represented groups, of selected tained for ca 2.00 Å, a value that corresponds to the Ru–C1 bond
orbitals for the complexes 1, 4 and 7 in vacuum. length for the optimized structure.
Compound MO E (eV) Ru Cp P CC T Y In our recent studies on analogue iron(II) derivatives [28], we
found that the most intense electronic transition governs the static
1 LUMO + 2 0.27 37 1 10 12 38 2
LUMO + 1 0.64 38 2 55 0 0 0 first hyperpolarizabilities and the two-level model (TLM) [41]
LUMO 0.98 37 22 40 1 0 0 seems to be applicable for the set of complexes with Y = H, CN,
HOMO 4.21 14 2 1 27 43 13 CHO with some deviation for Y = NO2 for which the TLM underes-
HOMO-1 5.10 43 7 5 37 6 2 timates the calculated static quadratic hyperpolarizability. This
HOMO-2 5.56 53 15 8 7 9 8
HOMO-3 5.84 40 11 5 19 11 14
model, which is a good approximation for estimating the second-
order polarizabilities in donor/acceptor NLO chromophores, as-
4 LUMO + 2 0.50 10 0 2 18 70 0
LUMO + 1 0.72 44 7 36 13 0 0
sumes that for the contribution to b only one excited state is cou-
LUMO 1.10 36 22 39 1 2 0 pled strongly enough to the ground-state by the applied electric
HOMO 4.63 24 4 3 34 35 0 field, and only one component of the b tensor dominates the NLO
HOMO-1 5.40 48 5 4 39 4 0 response (i.e., a unidirectional charge-transfer transition). With
HOMO-2 5.83 51 22 10 15 2 0
these assumptions, the second-order nonlinearity can be described
HOMO-3 6.24 41 9 5 10 35 0
as:
7 LUMO + 2 1.06 40 8 35 17 0 0
LUMO + 1 1.51 38 24 36 1 0 1 Dle:g: fe:g:
LUMO 2.41 2 0 1 9 39 49 b/ ð4Þ
HOMO 5.27 28 5 4 29 27 7 E3e:g:
HOMO-1 5.96 53 5 4 34 4 0
HOMO-2 6.35 53 21 9 17 0 0 where Dleg is the difference between the dipole moments of the
HOMO-3 6.80 37 10 4 11 34 4
ground (g) and the excited (e) state, feg is the oscillator strength
CC: acetylide group, T: tiophene ring, P: bidentate phosphine, Cp: cyclopentadienyl. and Eeg is the transition energy. Those factors (Dleg, feg and Eeg)
 P.J. Mendes et al. / Journal of Molecular Structure: THEOCHEM 946 (2010) 33–42 37

Fig. 3. Representation of selected orbitals for complexes 1, 4 and 7.

factors will provide a higher b value. For the complexes studied in

this work, btot is clearly dominated by the bxxx tensor component
(along the charge transfer axis - vide infra - Table 3) and an analysis
of the results according to the TLM seems to be reasonable for these
complexes.
The energies, oscillator strengths and the composition of the
main electronic transitions (only those with f P 0.10) in terms of
the contributions of groups of atoms involved, obtained by TD-
DFT calculations, are given in Table 4. The main optical feature
for 1–6 is the presence of an intense electronic transition (ET)
attributed to HOMO?LUMO + 2 or HOMO?LUMO charge transfer,
depending on the Y-substituent. For 7, besides an intense low-en-
ergy electronic transition attributed to HOMO?LUMO, two ETs at
higher energies are also found. In general, lower energy, higher
oscillator strength and higher difference between the dipole mo-
ments of the ground and the excited state for the main electronic
transition were obtained for the complexes with electron-acceptor
substituents. According to the TLM, this would enhance the sec-
ond-order NLO properties of this set of complexes, which is in
agreement with the calculated first hyperpolarizabilities. For the
Fig. 4. Energy levels of selected orbitals for complexes 1, 4 and 7. complexes with electron-releasing substitutents (1–3), the charac-
ter of the ETs can be assigned mainly to a thiophene ligand-to-me-
are all closely related, which are controlled by the electronic prop- tal charge transfer (LMCT) with some contribution of a CT within
erties of push–pull molecules. An optimal combination of the the organometallic fragment (i.e., from the Cp ring to the Ru-phos-
38 P.J. Mendes et al. / Journal of Molecular Structure: THEOCHEM 946 (2010) 33–42

Table 3
b components and btot values (1030 esu) for the compounds [Ru(g5-Cp)(H2P{CH2}2PH2)(CC{SC4H2}Y)].

