Mössbauer magnetization and nuclear magnetic resonance measurements on some iridium(I) complexes with fullerene ligands
M. Gál,a L. F. Kiss,b Z. Klencsár,c I. Pápai,d G. Schubert,d J. Rohonczy,a A. Vértes,c* F. E. Wagnere
a Department of General and Inorganic Chemistry, Eötvös L. University, H-1518 Budapest 112, P.O. Box 32, Hungary
b Research Institute for Solid State Physics and Optics, H-1525 Budapest, P.O. Box 49, Hungary
c Research Group for Nuclear Techniques in Structural Chemistry,
Hungarian Academy of Sciences at Eötvös L. University Budapest, H-1518 Budapest 112, P.O. Box 32, Hungary
d Chemical Research Center, Hungarian Academy of Sciences, H-1525 Budapest, P.O. Box 17, Hungary
e Physic-Department E-15, Technische Universität München, D-85474 Garching, Germany (Received September 15, 2003)
The isomer shifts of the measured Mössbauer spectra are in accordance with quantum chemical calculations and both techniques showed that the studied iridium(I) complexes have a strong covalent character, namely, there are a direct donation of the ligands into the 6s orbital of Ir. The coordination of further -acceptor ligands decreases the population of 6s orbital. The obtained quadrupole splittings detect significant difference among the electric field gradients at the nucleus of iridium in the discussed coordination compounds. In this paper we present the first attempt to obtain Mössbauer parameters of 193Ir by DFT calculations. The change in the measured quadrupole splittings due to the coordiantion of tetracyanoethylene to central iridium was excellently reproduced by theoretical calculation using ZORA scheme. The magnetization measurements proved that IrI has diamagnetic, singlet electronic structure in all the studied compounds. This finding was in accordance with density functional calculations as well. Nuclear magnetic resonance investigations on 31P nucleus could detect the decrease of the electric field gradient at the nucleus of 193Ir in IrCl(CO)(PPh3)2 due to the coordination of C60 to the central iridium atom.
Introduction
Iridium(I) has 5d8 electronic structure. It means that IrI is an adequate central atom for the -acceptor ligands like fullerene. It is the reason that several IrI coordination compounds were prepared after the discovery of fullerenes.
For examples: The BALCH’s group1 prepared chloro(carbonyl)bis(triphenylphosphine)iridium (CCTI) with buckminsterfullerene (C60) and RASINKANGAS and his co-workers2 produced di-µ-chloro-bis(1,5-cyclo- octadiene)-diiridium, (Ir2Cl2(1,5-COD)2) adduct with C60.
Iridium is a Mössbauer active atom and it makes possible to carry out Mössbauer studies on these complexes.3–5
The aim of this work is to complete the Mössbauer measurements by the spectrum of the Ir2Cl2(1,5-COD)2 with C70 ligand and to obtain information about the differences between the chemical interactions of C60 and C70 with IrI central ion. We recorded magnetization and nuclear magnetic resonance (NMR) curves on this materials to compare the Mössbauer results with the magnetic results. We also carried out quantum chemical calculations to provide preliminary tests for applying relativistic density functional theory to calculate Ir quadrupole splittings. It was the first and successful attempt to interpret 193Ir Mössbauer spectra by DFT calculations.
* E-mail: vertesa@ludens.elte.hu
The authors dedicate this work to Professor Rudolf Mössbauer on the occasion of his 75th birthday.
Experimental Preparation of the complexes
The preparation of the complexes studied were carried out using iridium(III) chloride 3-hydrate (Johnson Mattey) and fullerenes (C60, C70) of 99.99 purity (MER Company).
The precursors, the carbonylchlorobis(triphenyl- phosphine)iridium (Vaska’s complex) and the dichloro- bis(cyclooctadiene)diiridium were prepared by standard methods.6,7
The procedures for synthesis of the investigated iridium(I) complexes are reported elsewhere.1,2,8
A new di-adduct is obtained by the reaction of Ir2Cl2(cyclooctadiene)2 with C70 (in benzene solvent, at 55 °C). The yield is about 70%.
Mössbauer spectra
The Mössbauer spectra of the 193Ir 73 keV -transition were taken at 4.2 K. The source was metallic 193Os, which, being hexagonal, exhibits a quadrupole splitting of about 0.48 mm/s. The individual lines of the quadrupole doublets of the studied samples are, therefore, not simple Lorentzians. This circumstance is taken into account in the fitting procedure. The full widths at half- maximum of the individual superimposed Lorentzian lines obtained by the least squares fits are between
0.75 and 0.79 mm.s–1 and thus lightly larger than the natural line width of 0.60 mm.s–1. The isomer shift ( ) values are given relative to metallic Ir.
