www.elsevier.nl/locate/ica Inorganica Chimica Acta 300 – 302 (2000) 698 – 708
Electronic spectra of trans-[Ru(NH3)4(L)NO]3 + /2 + complexes Sergey I. Gorelsky a, Sebastia˜o C. da Silva b, A.B.P. Lever a,*, Douglas W. Franco b b
a Department of Chemistry, York Uni6ersity, Toronto, Ont., Canada M3J 1P3 Uni6ersidade de Sa˜o Paulo-USP, Instituto de Quı´mica de Sa˜o Carlos, Caixa Postal 780, 13560 -970 Sa˜o Carlos-SP, Brazil
Received 7 October 1999; accepted 29 November 1999
Abstract Density functional theory (DFT) with local, non-local and hybrid functionals has been used to obtain the geometry of a series of nitrosyl–metal complexes [Ru(NH3)4(L)NO]n + , where L = NH3, H2O, pyrazine and pyridine (n=3), Cl− and OH− (n =2). Based on the molecular orbital analysis and the time dependent DFT (TD-DFT) calculations, we discuss the electronic structure and the assignment of the bands in the electronic spectra of these complexes. © 2000 Elsevier Science S.A. All rights reserved. Keywords: Electronic spectra; Ruthenium complexes; Nitrosyl – metal complexes; DFT; TD-DFT
1. Introduction
2. Computational details
Studies of electronic structure and electronic spectroscopy of nitrosyl complexes of transition metals are dominated by investigations of the nitroprusside ion [Fe(CN)5NO]2 − , initiated by the pioneering work of Manoharan and Gray [1]. Hollauer and Olabe, and Westcott and Enemark have reviewed the progress in this field [2]. Ruthenium nitrosyl complexes, however, have not been the subject of as many studies as pentacyanonitrosometallates in spite of the fact that they are better suited for quantum chemical calculations. These nitrosyl complexes are amenable to simpler and more accurate calculations than the pentacyanonitrosometallates because they are positively charged. We reconsider the assignments of the electronic spectra of these [Ru(NH3)4(L)NO]n + species and discuss their electronic structure with respect to variation of ligand L. We are especially interested in how such variation of L influences the RuNO bonding interaction.
It was not clear, a priori, what is the best procedure to study the geometry and electronic structure of these complexes and therefore we tried several methods. We used the semi-empirical ZINDO method and the linear combination of Gaussian type orbitals-density functional (LCGTO-DF) method [3] with the following exchange-correlation functionals SVWN5 (functional 5 in [4]) and PVS1 [3], the GGA exchange and correlation functional of Perdew and Wang (PD86) [5], the GGA exchange functional of Becke [6], and nonlocal generalization of the correlation functional (LAP3). This is a gradient corrected functional that combines Becke’s exchange [6] to the kinetic energy density and Laplacian dependent correlation functional ‘LAP3’ of the LAP family developed by Proynov [3b,f,g,h] and employing the deMon-KS3p2 package [3c]. The split valence double-zeta orbital basis set DZVP [7]1 was used for all
* Corresponding author. Tel.: + 1-416-736 5246; fax: + 1-416-736 5936. E-mail address:
[email protected] (A.B.P. Lever)
1
Basis sets were obtained from the Extensible Computational Chemistry Environment Basis Set Database, Version 1.0, as developed and distributed by the Molecular Science Computing Facility, Environmental and Molecular Sciences Laboratory, which is part of the Pacific Northwest Laboratory, P.O. Box 999, Richland, Washington 99352, USA, and funded by the US Department of Energy. The Pacific Northwest Laboratory is a multi-program laboratory operated by Battelle Memorial Institute for the US Department of Energy under contract DE-AC06-76RLO 1830. Contact David Feller or Karen Schuchardt for further information.
0020-1693/00/$ - see front matter © 2000 Elsevier Science S.A. All rights reserved. PII: S 0 0 2 0 - 1 6 9 3 ( 9 9 ) 0 0 6 1 1 - 8
Table 1 Geometry of [Ru(NH3)5NO]3+ (C1) by X-ray
ZINDO/1
ZINDO/1
and DF methods and the comparison with X-ray data [19a] a
SVWN
SVWN5
PVS1
b
PD86 b
BLAP3 b
B3LYP
LanL2DZ
CEP-121G
DZVP
CEP-121G
DZVP
DZVP
DZVP
DZVP
3-21G
LanL2DZ
SDD
CEP-121G
DZVP
DZVP/6-31G
2.093(9)–2.100(8) 2.02(1) c 1.770(9) 1.172(14)
2.075 2.055 1.840 1.141
2.133–2.135 2.116 1.773 1.176
2.147–2.150 2.136 1.785 1.194
2.143–2.144 2.136 1.779 1.141
2.151–2.154 2.139 1.787 1.195
2.147–2.148 2.140 1.781 1.142
2.136–2.141 2.136 1.763 1.146
2.256–2.263 2.241 1.821 1.174
2.258–2.266 2.253 1.807 1.165
2.200–2.201 2.181 1.811 1.153
2.198 2.175 1.803 1.166
2.189–2.190 2.179 1.781 1.169
2.210–2.211 2.194 1.809 1.178
2.210 2.194 1.819 1.129
2.209–2.211 2.187 1.822 1.127
RuNO
172.8(9)
180
179
179
179
179
179
177
177
177
178
179
179
179
179
179
a b c
Basis set dimensions: 132 basis functions/237 primitive gaussians (3-21G); 115/281 (LanL2DZ); 139/293 (SDD); 163/249 (CEP-121G); 164/393 (DZVP); 183/367 (DZVP on Ru, 6-31++G(d,p) on other atoms). deMon-KS calculations, the rest are GAUSSIAN-98 calculations. Not reliable (see comments on p.1604 [19a]).