Comp. bxxx bxxy bxyy byyy bxxz bxyz byyz bxzz byzz bzzz btot
1 5.41 1.42 0.92 2.03 0.06 0.83 0.25 0.28 1.21 0.61 4.66
2 6.06 0.67 1.52 2.96 0.64 1.09 0.06 1.17 1.74 0.27 5.35
3 10.02 0.15 0.85 3.26 1.15 0.83 0.31 1.36 2.05 0.77 9.62
4 18.44 0.56 1.17 3.09 0.10 0.24 0.05 0.65 1.94 0.18 17.22
5 51.97 0.07 1.73 2.37 0.49 0.20 0.00 1.79 1.48 0.20 48.68
6 68.85 0.67 1.34 1.57 0.45 0.20 0.03 1.66 1.30 0.22 65.89
7 103.58 2.91 2.22 1.90 0.85 0.12 0.02 2.26 1.31 0.16 99.29

generated a charge transfer process between the metal moiety

and the Y-substituent. This donor-to-acceptor CT leads, compared
to the behaviour found for 1–3 complexes, to an effective charge-
transfer upon the excitation and thus higher first hyperpolarizabil-
ities. As the electron-withdrawing character of the Y-substituent
increases, a bathochromic shift and an enhanced contribution of
Y on the ETs are observed for these complexes. This red shift leads
to enhanced first hyperpolarizabilities as expected considering the
TLM. Nevertheless, Dleg and feg follow the opposite trend, which
can hamper the b values. According to the TLM, a linear correlation
is expected from a plot of the btot vs Dle:g: :f =E3e:g: if we assume that
only two electronic states and one tensor component along the
charge-transfer transition dominates the second-order NLO re-
sponse (Eq. 4). As mentioned above, for the complexes studied in
this work, btot  bxxx (the x-axis is directed along charge transfer
axis). If we consider that the low-energy intense electronic transi-
tion governs the static first hyperpolarizability, the quantitative
application of the TLM (Fig. 6) shows a large deviation for the com-
plex 7. This can be due to the fact that more than one excitation
contributes effectively to the magnitude of b for this compound.
Fig. 5. Effect of the Ru–C1 bond length on the first hyperpolarizability for
In fact, two additional important ETs were found for compound 7
compound 7.
(Table 4). The higher-energy HOMO-3?LUMO excitation arises
mainly from the organometallic fragment (with some contribution
phine moiety). For these complexes, the donor character of both Y- of the thiophene ring) to the nitro group, leading to a large increase
substituent and ruthenium moiety leads to an ineffective charge- on the dipolar moment upon excitation (Dleg = 7.160 D). This do-
transfer upon the excitation (small Dleg and feg values) and thus nor-to-acceptor CT can also contributes to the first hyperpolariz-
rather modest first hyperpolarizabilities. ability. The second optical transition (HOMO?LUMO + 1), closer
For 4–7 complexes, the main electronic transition is mostly a in energy to the HOMO?LUMO, can be attributed mainly to li-
metal-to-ligand charge transfer (MLCT) with some contribution gand-to-metal charge transfer (LMCT) since the CT arises mainly
of a CT within the organometallic fragment. For instance, from from the acetylide group and thiophene ring to the Cp and phos-
the electron densities of HOMO and LUMO for complex 7 shown phine co-ligands, as deduced from the corresponding electron den-
above, representative of the orbitals involved in the main ET for sities of the orbitals involved. This leads to a decrease in the dipolar
the set of compounds with acceptor substituents, it can be seen moment upon excitation and thus in a lower b.
that the HOMO has a significant contribution of the [RuAC„C] It is well-known that solvent polarity plays a significant role on
fragment whereas the LUMO are located mainly in the nitro accep- the geometry and electronic structure as well as on the first hyper-
tor. Thus, the ET between the occupied and unoccupied orbitals polarizabilities in dipolar molecules. In order to get an insight on

Table 4
Optical data for the compounds [Ru(g5-Cp)(H2P{CH2}2PH2)(CC{SC4H2}Y)] obtained by TD-DFT calculations.

Comp. kega (nm) Eegb (eV) fc Attribd Character of the CTe lg (D) le (D) Dl (D) btot  10–30 (esu)
1 351 3.53 0.234 H?L + 2 [CC-T-Y] (69); [Cp] (31)?[Ru-P] 3.888 4.361 0.473 4.66
2 346 3.59 0.175 H?L + 2 [CC-T-Y] (60); [Cp] (40)?[Ru-Cp] 3.659 4.752 1.093 5.35
3 345 3.60 0.232 H?L + 2 [CC-T-Y] (62); [Cp] (38)?[Ru-P] 6.067 6.399 0.332 9.62
4 338 3.67 0.284 H?L + 2 [Ru-Cp-CC] ? [T] (79); [P] (21) 5.381 8.145 2.764 17.22
5 387 3.20 0.498 H?L [Ru-Cp-CC] ? [T] (57); [P] (33); [Y] (10) 11.105 18.373 7.268 48.68
6 408 3.04 0.495 H?L [Ru-Cp-CC] ? [T] (36); [P] (18); [Y] (46) 8.939 14.951 6.013 65.89
7 470 2.64 0.356 H?L [Ru-Cp-CC] ? [T] (3); [P] (22); [Y] (75) 15.996 3.290
451 2.75 0.169 H?L + 1 [CC-T] ? [Cp-P] (86); [Y] (14) 12.706 10.585 2.121 99.29
310 4.01 0.103 H-3?L [Ru-Cp] (79); [T] (21] ? [Y] 19.866 7.160