Magnetic measurements
The magnetic measurements were performed by a Quantum Design MPMS-5S SQUID magnetometer at the temperature and field range of 1.8 K T 300 K and 0 H 5 T using a gelatine capsule as a sample holder for each powder sample between 60 and 120 mg. During the measurements the sample was surrounded by He as an exchange gas of several mbar pressure. The measured moment was corrected for the (almost diamagnetic) signal of the capsule whose contribution made up 30–
50% of the total measured moment.
NMR spectra
All NMR spectra were recorded on Bruker DRX 500 spectrometer at 11.744 Tesla. 31P T1 relaxation time measurements were performed at 202.404 MHz in a 7 mm MAS probe. Samples were rotated with 5 kHz speed at MAS conditions in ZrO2 rotor. One scan of 4k data points were acquired during 31 ms in a 322 ppm spectrum window after the inversion recovery pulse sequence (with 9.5 µs as 90 degree pulse at –6 dB using 300 W transmitter). 2000-second relaxation delay was used prior each scans to fulfill the 5 T1 time requirements. 2k real data points were observed after the exponential multiplication with LB=30 Hz and Fourier transform. All chemical shift values were referenced to external 85% phosphoric acid sample.
Results and discussion Mössbauer spectroscopy
The recorded 193Ir Mössbauer spectra are shown in Fig. 1, and their parameters, together with the earlier published results, are summarized in Table 1. For the better understanding of the chemical properties of fullerenes we used tetracyanoethylene (TCNE) for the adduct formation as well. The coordinating carbon atoms have only carbon neighbours in both TCNE and fullerene. Furthermore, TCNE is known as a strong -acceptor3 and we compared the effects of TCNE and C60.
The minor component (~5%) on Fig. 1b has an isomer shift equal to –1.04 mm.s–1 and its quadrupole splitting is 2.39 mm.s–1. The line width of this component (1.7 mm.s–1) is about two times larger than that of the major component. On the basis of these Mössbauer parameters9 we can suppose that IrI was oxidized into IrIII in a small part of the precursor Ir2Cl2(COD)2 due to the dissolved oxygen in the
solution. (It is worth mention that the Mössbauer measurements were carried out later than the magnetic investigations and the samples were prepared freshly for each experiments.)
The value of the isomer shift of CCTI,
=+0.02 mm.s–1, is about that of iridium metal. Note that typical isomer shifts of octahedral, ionic Ir+ compounds extrapolated to the 5d86s0 configuration would give an isomer shift about –4 mm.s–1 as it was published earlier.3 The more positive shift is attributed to direct donation into the 6s orbital as well as to the hybridization of 6s with 5dz2. On the other hand, the isomer shifts of the Ir2Cl2(COD)2 dimer and of its C60 adduct (–0.88 and –0.98 mm.s–1) are to our knowledge the lowest ones found for Ir(I) compounds.9 Hence, the COD ligands must be a very weak -donor. The small EQ value of 3.81 mm.s–1 found for Ir2Cl2(COD)2 may then be understood as the result of a better cancellation of the EFG’s (electric field gradients) from the dz2 and the “in-plane” -bonding electron density. Whether the resulting EFG is positive or negative cannot be decided so far, but it might be possible that the dz2 orbitals with their negative EFG make the larger contribution in this system since the isomer shift indicates that the - donation of the equatorial ligands is in fact particularly weak.
Upon coordination of fullerenes to CCTI and to Ir2Cl2(COD)2, a decrease of the isomer shift is observed in both cases ( of CCTI( 2–C60). 5C6H6: –0.26 mm.s–1; of {Ir2Cl2(COD)2}2(µ–C60)·2C6H6: –0.98 mm.s–1 and of {Ir2Cl2(COD)2}2(µ–C70)·2C6H6: –0.93 mm.s–1). Since C60 is a ligand capable of backbonding, a reduction of d-electron density is expected in the adducts which would shift to higher values as compared to the precursors. The fact that in contrary a decrease is observed must be due to a decrease of 6s electron density which overcompensated the reduction of shielding by d electrons in the C60 adducts.
The considerable similarity between the Mössbauer parameters of the coordination compounds with C60 and C70 ligands proves that the chemical structures of the coordinations of both fullerenes to IrI are very resembling. Fullerenes containing sixty or seventy carbon atoms have very similar -acceptor ability.
Nevertheless, the measured data in Table 1 belonging to the C70 ligands are slightly closer to those of the precursor than in the case of C60 ligands. This experimental finding means that C60 has a somewhat stronger -acceptor property than C70.
