SciELO - Scientific Electronic Library Online

 
vol.48 número3X-RAY CRYSTAL AND MOLECULAR STRUCTURE OF AN ORGANOMETALLIC HYDRAZONE CONTAINING A ZIGZAG p-CONJUGATED SPACER: [CpFe(h6-C6H5)-N(Me)N=CH-C 6H4-4-NMe2]+ PF6-A COMPARISON OF SEMIEMPIRICAL AND AB INITIO METHODS FOR CALCULATING THE ELECTRONIC STRUCTURE OF C60 AND C70 FULLERENES índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Journal of the Chilean Chemical Society

versión On-line ISSN 0717-9707

J. Chil. Chem. Soc. v.48 n.3 Concepción sep. 2003

http://dx.doi.org/10.4067/S0717-97072003000300016 

J. Chil. Chem. Soc., 48, N 3 (2003) ISSN 0717-9324

A FAMILY OF RYTHENIUM COMPLEXES CONTAINING THE NON-INNOCENT LIGAND O-BENZOQUINONEDIIMINE.
AN INFRARED STRUCTURAL INTERPRETATION.

D. Venegas-YazigiA* ; M. Campos-ValletteA; A.B.P. LeverB ; J. CostamagnaC ; R.O. LatorreA;
W. Hernández G.A

A Department of Chemistry, Faculty of Sciences, Universidad de Chile,
Las Palmeras 3425, Ñuñoa, Santiago-Chile
B Department of Chemistry, Faculty of Pure and Applied Sciences, York University,
4700 Keele St., Toronto, Ontario-Canada M3J 1P3.
C Department of Chemistry, Faculty of Chemistry and Biology, Universidad de Santiago de Chile,
Av. B. O'Higgins 3363, Santiago-Chile.
dvy@manquehue.net

(Received: March 31, 2003 ­ Accepted: June 05, 2003)

ABSTRCT

It has been shown that innocent ligands always contribute with a similar electron density to a metal centre in a complex regardless of which other innocent ligands are bonded to the same centre, while non-innocent ligands are capable of tuning electron density on the metal centre depending on the nature of the other ligands. The present work reports the IR spectral characterisation of four ruthenium complexes containing the non-innocent ligand o-benzoquinonediimine and different innocent ligands (Ru(C6H4{NH}2)(Cl) 2(PPh3)2; [Ru(CH3CN)(C6H4{NH} 2)(Cl)(PPh3)2][BF4 ]; [Ru(CH3CN)2(C6H 4{NH}2)(PPh3)2 ][BF4]2, and [Ru(C6H4{NH}2)({C 2H5}2NCS2) (PPh3)2][Cl]. The C=N vibration correlates with the EL values of the ligands. We found that the o-benzoquinonediimine ligand modulates the vibrational energies depending on the nature of the innocent ligands bonded to the ruthenium atom.

Keywords: benzoquinonediimine, non-innocent ligand, triphenylphosphine, ruthenium, complexes.

INTRODUCTION

Since the publication of the electrochemical ligand parameter (EL) series [1], our group has been concerned with the behaviour of a family of ligands characterised by a high degree of mixing with the metal and more than one stable oxidation state. To distinguish both series, Jorgensen called the former "innocent" and the latter "non-innocent" ligands [2].

The EL value of the innocent ligands show additivity leading to the possibility to predict the experimental value of a metal centred redox process, i.e. a complex Ru(L1)(Lx)5 where L1 and Lx are innocent ligands, the experimental value (Ecalc) of the RuIII/RuII electrochemical process corresponds to:

Ecalc = EL1 + 5 * ELx

respect to NHE. The additivity of the electrochemical parameter, i.e. EL1 does not change upon the chemical nature of Lx.

We are presently focusing our study on complexes with one of these non-innocent ligands, o-benzoquinonediimine. This species has three stable oxidation states: the o-phenylenediamine (A), semiquinonediimine (B) and quinonediimine (C) forms.

Each oxidation state may be viewed as a different ligand by itself. We are evaluating the question of whether, if o-benzoquinonediimine is coordinated to the metal in one of its oxidation states, the other ligands are able to exert sufficient influence to change its electron density. In this context, it is important to assess the electronic properties of complexes containing this non-innocent ligand and a variety of ligands with different electron donor properties.

The crystal structures of four complexes Ru(C6H4{NH}2)(Cl) 2(PPh3)2, (2CL); [Ru(CH3CN)(C6H4{NH} 2)(Cl)(PPh3)2][BF4 ],(MECNCL); Ru(CH3CN)2(C6H 4{NH}2) (PPh3)2][BF4] 2, (2MECN) and [Ru(C6H4{NH}2)({C 2H5}2NCS2)(PPh 3)2] [Cl], (DEDTC) have been published [3-6]. The chemical structures of all four complexes are shown in scheme 1. These complexes contain one o-benzoquinonediimine chelating the ruthenium atom, plus two triphenylphosphine ligands trans to each other, and two additional ligands in the plane of the benzoquinonediimine ligand. Two short C=C bond distances in the BQDI ring in each crystal structure is indicative of a quinone arrangement (C) of the coordinated ligand [3-6]. Note that for a semiquinonediimine (B) and a phenylenediamine (A) arrangement of the ligand, equal C=C bond distances in each ring are expected.