Table 2 Geometry of [Ru(NH3)4(OH)NO]2+ (Cs ) by X-ray
ZINDO/1
and DF methods and the comparison with X-ray data [19a] a
ZINDO/1
SVWN
PD86 b
BLAP3 b
B3LYP
DZVP
DZVP
DZVP
DZVP
3-21G
LanL2DZ
SDD
CEP-121G
DZVP
DZVP/6-31G++
2.115 2.126 1.933 1.783 1.154
2.129
2.242
2.247
1.853 1.758 1.189
1.915 1.809 1.225
1.906 1.795 1.217
2.163 2.182 1.936 1.820 1.171
2.161 2.178 1.946 1.799 1.183
2.155 2.170 1.944 1.777 1.186
2.176 2.190 1.959 1.803 1.195
2.179 2.193 1.960 1.817 1.144
2.174 2.192 1.961 1.814 1.144
177 124
180 180
180 180
180 180
176 123
176 131
177 132
177 130
177 122
176 126
2.099(3)–2.106(3)
2.068–2.075
RuO(H) RuN(O) NO
1.961(3) 1.735(3) 1.159(5)
1.870 1.807 1.154
2.100 2.113 1.922 1.771 1.190
RuNO RuOH
173.8(3)
179 117
177 134
b
b
LanL2DZ RuN(H3)
a
PVS1
S.I. Gorelsky et al. / Inorganica Chimica Acta 300–302 (2000) 698–708
RuN(H3)eq R-N(H3)ax RuN(O) NO
Basis set dimensions: 128 basis functions/231 primitive gaussians (3-21G); 111/273 (LanL2DZ); 135/285 (SDD); 157/239 (CEP-121G); 160/383 (DZVP); 170/364 (DZVP on Ru, 6-31++G(d,p) on other atoms). deMon-KS calculations on C46 structure with the linear RuOH group, the rest are GAUSSIAN-98 calculations.
699
S.I. Gorelsky et al. / Inorganica Chimica Acta 300–302 (2000) 698–708
700
Table 3 Geometry of [Ru(NH3)4(Cl)NO]2+ (C46) by X-Ray
RuN(H3)
ZINDO/1
ZINDO/1
and DF methods and the comparison with X-ray data [19c] BLAP3 a
B3LYP
DZVP
DZVP
3-21G
LanL2DZ
CEP-121G
DZVP
SVWN
SVWN5
PVS1
DZVP
DZVP
a
2.062
2.129
2.133
2.138
2.265
2.179
2.168
2.187
2.194
RuCl RuN(O) NO
2.101(8) 2.109(7) 2.355(3) 1.79(1) 1.08 b
2.288 1.796 1.150
2.288 1.784 1.152
2.292 1.786 1.153
2.216 1.780 1.177
2.288 1.832 1.201
2.353 1.805 1.167
2.383 1.795 1.179
2.375 1.803 1.191
2.330 1.819 1.142
RuNO
174.9(3)
180
180
180
180
180
180
180
180
180
a b
deMon-KS calculations, the rest are GAUSSIAN-98 calculations. Not reliable due to disorder problems [19c]
deMon calculations. The auxiliary basis sets (fit functions), required by the LCGTO-DF model to fit the electron density and the exchange-correlation potential, were chosen as Ru(5,5;5,5), C(4,1,2;4,1,2), N(4,1,2;4,1,2) and H(3,1;3,1) [3d,e] with the SCF convergence tolerance 10 − 7 Hartrees, electronic density convergence tolerance 1× 10 − 7, and geometry optimization tolerance 1× 10 − 5 Hartrees/Bohr (5×10 − 5 Hartrees/Bohr for calculations with the BLAP3 potential). We also used local spin density exchange-correlation functionals SVWN (functional III in [4]) and SVWN5 (functional V in [4]), and Becke’s three parameter hybrid functional [8] with LYP correlation functional [9] (B3LYP) within the GAUSSIAN-98 program suite [10] employing an effective core potential (ECP) triple-split basis set (CEP-121G) [11], 3-21G [12], LanL2DZ ECP [13], Stuttgart/Dresden ECP (SDD) [14], DZVP [7] and 6-31+ + G(d,p) basis sets [15]. As a starting geometry, we used the X-ray data, when available, or an optimized geometry at the semiempirical level, obtained by the ZINDO/1 method [16] with b(d)= −20 eV for Ru [16c]. The stability of the DFT wavefunctions was tested with respect to relaxing spin and symmetry constraints. The harmonic frequency calculations verifying the true nature of the optimized minima were carried out as well. The MO energies were calculated with ZINDO/S [16], utilizing the Ru and Cl INDO/S parameter obtained from [17]2. The overlap weighting factors s – s and p–p were set at 1.265 and 0.585, respectively [16]. In addition to ZINDO/S, TD-DFT [18] calculations were conducted. For TD-DFT calculations the energies and intensities of the lowest 30 – 40 singlet – singlet transitions were calculated. The TD-DFT method is gaining prominence as an accurate procedure to derive excited state transition energies [18,21].
2 Is(cl)= 25.23 eV, Ip = 15.03 eV, bs = bp = − 10.5 eV, zs = zp = 2.033, F 2(p,p)=5.21 eV, G 1(s,p) =2.66 eV, gss = 11.41 eV (taken from Ref. [17b]).
3. Results and discussion
3.1. Geometry optimization The results from the optimization of the ions [Ru(NH3)4(L)NO]n + using ZINDO/1 and various DFT/ basis set combinations are listed in Tables 1–6 and for comparison, the X-ray results are reported for the related structures [19]. For complexes with pyridine and pyrazine ligands, there are no experimental X-ray data but their structures are closely related to the structure of a similar complex with the nicotinamide ligand, [Ru(NH3)4(nic)NO]3 + [19b], so that we can compare the bond distances in the nicotinamide complex with the corresponding distances in the pyridine and pyrazine complexes. All complexes show a pseudo-octahedral arrangement of ligands around the metal atom. The RuNO angle is close or equal to 180°, thus confirming the presence of the RuIINO+ moiety in the complexes studied. The ZINDO/1 method is the simplest method and we used it initially to optimize the geometry of the complexes, however the quality of the results is not very impressive. The equatorial RuN(H3) and RuL distances are underestimated, and RuN(O) distances, on the contrary, are significantly overestimated. However the DFT procedures, in general, gave optimized structures where both the RuNH3 and RuNO distances agreed much more closely with the experimental data. From comparison of the calculated and the experimental geometry it is noticeable that nonlocal and mixed functionals somewhat overestimate the RuNH3 distances while the gradientless SVWN, SVWN5 and PVS1 functionals give RuNH3 distances very close to the X-ray results. This is the usual difference between the LSDA and GGA results [20]. For other atomic distances the popular mixed B3LYP functional and SVWN functional give results close to the experimental data. Based on our observations as well as on a general agreement in the literature that the B3LYP
2.178 1.794 1.129 180
2.199
2.141 1.761 1.169 180
2.155 1.790 1.177 180
2.196 2.206 2.171 1.794 1.129 180 2.201 2.181
2.177 2.186 2.136 1.761 1.169 180 2.115 1.792 1.152 180
2.143 1.780 1.165 180
2.185 2.194 2.138 1.780 1.166 180 2.189 2.192
calculations. GAUSSIAN-98
deMon-KS calculations, the rest are
2.022 1.835 1.142 180 2.035(5) 1.715(5) 1.142(7) 178.1(5) RuO(H2) RuN(O) NO RuNO
2.069
2.069 2.070 2.022 1.835 1.142 180 2.093(5)–2.107(5) RuN(H3)