Magnetization measurements
The magnetization measurements were carried out as a function of the magnetic field (0 H 5 T) at T=1.8, 5 and 300 K and temperature (1.8 K T 300 K) at H=5 T.
A few magnetization curves are shown in Figs 2 to 4.
The magnetization vs. magnetic field curves could be decomposed into three types of magnetic behavior:
M i H B
n
M = µH (H,T)+ dia + 0 (1) where the first term is a paramagnetic contribution described by a Brillouin function:
T k
H J J
T k
H J J J
B J
B H
B H
2 coth 1 2
1
2 1 coth2 2
1 ) 2
T H, (
µ µ +
= +
(2)
where µH denotes the maximum moment of each moment-carrying entity (atom) in the direction of the field, n their number per unit mass (or mol), J the total angular momentum and kB Boltzmann’s constant. The second term is a diamagnetic contribution characterized by a temperature and magnetic field independent susceptibility, dia, the third term is a temperature and field independent magnetization, M0i, originating probably from ferromagnetic impurities
Fig. 1. Mössbauer spectra of Ir2Cl2(1.5 – COD)2 (a) and C70·[Ir2Cl2(1.5 – COD)2]2·2C6H6 (b). Spectra were recorded at 4.5 K
Table 1. Recorded 193Ir Mössbauer spectra, their parameters and references No. of
samples
Iridium compound Isomer shifta
(IS), mm.s–1
Quadrupole splitting ( EQ), mm.s–1
Reference 1
2 3 4 5 6 7
IrCl(CO)(PPh3)2
IrCl(CO)(PPh3)2 ( 2–TCNE)b ( 2–C60)·[IrCl(CO)(PPh3)2].5C6H6 Ir2Cl2(1,5 – COD)2
C60·[Ir2Cl2(1,5 – COD)2]2.2C6H6 Ir2Cl2(1,5 – COD)2
C70·[Ir2Cl2(1,5 – COD)2]2.2C6H6
+0.02 ± 0.02 –0.19 ± 0.02 –0.26 ± 0.02 –0.88 ± 0.02 –0.98 ± 0.02 –0.87 ± 0.02 –0.93 ± 0.02 –1.04 ± 0.02c
6.52 ± 0.03 1.84 ± 0.02 2.71 ± 0.02 3.81 ± 0.02 6.18 ± 0.02 3.85 ± 0.02 5.99 ± 0.02 2.39 ± 0.02
3 3 3 4 4
*
*
a With respect to metallic Ir.
b The adduct of CCTI with tetracyanoethylene (TCNE) was measured for comparison with fullerene. Namely, it is known that TCNE is a strong -acceptor.
c The area ratio of this minor component is 5%.
* Present paper.
Fig. 2. Magnetization of CCTI as a function of magnetic field at T=5 and 300 K; (a) The continuous lines are fits to Eq. (1) using the parameters listed in Table 2. Magnetization of CCTI as a function of temperature at H=5 T (b)
Fig. 3. Magnetization of CCTI - C60 as a function of magnetic field at T=1.8, 5 and 300 K; (a) The continuous lines are fits to Eq. (1) using the parameters listed in Table 2. Magnetization of CCTI - C60 as a function of temperature at H=5 T (b)
It is evident from the measurements that at low temperatures (below T=150–200 K) the magnetic behavior is dominated by the paramagnetic contribution.
On the contrary, at higher temperatures (above T=150–
200 K) the dominating part is the diamagnetic contribution which makes possible to separate the ferromagnetic contribution, M0i, considered to be constant in the course of the fitting procedure. The total angular momentum of the paramagnetic entity, J, turned out to be a relative insensitive fitting parameter. The most reliable determination of J is expected for CCTI- C60 with the highest paramagnetic contribution, giving J=4, which was then used as a constant for all the samples studied. All fitting parameters are summarized for all the samples studied in Table 2.
Since the paramagnetic moment density is of the order of several 10–20 mol–1, the concentration of the moment-carrying atoms or ions should be less than about 1000 ppm (see in detail below). It means that the overwhelming majority of the ions in the studied complexes including IrI gives a diamagnetic contribution to the magnetization. The deduced values of the diamagnetic susceptibility (Table 2) are systematically higher for the co-ordination compounds containing fullerene compared to those of the pure precursors. The differences between them are close to the dia values for pure C60 and C70.10–12 The magnetic measurements show that IrI has a diamagnetic, singlet electronic structure in all the studied materials.