Scheme 1: Structures of the complexes.


Infrared spectroscopy is a useful tool to assess the strength of the interaction of the BQDI ligand with the rest of the complex. Although systematic studies appear to be lacking, it seems reasonable to expect a correlation between EL value of an innocent ligand and the energies of some common vibrators. The present work shows a characterisation of the most common vibrators of the above mentioned ruthenium complexes incorporating one BQDI ligand and four innocent ligands.

EXPERIMENTAL SECTION

Synthesis of the complexes.

The complexes studied, Ru(C6H4{NH}2)(Cl) 2(PPh3)2, (2CL); [Ru(CH3CN) (C6H4{NH}2)(Cl)(PPh 3)2] [BF4], (MECNCL); [Ru(CH3CN)2(C6H 4{NH}2)(PPh3)2 ] [BF4]2, (2MECN) and [Ru(C6H4{NH}2)({C 2H5}2NCS2)(PPh 3)2][Cl], (DEDTC) were synthesised using published methods [6].

Infrared spectra

Fourier transform infrared (FT-IR) spectra of the complexes as KBr and polyethylene pellets were recorded in the spectral domains 4000-600 cm-1 and 600-250 cm-1 using a Bruker Vector 22 instrument. Two hundred scans were accumulated for the mid- and far- IR measurements, scanning with a resolution of 4 cm-1.

RESULTS AND DISCUSSION

Infrared spectral analysis

A list of the infrared frequencies is given in Table 1. Most assignments were made by comparison with published data for related molecules and characteristic group frequencies [7-10]. The mid- and far-IR spectra share many bands indicating that the overall structures of the complexes are not very different (see Fig 1 and Table 1). Large spectral differences only arise in the regions where the innocent ligand vibrations are expected, as all four complexes have in common two triphenylphosphine groups (TPP) in a trans arrangement and the BQDI ligand in the equatorial position. We expect all the complexes to show the common vibrators for the TPP and BQDI groups. For other ruthenium complexes containing BQDI, one band has been reported for the N-H stretching between 3280 ­ 3290 cm-1 [10] and two bands at 3340 and 3310 cm-1 [11]. In three of our complexes (2MECN, 2CL, DEDTC) we found a single band for the N-H stretching, at 3246, 3281 and 3200 cm-1 respectively; but for the asymmetrical MECNCL complex we observed two N-H stretching bands (3300 and 3268 cm-1). We found this last situation reasonable in view of the fact that this complex has two different ligands in the quinonediimine plane making each N-H different, thus producing a different trans effect on each N-H group of the BQDI ring.

Table 1. Abbreviations: w, weak; vw, very weak; vvw, very very weak; m, medium; s, strong; ms, medium strong; br, broad; sh, shoulder; Ph, phenyl; BQDI, o-benzoquinonediimine; c, ring out of plane deformation.


Fig. 1: FTIR spectra of the four complexes in KBr from 1600 to 600 cm-1

For the C=N fragment of the BQDI ligand, we could expect a band between 1300 and 1400 cm-1 as stated in an early paper [12,13]. The bands we assign to stretching of the C=N fragment for the four complexes appear between 1392 and 1367 cm-1. Since the C=C stretching bands appear in the same region, to assign the bands we assumed that the C=N frequencies depend upon the other ligands present in the complex, and the C=C bands are invariant with regard to the chemical nature of the ligands present in the first co-ordination sphere.

The P-phenyl stretching band has been reported at 1092 cm-1 [9]. In our spectra we found an intense band between 1094 and 1084 cm-1 for all four complexes. For the 2MECN and MECNCL complexes, which have tetrafluoroborate as the counterion, there is an additional broad band at 1070 cm-1 (for 2MECN) or 1063 cm-1 (for MECNCL), which we assigned to B-F stretching. Bands of variable intensity between 1040 and 947 cm-1 are assigned to C-H deformations.

For the DEDTC complex, there are four single bands in the region of 1500 to 1200 cm-1 which are attributable to C-N (diethyldithiocarbamate ligand) stretching and/or NCS fragment stretching. These bands do not appear in the other three complexes.

The MECNCL complex shows a medium intensity band at 841 cm-1 for which we do not have a unequivocal explanation at this time. A possible explanation could be that the low symmetry of this complex allows forbidden bands.