a
2.106 1.756 1.141 180
2.100 1.733 1.150 180
2.241 1.773 1.171 180
2.187 2.198 2.111 1.792 1.153 180 2.258 2.140 2.137
2.133 2.142 2.101 1.756 1.141 180
b a b b a b a b a
b a
b b a
b
SDD LanL2DZ LanL2DZ 3-21G 3-21G DZVP DZVP DZVP
DZVP
B3LYP BLAP3 a a
PVS1 SVWN ZINDO/1
X-ray
Table 4 Geometry of [Ru(NH3)4(H2O)NO]3+ (configurations (a) and (b) (Fig. 1)) by
ZINDO/1
and DF methods the comparison with X-ray data [19c]
SDD
CEP-121G
DZVP
DZVP
S.I. Gorelsky et al. / Inorganica Chimica Acta 300–302 (2000) 698–708
701
functional usually gives the best results both for geometry optimization and electronic spectra [21], we opted to use it as the primary method in our studies. The structure of [Ru(NH3)4(Cl)NO]2 + was optimized with C46 symmetry. [Ru(NH3)4(OH)NO]2 + was also initially optimized with C46 symmetry but it turned out that the structure with the linear RuOH group does not correspond to true energy minimum. The Cs structure with bent RuOH group does. But apart from that the bond distances in C46 and Cs structures of [Ru(NH3)4(OH)NO]2 + are quite similar. There are two limiting conformational possibilities for the [Ru(NH3)4(H2O)NO]3 + , [Ru(NH3)4(py)NO]3 + and [Ru(NH3)4(pyz)NO]3 + complexes; the ligand, L, can lie on the same plane with Ru and the two nitrogen atoms of the equatorial NH3 ligands (Fig. 1(a)), or its plane can bisect the planes passing through the equatorial NH3 ligands (Fig. 1(b)). Both orientations were considered. The DFT optimization revealed that the lowest energy structures of [Ru(NH3)4(L)NO]3 + (L=py, pyz) have configuration (b), in agreement with the X-ray data of the similar complex, [Ru(NH3)4(nic)NO]3 + [19b]. For the complex with L= H2O, however, configuration (a) corresponds to the true energy minimum but the electronic energy difference between configurations (a) and (b) is very small (0.0054 eV (B3LYP/DZVP), 0.015 eV (B3LYP/SDD)) and comparable with the kT energy. At room temperature free rotation of the H2O ligand is expected. Thus, the experimental spectroscopic data correspond to the time-average of the spectra of the various conformers. The Onsager reaction field model [22] with a spherical cavity in a continuum with a dielectric constant of o= 78.5, corresponding to water at 298 K, was also used for modeling the interactions between the solute and the solvent. The results are presented in Table 6. There are no dramatic changes in the geometry of the complex, so that the vacuum geometry can be used in the calculation to compare the results with the experimental condensed phase data. Similar conclusions were made in [23].
3.2. Molecular orbital analysis Neglecting the effect of the hydrogen atoms (and considering a rotation of NH3 and H2O ligands), all [Ru(NH3)5NO]3 + , [Ru(NH3)4(Cl)NO]2 + , [Ru(NH3)4(H2O)NO]2 + and [Ru(NH3)4(OH)NO]2 + ions can be classified in the point group C46. The LRuNO vector is defined as z, while x and y bisect the two H3NRuNH3 vectors. The structure of the MNO bond is composed of a s-bond, using the nitrogen lone pair and two p-interactions involving the filled dxz and dyz orbital on ruthenium atom and the p*-orbital of NO (LUMO and LUMO + 1 in these complexes). The dz 2 and dxy orbitals are p* with respect to the MN5L
S.I. Gorelsky et al. / Inorganica Chimica Acta 300–302 (2000) 698–708
702
Table 5 Geometry of [Ru(NH3)4(py)NO]3+ (C26 ) by X-ray (L =nic)
ZINDO/1
ZINDO/1
and DF methods and the comparison with X-ray data [19b]
SVWN
PVS1 a
B3LYP
DZVP
DZVP
3-21G
LanL2DZ
SDD
CEP-121G
DZVP
RuN(H3) RuN(py) RuN(O) NO
2.06(2)–2.16(2) 2.14(2) 1.71(2) 1.17(3)
2.073 1.993 1.832 1.144
2.137 2.085 1.790 1.143
2.145 2.094 1.770 1.149
2.194 2.128 1.819 1.556
2.192 2.119 1.808 1.169
2.183 2.126 1.787 1.173
2.204 2.141 1.814 1.181
2.205 2.145 1.826 1.132
RuNO
177(1)
180
180
180
180
180
180
180
180
a
deMon-KS calculations, the rest are
GAUSSIAN-98
Table 6 Geometry of [Ru(NH3)4(pyz)NO]3+ (C26 ) by X-ray (L = nic)
ZINDO/1
ZINDO/1
calculations.
and DF methods and the comparison with X-ray data [19b]
SVWN
PVS1 a
B3LYP
DZVP
DZVP
3-21G
LanL2DZ
LanL2DZ +SCRF b
SDD
CEP-121G
DZVP
RuN(H3) RuN(pyz) RuN(O) NO
2.06(2)2.16(2) 2.14(2) 1.71(2) 1.17(3)
2.073 1.995 1.834 1.143
2.140 2.101 1.799 1.142
2.143 2.123 1.768 1.149
2.196 2.146 1.818 1.154
2.195 2.139 1.807 1.168
2.196 2.160 1.811 1.162
2.185 2.144 1.786 1.171
2.206 2.162 1.813 1.179
2.206 2.165 1.825 1.131
RuNO
177(1)
180
180
180
180
180
180
180
180
180
a
deMon-KS calculations, the rest are GAUSSIAN-98 calculations. DFT-SCRF calculation: the Onsager reaction field model [22] with a spherical cavity in a continuum with a dielectric constant of o = 78.5, corresponding to water at 298K. b
chromophore while the dx 2 − y 2 orbital, localized in the equatorial plane, has very little p-interaction with the NH3 ligands and remains non-bonding (Table 7). While b1(dx 2 − y 2) is the HOMO for the NH3 complex, e(dxz,dyz ) is the HOMO for the OH− and Cl− complexes. The DFT and ZINDO/S calculations give the same orbital ordering. The negatively charged ligands L (OH−, Cl−) ‘push’ the e(dxz,dyz ) orbital up in energy through interaction with ligand e(px,py ). For the [Ru(NH3)4(H2O)NO]3 + ion with C26 symmetry, the HOMO is again the nonbonding dx 2 − y 2 orbital, similar to the NH3 complex but now there is also a small (0.1 eV (ZINDO/S), 0.2 – 0.3 eV (B3LYP with different basis sets)) splitting between the dxz and dyz orbitals of Ru owing to the fact that the water molecule interacts in a slightly different fashion with these orbitals. On a time average basis, however, in C46 symmetry, they remain degenerate. Both [Ru(NH3)4(py)NO]3 + and [Ru(NH3)4(pyz)NO]3 + complexes have similar electronic structures and are defined with the pyridine and pyrazine rings in the yz plane of the C26 molecule. Unlike the previously described species, HOMO and HOMO-1 are localized on the heterocyclic ligand. Their energies are 0.5–2.0 eV above the set of d-orbitals (Table 7 and Fig. 2). So, in addition to d–d (Ru) and MLCT (Ru NO) bands in the electronic spectra of these complexes, we can also
Fig. 1. The structures of trans-[Ru(NH3)4(H2O)(NO)]3 + (C26 symmetry): (a) configuration a. (b) configuration b.