Fig. 4. Magnetization of C70.[Ir2Cl2(1,5 – COD)2]2.2C6H6 as a function of magnetic field at T=1.8, 5 and 300 K; (a) The continuous lines are fits to Eq. (1) using the parameters listed in Table 2. Magnetization of C70.[Ir2Cl2(1,5 – COD)2]2.2C6H6 as a function of temperature at H=5 T (b)
It is evident from Table 2 that the moment density is of the order of 1000 ppm. This quite small figure means that the atoms (ions) carrying moments between 2 and 6 µB can be considered to be an impurity. A paramagnetic impurity of about 100 ppm was found also in C60.10–12 (These results show that magnetization measurements can detect the very low concentration of impurities as well.) It is worth noting that the ground state of the free Ir atom is 4F9/2 (L=3, S=3/2 and J=9/2). This value for J (9/2) is very close to that extracted from our fitting (J=4). However, being an insensitive fitting parameter, the experimental points would allow for J to lie in the range higher than 3.5. The theoretical Lande factor, g, is calculated to be g=1.33 for the free Ir atom, giving µH=g J=6 µB for the
maximum moment of the atom in the direction of the field. Similar µH values were measured for CCTI-C60 and CCTI-C70, for the other samples studied they lie in the µH=2–5 µB range.
31P NMR measurements
The 31P NMR spectra of IrCl(CO)(PPh3)2 (A) and ( 2–C60)·[IrCl(CO)(PPh3)2].5C6H6 (B) are compared in Fig. 5. Asterisks mark the isotrope signals of PPh3 groups at 25.7 ppm and at –7.7 ppm in compound A and B, respectively. In both cases several rotational sidebands exhibit big chemical shielding anisotropies (CSA) of the phosphorous shielding tensors. This anisotropy seems to be smaller (fewer and more
symmetrical sidebands) in the case of B, where a C60 ligand is connected to the Ir atom with 2 bond. This - donor ligand can increase the -donor character of the Ir in the Ir-P bond resulting a bigger chemical shielding of the phosphorous atom. This leads to the observed –7.7 ppm chemical shift value, while the similar phosphorous signal of A is at 25.7 ppm. The 31P spin- lattice relaxation times of PPh3 groups in compounds A and B were measured by inversion recovery method at MAS (magic angle spinning) conditions (Fig. 6.). The relaxation time values were found 348±14 and 610±28 seconds in A and B, respectively. In the case of
compound B, the more symmetrical shielding of the phosphorous atom results in a longer T1 with a less efficient CSA relaxation mechanism. Another possible reason of the longer relaxation time is a weaker quadrupole interaction between the Ir nucleus and the neighbouring valence electrons (e.g., and -bonds between the Ir and P atoms). This weaker quadrupole interaction correlates with the smaller Mössbauer quadrupole splitting, which is 2.71 mm/s in the case of B, while this value is 6.52 mm.s–1 in the case of A (Table 1).
Table 2. Results of the fits to Eq. (1) Sample
mol. weight, g dia
,
×10–6emu.mol–1.Oe–1 M0i, emu.mol–1
µH (µB)
n,
×1020 mol–1 IrCl(CO)(PPh3)2 (CCTI)
780.25165
–371 0.382 1.94 1.84a
– IrCl(CO)(PPh3)2 ( 2–TCNE)
908.34535
–432 1.12 4.85 2.58a
2.29b ( 2–C60).[IrCl(CO)(PPh3)2] 5C6H6
1891.4942
–848 5.83 5.72 8.90a
8.67b ( 2–C70).[IrCl(CO)(PPh3)2] 2.5C6H6
1816.3189
–984 4.09 5.93 7.22a
7.04b Ir2Cl2(1.5 – COD)2
671.68
–254 0.329 3.34 1.04a
0.704b C60.[Ir2Cl2(1.5 – COD)2]2 2C6H6
2220.25
–905 0.688 4.42 8.21a
7.31b C70.[Ir2Cl2(1.5 – COD)2]2 2C6H6
2340.36
–1204 0.983 3.24 6.82a
6.62b
a 5 K, 300 K.
b 1.8 K.
Fitting parameters: dia temperature independent diamagnetic susceptibility, M0i temperature and field independent magnetization, µH maximum moment in field direction and n their number per mol.
Fig. 5. 31P MAS NMR spectra of IrCl(CO)(PPh3)2 (a) and ( 2–C60).[IrCl(CO)(PPh3)2].5C6H6 (b). Asterisk marks the isotrope signal of the PPh3 group
Fig. 6. 31P NMR T1 peak intensities versus delay time of IrCl(CO)(PPh3)2 (curve 1, signed with squares) and ( 2–C60).[IrCl(CO)(PPh3)2].5C6H6 (curve 2, signed with triangles)
DFT calculation
We carried out relativistic density functional calculations to gain more insight into the electronic structure of the investigated Ir complexes and to compare the results of the calculations and the parameters of the Mössbauer spectra. Two model compounds were considered to examine the changes occurring in the electronic structure of Ir atom upon the coordination of a -acceptor ligand to CCTI. In our models, the bulky PPh3 ligands were replaced by PH3 units, so as to keep the computations in reasonable bounds, and only the coordination of TCNE was considered.