Correlations

If we accept that the electrochemical parameter (EL) of a ligand [1] is a measure of its ability to donate or withdraw electron density, we may try to correlate vibrational frequencies with the electrochemical parameters. In figure 2 we show the correlation between the C=N stretching frequency, nC=N, and the sum of the electrochemical parameters of the ligands trans to the quinonediimine group. For this correlation we use the four complexes, 2CL, 2MECN, MECNCL and DEDTC. It may be seen in figure 2 that a higher value of the electrochemical parameter, which implies a greater electron-accepting character, is associated with an increase of the C=N stretching frequency, i.e. a strengthening of the C=N bond. This is due to the non-innocent character of the BQDI ligand, so the Ru atom can move electron density towards or away from it. In this way the properties of the BQDI ligand respond to a much greater degree than those of an innocent ligand like acetonitrile. So as the EL value gets larger, + 0.34 V for acetonitrile, more electron density is moved off the ruthenium to the ligand; this makes the ruthenium atom a poorer p electron donor. In this way less electron density is moved towards the BQDI ligand, so it becomes more quinone-like leading to an increase of the nC=N value.


Fig. 2: nC=N vs. EL of the innocent ligand (Chloride , -0.24 V ; diethyldithiocarbamate , -0.08 V ; acetonitrile , +0.34 V) for complexes 2CL, DEDTC and 2MECN.

Plotting the C=N-Ru deformation frequency, dC=N-Ru, versus the sum of the electrochemical parameters of both ligands trans to the quinonediimine, we did not obtain a clear trend, as the DEDTC complex shows a much lower frequency than expected (figure 3). Taken together with the electrochemical data for these complexes [6], we suggest that the diethyldithiocarbamate may be behaving as a non-innocent ligand, a proposal which obviously warrants additional studies.


Fig. 3: dC=N-Ru vs. EL for complexes 2CL, DEDTC, MECNCL and 2MECN.

For the four complexes we found that the Ru-P stretching frequency, nRu-P, are in the region of 280 cm-1 according with literature [9]. It is interesting to point out that for the four complexes the Ru-P stretching frequencies are similar which is in contrast with crystal data for Ru-P distance. Crystal data show Ru-P in the range from 2.393 to 2.409 A for the complexes 2CL, 2MECN and MECNCL, while the complex DEDTC shows shorter Ru-P distances, 2.377 and 2.387 A [6].

CONCLUSIONS

The C=N stretching frequencies show an approximately linear dependence with the EL values of the ligands. We think that the effect of the electronic behaviour of the ligand trans to the quinone may only be observed on bonds, which are directly involved, in the first coordination sphere and trans to it. This is probably the reason why the C=N stretching band plot is the only one to show the expected behaviour.

We believe that the systematic variation of ligands within a family of complexes is an important tool to understand the effect of ligands with different electron donor-acceptor behaviour on the characteristics of the chemical bond. Consequently, we will direct our attention to the synthesis of new complexes with ligands having different EL values.

ACKNOWLEDGMENTS

D.V.Y. wants to acknowledge CONICYT for the scholarship and to FONDECYT 2980026 for the financial support. D.V.Y. and J.C. want to acknowledge FONDECYT 8010006 for financial support.

REFERENCES

[1].- Lever A.B.P., Inorg. Chem. (1990) 29, 1271-1285.         [ Links ]

[2].- Jorgensen C.K. "Absorption Spectra and Chemical Bonding in Complexes", Pergamon Press, Oxford, 1962.         [ Links ]

[3].- Venegas-Yazigi D., Mirza H., Lever A.B.P., Lough A., Costamagna J., Vega A., Latorre R. Acta Cryst. (2000) C-56, e245-e246.         [ Links ]

[4].- Venegas-Yazigi D., Mirza H., Lever A.B.P., Lough A., Costamagna J., Latorre R. Acta Cryst. (2000) C-56, e247-e248.         [ Links ]

[5].- Venegas-Yazigi D., Mirza H., Lever A.B.P., Lough A., Costamagna J., Latorre R. Acta Cryst. (2000) C-56, e281-e282.         [ Links ]

[6].- Diego Venegas-Yazigi, Doctoral Thesis, Universidad de Chile, 2001.         [ Links ]

[7].- Nakamoto K. Infrared and Raman Spectra of Inorganic and Co ordination Compounds, 5th edn., Wiley-Interscience, London, 1997.         [ Links ]

[8].- Lien-Vien D., Colthup N.B., Fately W.G. and Grasseli J.G. The Handbook of Infrared and Raman Characteristic Frequencies of Organic Molecules, 1st edn., Academic Press, Boston, 1991.         [ Links ]

[9].- Socrates G. Infrared Characteristic Group Frequencies, John- Wiley and Sons, Bristol, 1980.         [ Links ]

[10].- Belser P. von Zelewsky A., Zehnder M., Inorg. Chem. (1981) 21, 3098-3103.         [ Links ]

[11].- Obeso R., Cole-Hamilton D., Wilkinson G., J. Chem.Soc. Dalton Trans. (1979) 1618-1623.         [ Links ]

[12].- Barron A., Wilkinson G., Montevalli M., Hursthouse M., Poly hedron (1987) 6, 1089-1095.         [ Links ]

[13].- Anillo A., Barrio C., García-Granda S., Obeso-Rosete R., J. Chem. Soc. Dalton Trans. (1993) 1125-1130         [ Links ]