S.I. Gorelsky et al. / Inorganica Chimica Acta 300–302 (2000) 698–708
703
Fig. 2. B3LYP/LanL2DZ Kohn–Sham orbitals for trans-[Ru(NH3)4(L)(NO)]3 + , L=NH3, py, pyz.
expect MLCT (RuL), LMCT (L Ru) and LLCT (L NO) transitions. For the neutral ligands considered here, the percentage Ru content in the LUMO primarily localized on the nitrosyl ligand is about 28% essentially independent of L (Table 7). However with the negative ligands hydroxide and chloride, this percentage increases up to 33% since the ruthenium becomes much more electron-rich. The non-bonding dx 2 − y 2 orbitals remain approximately 94 – 95% pure 4d Ru, while the purity of the dxz,yz varies quite a lot with L. Thus for the neutral ligands dxz,yz (when not coupled to pyridine or pyrazine) is 55 –59% pure 4d Ru through mixing with p*(NO) which comprises 37–40% in this orbital (Table 7). In the L= py, pyz species, the dxz couples with both p*(NO) and p* py or pyz. Coupling to p* pyrazine is particularly pronounced because of the greater acceptor character of that ligand compared with pyridine. This coupling causes dxz to be 40– 47% Ru and 28 – 29% p*(NO). For L = OH− the p*(NO) content remains around 35% but the 4d content for one of the dp orbitals drops to some 34% due to mixing with pp of hydroxide.
3.3. Electronic spectra In the earlier work of Schreiner et al. [24a] the electronic spectra of [Ru(NH3)4(L)NO]n + (L = NH3,
− OAc−, Cl−, OH−, N− 3 and Br ) were discussed in some detail. Recently Franco with co-workers [19b,24b] studied ruthenium nitrosyl complexes with N-heterocyclic ligands (L = imidazole, pyridine, pyrazine, nicotinamide) and proposed a tentative band assignment for the reported complexes taking Schreiner’s assignment scheme. In another contribution [19c] they also studied the aqua complex [Ru(NH3)4(H2O)NO]3 + and proposed the assignment of its electronic spectrum based on a ZINDO/S calculation. According to the experimental data in the UV–Vis region there are two major bands for [Ru(NH3)4(L)NO]3 + /2 + . The first one is a broad band in the 420–480 nm (2.6–2.9 eV) region. Schreiner et al. [24a] have assigned it to a singlet–triplet transitions 1 A1 3T1,3T2 and MLCT t2 p*(NO) (the notation from their work). It was also noted that the t2 p*(NO) charge transfer band has surprisingly low intensity. The proposed assignment of the second band at the 300–350 nm (3.5–4.1 eV) region was d–d 1A1 1A1, 1E. For some trans-ligands L there is a third electronic absorption band which is located above 5 eV. It was assigned to px,py (L) dz 2,dxy (Ru) LMCT. For complexes with imidazole, pyridine, nicotinamide and pyrazine ligands there is also an additional low-energy and lowintensity band in the 500–600 nm (2.0–2.5 eV) region
704
Table 7 Frontier molecular orbitals (orbital energies and atom contributions, %) of [Ru(NH3)4(L)NO]3+ by
calculations on SVWN/DZVP optimized structures
L= OH− L= Cl−
L= NH3
L =H2O (configuration b)
−o (eV)
Ru
NO
L
G
−o (eV)
Ru
NO
L
−o (eV)
Ru
NO
L
G
−o (eV)
Ru
NO
L
a1(dz 2) b2(dxy ) e(p*NO)
11.4 11.9 13.6
59 61 28
13 0 70
12 0 0 0 1
15 0 1 2 1 29 8
19 0 1
0 38
8 0 68 63 4 35 35
11 0 66
95 59
55 61 30 33 90 34 55
58 61 31
21.8 22.2
6.5 7.7 9.4 9.4 17.8 17.4 17.8
7.9 8.0 9.9
b1(dx 2−y 2) e(dxz,dyz )
a% a%% a% a%% a% a%% a%
18.2 18.1
95 35
0 29
0 36
a1 a2 b2 b1 a1 b1 b2
11.8 11.9 13.5 13.6 21.9 22.3 22.4
61 60 29 30 94 55 57
14 0 70 69 0 40 40
11 0 0 0 0 2.8 0
Orbital
L =py
LUMO+4 LUMO+3 LUMO+2 LUMO+1 LUMO HOMO HOMO-1 HOMO-2 HOMO-3 HOMO-4
L =pyz
G
−o (eV)
Ru
NO
L
G
−o (eV)
Ru
NO
L
a1(dz 2) b1(p*L) a2(dxy ) b2(p*NO) b1(p*NO) a2(pL) b1(pL) a1(dx 2−y 2) b2(dyz ) b1(dxz )
10.4 10.6 11.0 12.8 12.8 19.4 20.2 20.9 21.3 21.5
56 2 61 28 28 0 10 94 59 47
11 1 0 70 67 0 10 0 37 29
15 98 0 0 3 100 79 0 2 22
a1(dz 2) b1(p*L) a2(dxy ) b2(p*NO) b1(p*NO) a2(pL) a1(sL) a1(dx 2−y 2) b1(dxz ) b2(dyz )
10.7 11.2 11.2 13.0 13.1 19.4 20.2 21.2 21.2 21.5
57 2 61 28 27 0 1 95 40 59
11 1 0 70 68 0 0 0 28 37
14 96 0 0 4 100 99 0 30 1
S.I. Gorelsky et al. / Inorganica Chimica Acta 300–302 (2000) 698–708
Orbital (C46 group)
ZINDO/S
S.I. Gorelsky et al. / Inorganica Chimica Acta 300–302 (2000) 698–708