Cp: cyclopentadienyl, CC: acetylide group, T: tiophene ring, P: bidentate phosphine.

a
Absorption wavelength of the main transitions.
b
Energy of the transition.
c
Oscillator strength.
d
H-HOMO, L-LUMO.
e
Based on the represented fragments (overall % of the charge transfer in parentheses).
 P.J. Mendes et al. / Journal of Molecular Structure: THEOCHEM 946 (2010) 33–42 39

separated resonance form. Accordingly, the BLA parameter for

the thiophene ring is less negative and the ground-state dipole mo-
ments increase with the solvent polarity.
Inclusion of solvation effects leads also to changes on the
molecular orbital energies, in spite of maintaining a rather similar
composition in terms of the group contributions. In view of the
orbitals composition, in spite of the small differences, only a siz-
able change seems to be observed for HOMO-3 (the contribution
of C„C group is slightly increased at expenses of the Ru contribu-
tion) and HOMO (an enhanced contribution of Ru at expenses of
C„C and thiophene ring groups contributions). The effect of the
solvated media on the energies of selected orbitals are shown in
Table 6 and Fig. 7. In solution, the set of HOMOs are destabilized,
by 0.04–0.18 eV, with respect to the corresponding values in
vacuum. The destabilizing effect follows the order HOMO-
3 < HOMO < HOMO-1 < HOMO-2 and is more pronounced from
vacuum to chloroform since the energies of these orbitals for dif-
Fig. 6. Application of the two-level model to the complexes 1–7.
ferent solvents are very similar. Only a small increase (up to
0.04 eV) on the corresponding energies is observed as the solvent
polarity increases. On the other hand, the energies of LUMOs in sol-
the solvent-structure, solvent-optical and solvent-SONLO depen- vated media are lower than the corresponding values computed in
dence of the compound 7, for which the better first hyperpolariz- vacuum. Whereas the LUMO is stabilized by ca 0.18–0.25 eV, the
ability was found, we have used a self-consistent reaction field LUMO + 1 is destabilized in the same extent. For both LUMO orbi-
(SCRF) approach with polarizable continuum model (PCM) for sim- tals, as was observed for the set of HOMOs, the changes are more
ulating the interaction of the complex with chloroform, acetone evident from vacuum to chloroform (ca 0.18 eV) than within the
and methanol. different solvents (0.06–0.07 eV). As the solvent polarity increases,
The structural data for the optimized structures in the three the LUMO becomes more stabilized and LUMO + 1 is destabilized,
studied solvents is shown in Table 5 (the data in vacuum were also which results in an enhanced energy difference between these
included for comparison). The results show that the structure of orbitals. Altogether, the presence of the solvent leads to a decrease
compound 7 is affected by the polarity of the surrounding media of the HOMO–LUMO, HOMO-1-LUMO and HOMO-3-LUMO gaps,
with the main following general trends: (1) shortening of Ru–C1, which changes from 2.86, 3.55 and 4.39 eV in the gas-phase,
C2–C3, C4–C5 and C6–Y bonds; (2) lengthening of C1–C2, C3–C4,
and C5–C6 bonds and (3) slight deviation in the linearity of the
Ru-C1–C2–C3 system. In spite of the slight deviation in linearity Table 6
of the Ru–C1–C2–C3 system, the overall data are consistent with Energies of selected orbitals for the complex 7 in vacuum and solvated media.
an increasing contribution of the quinoidal resonance form in the MO E (eV)
ground-state with increasing solvent polarity (Fig. 2). This is not
Vacuum Chloroform Acetone Methanol
surprising since dipolar solvents will help to stabilize the charge-
LUMO + 1 1.51 1.32 1.27 1.26
LUMO 2.41 2.59 2.66 2.66
HOMO 5.27 5.20 5.19 5.19
Table 5 HOMO-1 5.96 5.85 5.83 5.83
Solvent-dependent calculated structural data and ground-state dipole moments for HOMO-2 6.35 6.21 6.17 6.16
the compound 7. HOMO-3 6.80 6.76 6.76 6.76

Vacuum Chloroform Acetone Methanol

Bond lengths (Å)
Ru-Cpa 1.930 1.934 1.934 1.934
Ru–C1 1.998 1.985 1.981 1.980
C1–C2 1.236 1.240 1.241 1.242
C2–C3 1.392 1.385 1.382 1.382
C3–C4 1.390 1.396 1.398 1.399
C4–C5 1.410 1.404 1.402 1.402
C5–C6 1.373 1.378 1.379 1.380
C6–Y 1.418 1.407 1.403 1.403
Ru–Pb 2.391 2.394 2.396 2.397
Bond angles (°)
Ru–C1–C2 179.6 178.4 177.7 177.8
C1–C2–C3 179.4 179.1 178.9 178.8
P1–Ru–P2 83.5 83.3 83.2 83.2
P1–Ru–C1 85.4 87.2 87.7 87.7
P2–Ru–C1 84.1 85.8 86.5 86.5
P–Ru–Cpa,c 131.2 130.2 129.8 129.8
BLAd (Å) 0.028 0.017 0.014 0.013
lg(Debye) 12.706 16.678 18.160 18.276
a
Cp centroid.
b
Average Ru–P1/P2 bond lengths (no significant changes was observed for Ru–P1
and Ru–P2 bond distances).
c
Average P1/P2-Ru-Cp bond angles (no significant changes was observed for P1–
Ru–Cp and P2–Ru–Cp bond angles). Fig. 7. Energy levels of selected orbitals for complex 7 in vacuum, chloroform,
d
BLA=(dC3–C4 + dC5–C6)/2-dC4–C5. acetone and methanol.
40 P.J. Mendes et al. / Journal of Molecular Structure: THEOCHEM 946 (2010) 33–42