[19b]. This band has been observed only in complexes where the trans-ligand is an aromatic ring. Initially we conducted semiempirical ZINDO/S calculations to predict the electronic spectra, but the agreement between calculated and experimental spectra was not very satisfactory. For the complexes with L= py, pyz, ZINDO/S was unable to predict the low-energy band in the electronic spectrum. On the other hand, TD-DFT (B3LYP) calculations appear to give very good predictions of the energies of low-lying excited states of clear valence type [21]. Table 8 shows the major results from the calculated electronic transitions and the experimental absorption bands for [Ru(NH3)4(L)NO]3 + /2 + . According to TD-DFT calculations the lowest energy spin singlet excited states of [Ru(NH3)4(L)NO]3 + with L=NH3, H2O correspond to the electron excitation HOMO(dx 2 − y 2Ru) LUMO +0,1(p*NO) (b1 e in C46, a1 b1,b2 in C26 ). Lying higher in energy are three of the possible four excited MLCT states originating from the electron excitation HOMO-1,2(dxz,dyz ) LUMO+ 0,1(p*NO) transition. This result is in disagreement with the assignment proposed in [19c], which is, however, based upon a ZINDO/S calculation where the energy of the HOMO-1,2 LUMO +0,1 transition is lower than HOMO LUMO +0,1. The energy difference between these transitions is not large although the difference in calculated intensity is significant (Table 8): the orbitally forbidden components of dxz,dyz (Ru) p*(NO) transitions (vide infra) have oscillator strengths fB 10 − 4 while the oscillator strengths of dx 2 − y 2(Ru) p*(NO) lie within the range f= 1 – 31 ×10 − 4. In the C46 point group the electronic transitions dxz,dyz (Ru)p*(NO) from the ground state 1A1 to the excited states 1A2, 1B1, and 1B2, are orbitally forbidden and only the transition to the excited state 1A1, is allowed. However the energy of this last electronic transition is calculated to be in the far UV region. There is no experimental data for these complexes in the far UV spectra, and this intense transition is absent from the regular UV spectra ( \220 nm) of these [Ru(NH3)4(L)NO]3 + /2 + species. Similarly ZINDO/S calculation predicts three low-lying and weak forbidden transitions and one high energy and intense allowed transition. For [Ru(NH3)4(L)NO]2 + with L =Cl− and OH−, the three states from the e(dxz,dyz ) e(p*NO) excitation are slightly lower in energy than the dx 2 − y 2(Ru) p*(NO) states because of different orbital order in complexes with neutrally and negatively charged ligands, L, but still remain low in intensity. The intense allowed transition is shifted to lower energy and should occur in the UV region (around 4.6 eV for [Ru(NH3)4(OH)NO]2 + , and around 3.7 eV for [Ru(NH3)4(Cl)NO]2 + ).
705
When the symmetry of the complex is lowered from C46 to C26, two of four transitions originating from the e(dxz,dyz ) e(p*NO) excitation remain orbitally forbidden (1A2 states). One of three forbidden (within C46 symmetry) transitions becomes allowed and increases in intensity. The calculated spectrum of the [Ru(NH3)4(H2O)NO]3 + complex depends very weakly on the orientation of the H2O ligand (Fig. 1) (the difference in optical transition energies for configuration (a) and (b) is around 0.03 eV for all bands in the UV–Vis region). This is a consequence of a small contribution that the H2O orbitals make to the frontier orbitals of the complex (Table 7). Thus, our calculations are in good agreement with the experimental data, which show that the MLCT d(Ru) p*(NO) band for these complexes is very weak. We do not expect any significant contribution from singlet–triplet transitions to the electronic spectrum since their intensity should be several orders of magnitude smaller than the intensity of the corresponding singlet–singlet transitions. There is a broad absorption band around 3.75–4 eV in all the complexes. It was assigned by Schreiner et al. [24a] and Franco [19c] to be a d–d transition. DFT calculations are mostly in agreement with this assignment (Table 8) with an additional contributor; some components of the MLCT d(Ru) p*(NO) transition are also located in this frequency region. For [Ru(NH3)4(py)NO]3 + and [Ru(NH3)4(pyz)NO]3 + DFT(B3LYP) calculations predict that this band is mostly due to MLCT d(Ru) p*(NO). For [Ru(NH3)4(py)NO]3 + the LMCT p(py) d(Ru) charge transfer band is also located in the 4 eV region with approximately the same intensity as MLCT d(Ru)p*(NO). For the corresponding pyrazine complex there should be an LMCT s(pyz) d(Ru) transition from a ‘lone pair’ on a nitrogen atom of the pyrazine ligand to ruthenium. The calculated energy of this transition is about 3 eV. In the experimental spectrum there is a band at 2.67 eV but its intensity is not as large as calculated. For the [Ru(NH3)4(py)NO]3 + and [Ru(NH3)4(pyz)NO]3 + complexes, DFT calculations give a number of intense internal ligand–ligand (L) transitions in accordance with the experimental data (4.6 eV with o= 2300 for the complex with L= py; 4.5 eV with o =4400 l mol − 1 cm − 1 for the complex with L=pyz). In the visible region, the spectrum of [Ru(NH3)4(py)NO]3 + shows a broad band. This band can be deconvoluted into two peaks, at 2.43 and 2.67 eV. The band at 2.67 eV is assigned to d(Ru) p*(NO) charge transfer, as in the systems discussed before, and to one of the components of p(py) p*(NO). This can explain why the experimental intensity of this band is almost twice as high as the intensity of the corresponding band in the other complexes. The band at 2.43 eV in