respectively, to 2.53, 3.17 and 4.10 eV in methanol and an increase

of the HOMO–LUMO + 1 gap from 3.76 eV in the gas-phase to
3.93 eV in methanol.
These changes play an important role in the optical data and
first hyperpolarizabilities for the complex 7 when solvation effects
are taken into account (Table 7). The calculated optical spectra in
methanol and vacuum (for comparison) are shown in Fig. 8 (the
spectra of the complex in chloroform and acetone are qualitatively
similar to that found in methanol). As was observed in vacuum, the
spectrum in solvated media shows a low-intensity band mainly
attributed to a single ET and an intense lowest-energy band which
has a contribution of two excitations closer in energy. In solvated
media, the highest-energy band (attributed mainly to HOMO-
3?LUMO ET) can be clearly assigned to a MLCT, according to the
composition of the orbitals involved, since the excitation arises
mainly from the organometallic fragment to the nitro group (with
a small contribution of the thiophene ring). Compared to the
behaviour found in vacuum, the main difference in solvated media
Fig. 8. TD-DFT calculated absorption spectra for 7 in vacuum (- - -) and methanol
is the increasing of electronic density on the thiophene ring upon (—).
excitation, which contributes for an improved efficiency of the
charge transfer between the organometallic fragment and the thi- 1?LUMO) in solvated media. This is not surprising since, as men-
ophene ligand. The energy of this ET is red-shifted when the sol- tioned above, the presence of the solvents leads to an important
vent effects are taken into account and follow the same trend decrease of the HOMO-1-LUMO and an increase of the HOMO–
that was observed for the HOMO-3-LUMO gap in different solvents. LUMO + 1 gaps. It is also interesting to note that the relative inten-
The main changes of the inclusion of solvation effects are observed sity of the two ETs in solvated media depends on the solvent polar-
on the intense lowest-energy band, which possesses a contribution ity: as the solvent polarity increases, the oscillator strength of
of two ETs in both gas-phase and solvated media. In gas-phase the HOMO-1?LUMO is enhanced whereas for HOMO?LUMO a
lowest-energy ET (HOMO?LUMO) is assigned to a MLCT with decreasing effect is observed, thus leading to a larger contribution
some contribution of CT within the organometallic fragment, of the former ET to the lowest-energy band in the optical spectra.
according to the composition of the corresponding orbitals, The resulting band is centred at 2.50, 2.46 and 2.45 eV for
whereas the second ET is assigned mainly to a LMCT, in which chloroform, acetone and methanol, respectively, thus showing a
the LUMO + 1 orbital is involved in the excitation (HOMO?LU- red-shift with respect to gas-phase and with increasing solvent
MO + 1). These two contributions results in a band centred at polarity. This positive solvatochromism, characteristic of electronic
2.68 eV. For all the studied solvents, the two contributing ETs transitions with an increase of the dipole moment upon photo-
for the considered band have very similar character and attribu- excitation, can be related to an improved stabilization of the LUMO
tion: the lower-energy ET is assigned to HOMO?LUMO and the for more polar solvents. In fact, the energies of HOMO-1 and HOMO
second ET is mainly assigned to HOMO-1?LUMO CT. These ETs are relatively insensitive to the presence of the different studied
can be clearly attributed to MLCT excitations i.e., arising from the solvents whereas the LUMO is stabilized by ca 0.07 eV from chloro-
organometallic fragment to the thiophene ring (ca 30–40%) and ni- form to methanol (Table 6). A positive solvatochromism was found
tro group (ca 60–70%), according to the composition of the orbitals experimentally for the related compound [RuCp(dppe)(CC{SC4H2}-
involved. This is an important difference when compared to the NO2] (dppe: 1,2-Bis(diphenylphosphino)ethane) [27]. Also, the
behaviour found in gas-phase. In vacuum, the second ET has a main spectral features are well reproduced by the theoretical mod-
LMCT character in which the LUMO + 1 orbital is involved in the el with, for instance, a blue-shift by ca 0.06 eV with respect to the
excitation (HOMO?LUMO + 1) instead of the HOMO-1 (HOMO- experiment for the lowest-energy band.