706
S.I. Gorelsky et al. / Inorganica Chimica Acta 300–302 (2000) 698–708
Table 8 Calculated (TD-DFT) and experimental electronic spectral data a for [Ru(NH3)4(L)NO]n+ (in parenthesis: electric dipole oscillator strengths b f×104, for calculated spectra; oscillator strengths, f×104 (where available), and molar absorption coefficients (o, l mol−1 cm−1) for experiment) Electronic transition c (principal excitation)
L=NH3 HOMO LUMO HOMO LUMO+1 HOMO-1,2 LUMO+0,1
G
TD-DFT(B3LYP) excitation energy (intensity) (eV) 3-21G
LanL2DZ
SDD
CEP-121G
DZVP
Experiment 2.77(o= 15)
dd(Ru) p*(NO) dd(Ru) p*(NO) dp(Ru) p*(NO) dp(Ru) p*(NO) dp(Ru) p*(NO) dd(Ru) ds(Ru) dp(Ru) ds(Ru) dp(Ru) ds(Ru)
1
A A 1 A 1 A 1 A 1 A 1 A 1 A
2.69 2.70 3.13 3.40 3.59 3.46 4.64 4.65
2.75(1) 2.76(1) 3.16(0) 3.44(0) 3.59(0) 3.58(0) 4.79(6) 4.80(13)
2.82 2.82 3.36 3.55 3.71 3.61 4.93 4.94
2.72 2.73 3.12 3.34 3.49 3.46 4.65 4.66
2.99(2) 3.00(2) 3.46(0) 3.58(0) 3.82(0) 3.36(0) 4.58(6) 4.59(13)
dp(Ru) p*(NO) dp(Ru) p*(NO) dp(Ru) p*(NO) dp(Ru) p*(NO) dd(Ru) p*(NO) dd(Ru) p*(NO) dp(Ru) ds(Ru) dp(Ru) ds(Ru)
1
A%% A% 1 A%% 1 A% 1 A% 1 A%% 1 A% 1 A%%
2.37(0) 2.93(38) 3.20(0) 4.36(467) 3.09(2) 3.12(0) 3.90(47) 4.23(54)
2.48(0) 2.95(26) 3.14(0) 4.38(588) 3.16(2) 3.20(1) 4.07(56) 4.32(50)
2.67(0) 3.14(24) 3.32(0) 4.61(512) 3.22(2) 3.25(1) 4.24(62) 4.53(61)
2.53(0) 2.95(26) 3.13(0) 4.39(466) 3.12(2) 3.17(1) 4.06(60) 4.30(56)
2.66(0) 3.17(61) 3.49(0) 4.47(388) 3.34(5) 3.37(1) 3.76(40) 4.20(46)
dp(Ru) p*(NO) dp(Ru) p*(NO) dp(Ru) p*(NO) dp(Ru) p*(NO) dd(Ru) p*(NO) dp(Ru) ds(Ru) dp(Ru) ds(Ru)
1
A2 B1 1 B2 1 A1 1 E 1 E 1 E
2.63 2.84 2.87 3.60 2.88 4.02 4.35
2.60(0) 2.80(0) 2.81(0) 3.51(436) 2.92(1) 4.04(33) 4.42(30)
2.78 2.99 3.02 3.77 2.97 4.28 4.58
2.62 2.80 2.82 3.62 2.90 4.05 4.33
2.73(0) 2.92(0) 2.94(0) 3.74(568) 3.15(1) 4.03(8) 3.95(59)
1
B2 B1 1 A2 1 A1 1 A2 1 A1 1 B1 1 B2
2.81 2.82 3.19 3.38 3.71 4.11 4.34 4.51
2.86(1) 2.87(1) 3.26(0) 3.44(3) 3.74(0) 4.20(6) 4.45(21) 4.60(11)
2.91 2.93 3.36 3.54 3.85 4.28 4.60 4.73
2.82 2.83 3.21 3.32 3.63 4.07 4.32 4.44
3.12(2) 3.13(1) 3.50(0) 3.58(0) 4.00(0) 4.19(9) 4.23(20) 4.39(12)
2.91(o=45)
HOMO LUMO+2 HOMO-1 LUMO+2 HOMO-2 LUMO+2
dd(Ru) p*(NO) dd(Ru) p*(NO) dp(Ru) p*(NO) dp(Ru) p*(NO) dp(Ru) p*(NO) dd(Ru) ds(Ru) dp(Ru) ds(Ru) dp(Ru) ds(Ru)
L = py HOMO LUMO HOMO LUMO+1 HOMO-1 LUMO HOMO-1 LUMO+1 HOMO-2 LUMO HOMO-2 LUMO+1 HOMO-5 LUMO HOMO-1LUMO+2 HOMO-1LUMO+3
p(py) p*(NO) p(py) p*(NO) p(py) p*(NO) p(py)p*(NO) dd(Ru) p*(NO) dd(Ru) p*(NO) dp(Ru) p*(NO) p(py) ds(Ru) p(py) ds(Ru)
1
B2 B1 1 A1 1 A2 1 B1 1 B2 1 A1 1 B2 1 B1
2.55 2.55 3.01 2.73 2.71 2.72 3.57 4.45 4.41
2.39(3) 2.39(0) 2.89(521) 2.62(0) 2.77(1) 2.77(1) 3.57(110) 4.37(21) 4.44(27)
2.63 2.63 3.10 2.82 2.83 2.83 3.70 4.57 4.64
2.36 2.36 2.85 2.59 2.74 2.75 3.49 4.28 4.32
2.37(2) 2.37(0) 2.96(464) 2.75(0) 2.99(1) 3.00(2) 3.70(51) 4.03(16) 4.14(21)
2.43(oB15)
L =pyz HOMO LUMO HOMO LUMO+1 HOMO-1 LUMO HOMO-1 LUMO+1 HOMO-2 LUMO HOMO-2 LUMO+1 HOMOLUMO+2 HOMO LUMO+3 HOMOLUMO+4 HOMO-3,5 LUMO
s(pyz) p*(NO) s(pyz) p*(NO) p(pyz) p*(NO) p(pyz) p*(NO) dd(Ru) p*(NO) dd(Ru) p*(NO) s(pyz)ds(Ru) s(pyz) ds(Ru) s(pyz) p(pyz) dp(Ru) p*(NO)
1
1.22 1.28 2.70 2.71 2.69 2.70 2.91 2.98 3.14 4.14
1.10(0) 1.16(0) 2.53(8) 2.54(0) 2.74(1) 2.75(1) 2.84(0) 2.98(376) 2.99(25) 4.06(495)
1.34 1.41 2.77 2.78 2.80 2.81 3.06 3.20 3.01 4.26
1.23 1.29 2.47 2.49 2.71 2.72 2.91 3.00 3.11 4.00
1.53(0) 1.58(0) 2.43(6) 2.44(0) 2.97(1) 2.98(1) 2.82(0) 3.03(465) 3.39(18) 4.14(377)
HOMO LUMO+3 HOMO-1 LUMO+2 HOMO-2 LUMO+2 L = OH− HOMO LUMO+0,1 HOMO-1 LUMO+0,1 HOMO-2LUMO+0,1 HOMOLUMO+2 HOMOLUMO+3 L = Cl− HOMO-0,1 LUMO+0,1
HOMO-2 LUMO+0,1 HOMO-0,1 LUMO+2 HOMO-0,1 LUMO+3 L = H2O (configuration b) HOMO LUMO HOMO LUMO+1 HOMO-0,1 LUMO+0,1
a
1
1
1
1
1
B1 B2 1 B2 1 B1 1 B1 1 B2 1 A2 1 A1 1 B1 1 A1 1
3.68 d
4.22(15, o = 56) d
2.90(o= 25) 5.44(o=5700) 3.71(45, o = 211)
2.81(o=17) 3.76(40, o = 264) d 2.81(o= 17) 3.76(40, o= 264) d
3.83(o=280)
2.67(o=23)
3.83(o= 160)
2.12(o=40) 2.67(o=43) d
2.67(o=43) d 4.07(o= 660)
Calculated major absorption bands assigned to experimentally observed features, are shown in bold. Intensities predicted by the calculation with other basis sets are reported only for [Ru(NH3)4(OH)NO]2+ but generally agree satisfactorily with those listed in the DZVP and LanL2DZ columns. c Note that the sequence of energy levels derived from ZINDO/S does not always exactly agree with that derived from DFT. Use Fig. 2 for the DFT sequence and Table 7 for the ZINDO/S sequence. d Overlapping absorption bands. b
S.I. Gorelsky et al. / Inorganica Chimica Acta 300–302 (2000) 698–708
707
Acknowledgements We thank Professor R. Fournier for discussion and his assistance with DFT calculations. We also thank the Province of Ontario for an Ontario Graduate Fellowship to SIG.
References
Fig. 3. Comparison between calculated (B3LYP/DZVP) and experimental electron transition energies for trans-[Ru(NH3)4(L)(NO)]3 + / 2+ complexes.
[Ru(NH3)4(py)NO]3 + is also assigned to charge transfer from the pyridine ligand to the nitrosyl ligand. For [Ru(NH3)4(pyz)NO]3 + we observe a similar picture, the only difference being a shift of the first band to lower frequency. Summarizing our TD-DFT results, we can state that our calculations with the B3LYP functional reproduce all features of the experimental electronic spectra. Fig. 3 shows a plot of experimental versus predicted energies illustrating the generally excellent agreement. We may conclude that DFT and TD-DFT are valuable tools for estimating the properties of chemical compounds including transition metal complexes. These methods appear to be more reliable in predictions of geometry and electronic spectra than ZINDO, which is commonly used. We finish by assessing whether the data presented in this paper correlate with other properties of these molecules, specifically with their reduction potentials, localized on the NO fragment and with the EL(L) electrochemical parameter of the ligand L [26,27]. There is an excellent linear relationship (not shown) between the NO stretching frequency and the EL(L) parameter for the data presented here, a fact that had been previously noted [25]. The relationship with reduction potentials is more complex. Considering the neutral ligand species (overall 3+ charge on the complex) there is a rough linear correlation with the LUMO energy but the slope is the inverse of that expected, i.e. the more negative the LUMO energy, the more difficult the reduction; however the spread in LUMO energies for these neutral ligands is probably too small for the result to be statistically meaningful.
[1] (a) P.T. Manoharan, H.B. Gray, J. Am. Chem. Soc. 87 (1965) 3340. (b) P.T. Manoharan, H.B. Gray, Inorg. Chem. 5 (1966) 823. [2] (a) E. Hollauer, J.A. Olabe, J. Braz. Chem. Soc. 8 (1997) 495. (b) D.A. Estrin, L.M. Baraldo, L.D. Slep, B.C. Barja, J.A. Olabe, 35 (1996) 3897. (c) B.L. Westcott, J.H. Enemark, in: E.I. Solomon, A.B.P Lever (Eds.), Inorganic Electronic Structure and Spectroscopy, vol. 2, Wiley, New York, 1999, p. 403. [3] (a) E.I. Proynov, A. Vela, D.R. Salahub, Phys. Rev. A 50 (1994) 3766. (b) E.I. Proynov, E. Ruiz, A. Vela, D.R. Salahub, Int. J. Quantum Chem. Symp. 29 (1995) 61. (c) M.E. Casida, C. Daul, A. Goursot, A. Koester, L. Pettersson, E. Proynov, A. StAmant, D. Salahub, H. Duarte, N. Godbout, J. Guan, C. Jamorski, M. Leboeuf, V. Malkin, O. Malkina, F. Sim, A. Velade, Mon-KS Module Release 3.2, 1997. (d) N. Godbout, D.R. Salahub, J. Andzelm, E. Wimmer, Can. J. Chem. 70 (1992) 560. (e) A. St-Amant, Ph.D. Thesis, Universite´ de Montre´al, 1992. (f) E.I. Proynov, D.R. Salahub, Phys. Rev. B (1994) 7874. (g) E.I. Proynov, D.R. Salahub, Phys. Rev. B (1998) 12616. (h) E.I. Proynov, A. Vela, D.R. Salahub, Chem. Phys. Lett. (1994) 419. (i) E.I. Proynov, A. Vela, D.R. Salahub, Chem. Phys. Lett. (1995) 462. (j) E.I. Proynov, S. Sirois, D.R. Salahub, Int. J. Quantum Chem. (1997) 427. [4] S.H. Vosko, L. Wilk, M. Nusair, Can. J. Phys. 58 (1980) 1200. [5] (a) J.P. Perdew, Phys. Rev. B 33 (1986) 8822. (b) J.P. Perdew, Y. Wang, Phys. Rev. B 33 (1986) 8800 [6] A.D. Becke, Phys. Rev. A 38 (1988) 3098. [7] N. Godbout, D.R. Salahub, J. Andzelm, E. Wimmer, Can. J. Chem. 70 (1992) 560. [8] A.D. Becke, J. Chem. Phys. 98 (1993) 5648. [9] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785. [10] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, V.G. Zakrzewski, J.A. Montgomery, Jr., R.E. Stratmann, J.C. Burant, S. Dapprich, J.M. Millam, A.D. Daniels, K.N. Kudin, M.C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G.A. Petersson, P.Y. Ayala, Q. Cui, K. Morokuma, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J. Cioslowski, J.V. Ortiz, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, J.L. Andres, C. Gonzalez, M. Head-Gordon, E.S. Replogle, J.A. Pople, GAUSSIAN-98, Revision A.7, Gaussian, Inc., Pittsburgh, PA, 1998. [11] (a) W.J. Stevens, H. Basch, J. Krauss, J. Chem. Phys. 81 (1984) 6026. (b) W.J. Stevens, M. Krauss, H. Basch, P.G. Jasien, Can. J. Chem. 70 (1992) 612. (c) T.R. Cundari, W.J. Stevens, J. Chem. Phys. 98 (1993) 5555. [12] (a) J.S. Binkley, J.A. Pople, W.J. Hehre, J. Am. Chem. Soc. 102 (1980) 939 . (b) M.S. Gordon, J.S. Binkley, J.A. Pople, W.J. Pietro, W.J. Hehre, J. Am. Chem. Soc. 104 (1983) 2797. (c) K.D. Dobbs, W.J. Hehre, J. Comput. Chem. 7 (1986) 359 . (d) K.D. Dobbs, W.J. Hehre, J. Comput. Chem. 8 (1987) 861. (e) K.D. Dobbs, W.J. Hehre, J. Comput. Chem. 8 (1987) 880.