Table 7
Optical data and first hyperpolarizabilities (btot) for compound 7 in different surrounding media.

kega (nm) Eegb (eV) fc Attribd Character of the CTe btot  1030 (esu)
Vacuum 470 2.64 0.356 H?L [Ru-Cp-CC] ? [T] (3); [P] (22); [Y] (75)
451 2.75 0.169 H?L + 1 [CC-T] ? [Cp-P] (86); [Y] (14) 99.29
310 4.01 0.103 H-3?L [Ru-Cp] (79); [T] (21] ? [Y]
Chloroform 501 2.48 0.426 H?L [Ru-Cp-P-CC] ? [T] (31); [Y] (69)
480 2.59 0.157 H-1?L [Ru-Cp-P-CC] ? [T] (39); [Y] (61) 229.30
325 3.82 0.120 H-3?L [Ru-Cp-P-CC] ? [T] (5); [Y] (95)
Acetone 516 2.40 0.300 H?L [Ru-Cp-P-CC] ? [T] (34); [Y] (66)
492 2.53 0.309 H-1?L [Ru-Cp-P-CC] ? [T] (37); [Y] (63) 290.28
330 3.77 0.110 H-3?L [Ru-Cp-P-CC] ? [T] (10); [Y] (90)
Methanol 518 2.40 0.285 H?L [Ru-Cp-P-CC] ? [T] (34); [Y] (66)
494 2.52 0.325 H-1?L [Ru-Cp-P-CC] ? [T] (36); [Y] (64) 298.81
330 3.76 0.106 H-3?L [Ru-Cp-P-CC] ? [T] (11); [Y] (89)

Cp: cyclopentadienyl, CC: acetylide group, T: Tiophene ring, P: bidentate phosphine; fbtot  bxxx (the x-axis is directed along charge transfer axis) in all surrounding media.
a
Absorption wavelength of the main transitions.
b
Energy of the transition.
c
Oscillator strength.
d
H-HOMO, L-LUMO.
e
Based on the represented fragments (overall% of the charge transfer in parentheses).
 P.J. Mendes et al. / Journal of Molecular Structure: THEOCHEM 946 (2010) 33–42 41

The overall results suggest an expected improvement of the first

hyperpolarizabilities for the compound 7 when the solvation ef-
fects are taken into account, as shown in Table 7. The b values in-
crease more than two times from vacuum to chloroform and in a
lesser degree within the different solvents. The dependence of
the first hyperpolarizability of the studied compound both on the
dielectric constant of the media and the Onsager function [80]
(Fig. 9) is typical for a dipolar reaction field interaction in the sol-
vation process [43,80–82]. Thus, the electronic reorganization that
takes place in solution for compound 7 plays an important role on
the resulting first hyperpolarizabilities. It is interesting to note a
linear dependence of b values with BLA in different media for this
compound (Fig. 10), which reveals that the degree of charge delo-
calization in the ground-state can be important concerning the
SONLO properties of this compound. The involvement of the
HOMO-1 in the electronic transfer in solvated media and its grow-
ing importance as the solvent polarity increases seems also to play
a role. In fact, as mentioned above, the HOMO-1?LUMO ET results Fig. 10. Dependence of b on BLA values for complex 7 in vacuum and solvated
in an effective donor-to-acceptor charge transfer in solvated media media.
whereas for gas-phase the involvement of the LUMO + 1 in the
HOMO?LUMO + 1 ET (LMCT) can result in an hampering effect
on b. In addition, the HOMO?LUMO excitation gives a further organometallic fragment and the thiophene ligand was also identi-
effectiveness of donor-to-acceptor charge transfer in solvated fied in solvated media. Thus, the better first hyperpolarizabilities
media (clearly a MLCT) since in the gas-phase this ET has some for the compound 7 in solvated media when compared to that
contribution of a charge transfer within the organometallic frag- found in gas-phase can be explained by both lower energies and
ment. Regarding the highest-energy ET (HOMO-3?LUMO) an im- better CT efficiencies from the ruthenium moiety to the thiophene
proved effectiveness of the charge transfer between the ligand.

4. Conclusions

The electronic, structural and spectroscopic properties as well

as the static first hyperpolarizabilities of the complexes
[RuCp(H2PCH2CH2PH2)(CC{SC4H2}Y)] (Y = NMe2, NH2, OMe, H,
CHO, CN, NO2) in gas-phase and, for the complex [RuCp(H2PCH2
CH2PH2)(CC{SC4H2}NO2)], also in chloroform, acetone and metha-
nol have been investigated by means of combined DFT/TD-DFT cal-
culations. The studies have shown that, as expected, the ruthenium
organometallic fragment is an effective p-donor in combination
with electron-acceptor substituents in the thiophene acetylide li-
gand. From the BLA formalism, an enhanced ground-state polariza-
tion due to an increasing of the donor–acceptor strength was
identified for these complexes, which is further improved in sol-
vated media for the complex [RuCp(H2PCH2CH2PH2)(CC{SC4H2}-
NO2)]. The improved coupling between the organometallic
ruthenium electron-donor and the electron-acceptor groups (CN,
CHO and NO2) in the studied push–pull complexes explains the
better static first hyperpolarizabilities for these complexes. The
analysis of the main electronic transitions composing the optical
spectra of the studied compounds has provided a deeper insight
into the electronic factors governing the corresponding second-or-
der nonlinear optical properties. In gas-phase studies, lower ener-
gies, higher oscillator strengths and higher differences between the
dipole moments of the ground and the excited state for the main
electronic transition were obtained for the complexes with elec-
tron-acceptor substituents. These combined factors enhance the
first hyperpolarizabilities, according to the two-level model. Nev-
ertheless, the application of this model underestimates the calcu-
lated quadratic hyperpolarizability for the complex [RuCp(H2P
CH2CH2PH2)(CC{SC4H2}NO2)], for which the better first hyperpo-
larizability was found, since additional excited states seem to
contribute effectively to the magnitude of b. The first hyperpolar-
i-zability of this compound is further increased if solvation effects
are taken into account in the calculations. The better SONLO prop-
erties of this compound in solvated media when compared to that
Fig. 9. Dependence of b for compound 7 on the dielectric constant (a) and Onsager found in gas-phase can be related to both lower energies and better
function (b). charge-transfer efficiencies arising from the ruthenium moiety to
42 P.J. Mendes et al. / Journal of Molecular Structure: THEOCHEM 946 (2010) 33–42