708
S.I. Gorelsky et al. / Inorganica Chimica Acta 300–302 (2000) 698–708
[13] (a) T.H. Dunning, Jr., P.J. Hay, in: H.F. Schaefer III (Ed.), Methods of Electronic Structure, Theory, vol. 2, Plenum, New York, 1977. (b) P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 270. (c) P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 284. (d) P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 299. [14] (a) P. Fuentealba, H. Preuss, H. Stoll, L.V. Szentpaly, Chem. Phys. Lett. 89 (1989) 418. (b) L.V. Szentpaly, P. Fuentealba, H. Preuss, H. Stoll, Chem. Phys. Lett. 93 (1982) 555. (c) P. Fuentealba, H. Stoll, L.V. Szentpaly, P. Schwerdtfeger, H. Preuss, J. Phys. B 16 (1983) 1323. (d) H. Stoll, P. Fuentealba, P. Schwerdtfeger, J. Flad, L.V. Szentpaly, H. Preuss, J. Chem. Phys. 81 (1984) 2732. (e) P. Fuentealba, L.V. Szentpaly, H. Preuss, H. Stoll, J. Phys. B 18 (1985) 1287. [15] (a) R. Krishnan, J.S. Binkley, R. Seeger, J.A. Pople, J. Chem. Phys. 72 (1980) 650. (b) J.-P. Blaudeau, M.P. McGrath, L.A. Curtiss, L. Radom, J. Chem. Phys. 107 (1997) 5016. (c) T. Clark, J. Chandrasekhar, P.V.R. Schleyer, J. Comp. Chem. 4 (1983) 294. [16] (a) HYPERCHEM for Windows, Release 5.1 Professional Version, Hypercube, Inc., Gainesville, FL, USA, 1997. (b) M.C. Zerner, ZINDO program, version 98.1, Quantum Theory Project, University of Florida, Gainesville, FL, USA. (c) S.I. Gorelsky, E.S. Dodsworth, A.B.P. Lever, A.A. Vlcek, Coord. Chem. Rev. 174 (1998) 469. [17] (a) K. Krogh-Jespersen, J.D. Westbrook, J.A. Potenza, H.J. Schugar, J. Am. Chem. Soc. 109 (1987) 7025. (b) S.I. Gorelsky, V. Yu. Kotov, Russ. J. Coord. Chem. 24 (1998) 491. [18] (a) M.E. Casida, in: D.P. Chong (Ed.), Recent Advances in Density Functional Methods, v.1, World Scientific, Singapore, 1995. (b) M.E. Casida, in: J.M. Seminario (Ed.), Recent Developments and Applications of Modern Density Functional Theory, Theoretical and Computational Chemistry, v.4, Elsevier, Amsterdam, 1996. (c) M.E. Casida, C. Jamorski, K.C. Casida,
.
[19]
[20]
[21]
[22]
[23] [24]
[25] [26] [27]
D.R. Salahub, J. Chem. Phys. 108 (1998) 4439. (d) R.E. Stratmann, G.E. Scuseria, M.J. Frisch, J. Chem. Phys. 109 (1998) 8218. (a) F. Bottomley, J. Chem. Soc., Dalton Trans. 15 (1974) 1600. (b) S. Da, S.S. Borges, C.U. Davanzo, E.E. Castellano, J.Z. Schpector, S.C. Silva, D.W. Franco, Inorg. Chem. 37 (1998) 2670. (c) C.W.B. Bezerra, S.C. da Silva, M.T.P. Gambardella, R.H.A. Santos, L.M.A. Plicas, E. Tfouni, D.W. Franco, Inorg. Chem. 38 (1999) 5660. (a) C. Daul, H.-U. Gudel, J. Weber, J. Chem. Phys. 98 (1993) 4023. (b) M.R. Bray, R.J. Deeth, V.J. Paget, P.D. Sheen, Int. J. Quantum Chem. 61 (1996) 85. (a) R. Bauernschmitt, R. Ahlrichs, Chem. Phys. Lett. 256 (1996) 454. (b) R. Bauerschmitt, R. Ahlrichs, F.H. Hennrich, M.M. Kappes, J. Am. Chem. Soc. 120 (1998) 5052. (a) M.W. Wong, M.J. Frisch, K.B. Wiberg, J. Am. Chem. Soc. 113 (1991) 4776. (b) M.W. Wong, K.B. Wiberg, M.J. Frisch, J. Am. Chem. Soc. 114 (1992) 523. (c) M.W. Wong, K.B. Wiberg, M.J. Frisch, J. Chem. Phys. 95 (1991) 8991. (d) M.W. Wong, K.B. Wiberg, M.J. Frisch, J. Am. Chem. Soc. 114 (1992) 1645 O.Y. Hamra, L.D. Slep, J.A. Olabe, D.A. Estrin, Inorg. Chem. 37 (1998) 2033. (a) A.F. Schreiner, S.W. Lin, P.J. Hauser, B.A. Hopcus, D.J. Hamm, J.D. Gunter, Inorg. Chem. 11 (1972) 880. (b) M.G. Gomes, C.U. Davanzo, S.C. Silva, D.W. Franco, J. Chem. Soc., Dalton Trans. (1998) 601. (c) S.C. Silva, D.W. Franco, Spectrosc. Acta Part A 55 (1999) 1515. L.G.F. Lopes, M.G. Gomes, S.S.S. Borges, D.W. Franco, Aus. J. Chem. 51 (1998) 865. A.B.P. Lever, Inorg. Chem. 29 (1990) 1271. A.B.P. Lever, E.S. Dodsworth, in: E.I. Solomon, A.B.P Lever (Ed.), Inorganic Electronic Structure and Spectroscopy, vol. 2, Wiley, New York, 1999, p. 227.