the thiophene ligand. This is a result of a suitable electronic reorga- [34] Chaoyong Mang, Kechen Wu, Int. J. Quant. Chem. 106 (2006) 2529.
[35] Jian Lin, Kechen Wu, Mingxin Zhang, J. Comput. Chem. 30 (2009) 2056.
nization that takes place in solvated media, as suggested from
[36] Y. Liu, Y.H. Geng, Y.J. Liu, C.B. Liu, Chin. Chem. Lett. 13 (2002) 45.
structural data and energies and composition of important orbitals, [37] P.S. Liyanage, R.M. de Silva, K.M.N. de Silva, J. Mol. Struct. (Theochem) 639
which produces important changes on their optical properties. The (2003) 195.
magnitude of b was found to increase with the solvent polarity. The [38] C.E. Powell, M.P. Cifuentes, A.M. McDonagh, S.K. Hurst, N.T. Lucas, C.D. Delfs, R.
Stranger, M.G. Humphrey, S. Houbrechts, I. Asselberghs, A. Persoons, D.C.R.
observed dependence is typical for a dipolar reaction field interac- Hockless, Inorg. Chim. Acta 352 (2003) 9.
tion in the solvation process. [39] C. Mang, K. Wu, M. Zhang, T. Hong, Y. Wei, J. Mol. Struct. (Theochem) 674
(2004) 77.
[40] S. Curreli, P. Deplano, C. Faulmann, A. Ienco, C. Mealli, M.L. Mercuri, L. Pilia, G.
Acknowledgement Pintus, A. Serpe, E.F. Trogu, Inorg. Chem. 43 (2004) 5069.
[41] Qiaohong Li, Rongjian Sa, Caiping Liu, Kechen Wu, J. Phys. Chem. A 111 (2007)
7925.
The authors are greatly thankful to Fundação para a Ciência e
[42] P.C. Ray, Chem. Phys. Lett. 395 (2004) 269.
Tecnologia (FCT) for financial support (Project n° PTDC/QUI/ [43] P.C. Ray, Jerzy Leszczynski, Chem. Phys. Lett. 399 (2004) 162.
67362/2006). [44] Talgat M. Inerbaev, Rodion V. Belosludov, Hiroshi Mizuseki, Masae Takahashi,
Yoshiyuki Kawazoe, J. Chem. Theory Comput. 2 (2006) 1325.
[45] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,
References J.A. Montgomery Jr., T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar,
J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A.
[1] D.R. Kanis, M.A. Ratner, T.J. Marks, Chem. Rev. 94 (1994) 195. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
[2] N.J. Long, Angew. Chem., Int. Ed. Engl. 34 (1995) 21. M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X.Li, J.E. Knox,
[3] S. Di Bella, Chem. Soc. Rev. 30 (2001) 355. H.P. Hratchian, J.B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.E,
[4] E. Goovaerts, W. Wenseleers, M.H. Garcia, G.H. Cross, in: H.S. Nalwa (Ed.), Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pommelli, J.W. Ochterski, P.Y.
Handbook of Advanced Electronic and Photonic Materials, Vol. 9, Academic Ayala, K. Morokuma, G.A. Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S.
Press, 2001, p. Chapter 3. Dapprich, A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K.
[5] F. Paul, K. Costuas, I. Ledoux, S. Deveau, J. Zyss, J.F. Halet, C. Lapinte, Raghavachari, J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J.
Organometallics 21 (2002) 5229. Cioslowski, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.L.
[6] C.E. Powell, M.P. Cifuentes, J.P. Morrall, R. Stranger, M.G. Humphrey, M. Samoc, Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, M.
B. Luther-Davies, G.A. Heath, J. Am. Chem. Soc. 125 (2003) 602. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, .
[7] Clem E. Powell, Mark G. Humphrey, Coord. Chem. Rev. 248 (2004) 725. [46] A.D. Becke, J. Chem. Phys. 98 (1993) 5648.
[8] O. Maury, L. Viau, K. Senechal, B. Corre, J.P. Guegan, T. Renouard, I. Ledoux, J. [47] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785.
Zyss, H. Le Bozec, Chem. Eur. J. 10 (2004) 4454. [48] Yan Zhao, Donald G. Truhlar, Acc. Chem. Res. 41 (2008) 157.
[9] Maurizio Bruschi, Piercarlo Fantucci, Maddalena Pizzotti, J. Phys. Chem. A 109 [49] Sérgio Filipe Sousa, Pedro Alexandrino Fernandes, Maria João Ramos, J. Phys.
(2005) 9637. Chem. A 111 (2007) 10439.
[10] Ireshika C. de Silva, Rohini M. de Silva, K.M. Nalin de Silva, J. Mol. Struct. [50] C. Cardoso, P.E. Abreu, F. Nogueira, J. Chem. Theory Comput. 5 (2009) 850.
(Theochem) 728 (2005) 141. [51] P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 270.
[11] P.J. Mendes, J.P. Prates Ramalho, A.J.E. Candeias, M.P. Robalo, M.H. Garcia, J. [52] P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 284.
Mol. Struct. (Theochem) 729 (2005) 109. [53] P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 299.
[12] Jean-Luc Fillaut, Johann Perruchon, Philippe Blanchard, Jean Roncali, Stephane [54] T.H. Dunning, P.J. Hay, in: H.F. Schaefer (Ed.), Modern Theoretical Chemistry,
Golhen, Magali Allain, Anna Migalsaka-Zalas, Ivan V. Kityk, Bouchta Sahraoui, Vol. 3, III Plenum, New York, 1976, p. 1.
Organometallics 24 (2005) 687. [55] C.E. Check, T.O. Faust, J.M. Bailey, B.J. Wright, T.M. Gilbert, L.S. Sunderlin, J.
[13] B.J. Coe, Acc. Chem. Res. 39 (2006) 383. Phys. Chem. A 105 (2001) 8111.
[14] Elena Cariati, Maddalena Pizzotti, Dominique Roberto, Francesca Tessore, [56] D. Feller, J. Comp. Chem. 17 (1996) 1571.
Renato Ugo, Coord. Chem. Rev. 250 (2006) 1210. [57] K.L. Schuchardt, B.T. Didier, T. Elsethagen, L. Sun, V. Gurumoorthi, J. Chase, J. Li,
[15] Maik Malessa, Jürgen Heck, Jürgen Kopf, M. Helena Garcia, Eur. J. Inorg. Chem. T.L. Windus, J. Chem. Inf. Model. 47 (2007) 1045.
(2006) 857. [58] D.A. Kleinman, Phys. Rev. 126 (1962) 1977.
[16] Ping Yuan, Jun Yin, Guang-ao Yu, Quanyuan Hu, Sheng Hua Liu, [59] E. Runge, E.K.U. Gross, Phys. Rev. Lett. 52 (1984) 997.
Organometallics 26 (2007) 96. [60] M.W. Schmidt, K.K. Baldridge, J.A. Boatz, S.T. Elbert, M.S. Gordon, J.H. Jensen, S.
[17] J.P. Morral, G.T. Dalton, M.G. Humphrey, M. Samoc, Adv. Organomet. Chem. 55 Koseki, N. Matsunaga, K.A. Nguyen, S.J. Su, T.L. Windus, M. Dupuis, J.A.
(2008) 61. Montgomery, J. Comput. Chem. 14 (1993) 1347.
[18] Ian R. Whittall, Mark G. Humphrey, Organometallics (1996) 1935. [61] A. Dreuw, M. Head-Gordon, Chem. Rev. 105 (2005) 4009.
[19] WimWenseleers, Abraham W. Gerbrandij, Etienne Goovaerts, M. Helena [62] A. Dreuw, G.R. Fleming, M. Head-Gordon, Phys. Chem. Chem. Phys. 5 (2003)
Garcia, M. Paula Robalo, Paulo J. Mendes, João C. Rodrigues, Alberto R. Dias, J. 3247.
Mater. Chem. 8 (1998) 925. [63] R. Cammi, B. Menucci, J. Tomasi, J. Phys. Chem. A 104 (2000) 4690.
[20] M.H. Garcia, A.R. Dias, M. PaulaRobalo, M.G. Humphrey, A.M. McDonagh, S. [64] Y. Luo, P. Norman, P. Macak, H. Agren, J. Chem. Phys. 111 (1999) 9853.
Urst, E. Goovaerts, W. Wenseleers, Organometallics 21 (2002) 2107. [65] B. Mennucci, J. Tomasi, J. Chem. Phys. 106 (1997) 5151.
[21] E. Goovaerts, W. Wenseleers, P. Hepp, M.H. Garcia, M.P. Robalo, A.R. Dias, [66] A. Bacchi, M. Balordi, R. Cammi, Eur. J. Inorg. Chem. 28 (2008) 4462.
M.F.M. Piedade, M.T. Duarte, Chem. Phys. Lett. 367 (2003) 390. [67] N. Sieffert, M. Buhl, Inorg. Chem. 48 (2009) 4622.
[22] M.H. Garcia, Paulo J. Mendes, A. Romão Dias, J. Organomet Chem. 690 (2005) [68] Luca Salassa, Claudio Garino, Giovanni Salassa, Carlo Nervi, Roberto Gobetto,
4063. Carlo Lamberti, Diego Gianolio, Ranieri Bizzarri, Peter J. Sadler, Inorg. Chem. 48
[23] M.P. Robalo, A. Teixeira, M.H. Garcia, M.F.M. Piedade, M.T. Duarte, A.R. Dias, J. (2009) 1469.
Campo, W. Wenseleers, E. Goovaerts, Eur. J. Inorg. Chem. (2006) 2175. [69] N.M. O’Boyle, A.L. Tenderholt, K.M. Langner, J. Comp. Chem. 29 (2008) 839.
[24] M.H. Garcia, P.J. Mendes, M.P. Robalo, A.R. Dias, J. Campo, W. Wenseleers, E. [70] Simona Fantacci, Filippo De Angelis, Annabella Selloni, J. Am. Chem. Soc. 125
Goovaerts, J. Organomet. Chem. 692 (2007) 3027. (2003) 4381.
[25] M. Helena Garcia, Pedro Florindo, M. Fátima, M. Piedade, M. Teresa Duarte, M. [71] Prakriti Ranjan Bangal, J. Phys. Chem. A 111 (2007) 5536.
Paula Robalo, Etienne Goovaerts, Wim Wenseleers, J. Organomet. Chem. 694 [72] O. Gritsenko, E.J. Baerends, J. Chem. Phys. 121 (2004) 655.
(2009) 433. [73] G.A. Zhurko, D.A. Zhurko, Build 284, 2008.
[26] M. Helena Garcia, Paulo J. Mendes, M. Paula Robalo, M. Teresa Duarte, Nelson [74] Ian R. Whittall, Mark G. Humphrey, David C.R. Hockless, Brian W. Skelton,
Lopes, J. Organomet. Chem. 694 (2009) 2888. Allan H. White, Organometallics 14 (1995) 3970.
[27] Tiago J. L. Silva, ‘‘Synthesis, characterization and prediction of the non-linear [75] S.R. Marder, J.W. Perry, G. Bourhill, C.B. Gorman, B.G. Tiemann, K. Mansour,
optical response of organometallic compounds of Ru(II)”, Msc Thesis, 2008. Science 261 (1993) 186.
[28] Paulo J. Mendes, A.J. Palace Carvalho, J.P. Prates Ramalho, J. Mol. Struct. [76] S.R. Marder, C.B. Gorman, B.G. Tiemann, L.-T. Cheng, J. Am. Chem. Soc. 115
(Theochem) 900 (2009) 110. (1993) 3006.
[29] B.J. Coe, L.A. Jones, J.A. Harrs, B.S. Brunschwig, I. Asselberghs, K. Clays, A. [77] S.R. Marder, L.-T. Cheng, B.G. Tiemann, A.C. Friedli, M. Blanchard-Desce, J.W.
Persoons, J. Am. Chem. Soc. 125 (2003) 862. Perry, J. SkindhØj, Science 263 (1994) 511.
[30] B.J. Coe, L.A. Jones, J.A. Harrs, B.S. Brunschwig, I. Asselberghs, K. Clays, A. [78] Izabela Janowska, Janusz Zakrzewski, Keitaro Nakatani, Marcin Palusiak,
Persoons, J. Garin, J. Orduna, J. Am. Chem. Soc. 126 (2004) 3880. Marcin Walak, Henryk Scholl, J. Organomet. Chem. 691 (2006) 323.
[31] L. Jensen, P.T. Van Duijnen, J.P. Snijders, J. Chem. Phys. 119 (2003) 12998. [79] Chun-Guang Liu, Yong-Qing Qiu, Shi-Ling Sun, Na Li, Guo-Chun Yang, Zhong-
[32] Patrizia Calaminici, Andreas M. Koster, Karl Jug, David Gray, Werner. Blau, J. Min Su, Chem. Phys. Lett. 443 (2007) 163.
Mol. Struct. (Theochem) 762 (2006) 87. [80] L. Onsager, J. Am. Chem. Soc. 58 (1936) 1486.
[33] P.L. Franzen, S.C. Zilio, A.E.H. Machado, J.M. Madurro, A.G. Brito-Madurro, L.T. [81] K. Clays, A. Persoons, Phys. Rev. Lett. 66 (1991) 2980.
Ueno, R.N. Sampaio, N.M. Barbosa Neto, J. Mol. Struct. 892 (2008) 254. [82] Hochan Lee, Sun-Young An, Minhaeng Cho, J. Phys. Chem. B 103 (1999) 4992.