The geometries of the two complexes were first optimized at the LDA (Local Density Approximation) and the GGA (General Gradient Approximation) levels of DFT using the VWN-5 local13 and the BP8614,15 gradient-corrected exchange-correlation functionals along with the ZORA/TZP (triple-zeta plus polarization functions) basis set. The core orbitals were kept frozen and scalar relativistic effects were included in these calculations.
The optimized geometries and some selected structural parameters of the IrCl(CO)(PH3)2 and IrCl(CO)(PH3)2( 2-TCNE) model compounds are illustrated in Fig. 7. As we can see on this figure, the calculated bond distances and those obtained by X-ray diffraction1 on ( 2-C60).[IrCl(CO)(PPh3)2] are very similar. It suggests that the used model compound gives
a correct basis for the calculations of the parameters of the electronic structure of iridium(I) central atom and of its Mössbauer parameters.
We note that both compounds (CCTI and its adduct with TCNE) have a singlet electronic ground state and the states with higher spin multiplicities lie far above the ground state. For instance, the triplet state of the TCNE complex is 75 kcal/mol less stable than the singlet state.
As expected, the coordination of TCNE induces drastic changes in the structure of the IrCl(CO)(PH3)2 unit. The two phosphine ligands are bent away from the coordinating TCNE giving rise to a pseudo trigonal bipyramidal arrangement of the ligands. These structural changes, along with the formation of the Ir-TCNE dative bond, imply significant variations in the electron distribution around the Ir atom. The electron configuration of Ir in IrCl(CO)(PH3)2 is 6s0.75 6p0.42 5d7.90 as obtained from Mulliken population analysis, which is modified to 6s0.55 6p0.86 5d7.74 after the coordination of TCNE. These changes underline the importance of s-d and s-p hybridization mechanisms in the formation of the Ir-TCNE dative bond. The results of calculations are consistent with Mössbauer measurements namely the decrease of 6s population due to the TCNE coordination explains the decrease of isomer shifts of the Mössbauer spectra (Table 1). These results suggest, again, that the used model compounds substitute correctly for the IrI complexes used for Mössbauer studies.
Fig. 7. Optimized geometries and selected structural parameters for IrCl(CO)(PH3)2 and IrCl(CO)(PH3)2( 2-TCNE). Bond distances are given in Ångstroms as obtained from GGA calculations, numbers in parantheses are values from LDA calculations and numbers in square brackets were
obtained by single crystal X-ray diffraction on ( 2–C60).[IrCl(CO)(PPh3)2]. (The measured values were taken from Reference 1)
Table 3. Calculated EFG components (Vzz and ) and quadrupole splittings ( EQ) for the IrCl(CO)(PH3)2 and IrCl(CO)(PH3)2( 2-TCNE) model complexes
IrCl(CO)(PH3)2 IrCl(CO)(PH3)2( 2-TCNE)
ZORA EFG/GEOMa Vzz, a.u. EQ, mm.s–1 Vzz, a.u. EQ, mm.s–1
SRb GGA/GGA 2.530 –0.190 3.81 1.234 –0.410 1.90
GGA/LDA 3.410 –0.049 5.11 1.264 –0.374 1.94
LDA/GGA 2.416 –0.231 3.65 1.239 –0.467 1.92
LDA/LDA 3.297 –0.073 4.95 1.268 –0.431 1.96
SOc GGA/GGA 3.000 –0.084 4.50 1.154 –0.385 1.77
expd 6.52 1.84
ZORA-4
SRb GGA/GGA 1.365 –0.582 2.16 1.396 –0.556 2.20
GGA/LDA 2.208 –0.229 3.34 1.425 –0.520 2.23
LDA/GGA 1.285 –0.676 2.07 1.396 –0.602 2.21
LDA/LDA 2.129 –0.270 3.23 1.425 –0.566 2.25
SOc GGA/GGA 1.487 –0.572 2.35 1.394 –0.591 2.21
expd 6.52 1.84
a In EFG/GEOM, EFG denotes the functional used for the EFG calculation, GEOM refers to the functional used for geometry optimization.
b SR: ZORA scalar relativistic approach.
c SO: Relativistic calculation including spin-orbit coupling.
d exp: Experimental Mössbauer data refer to chloro(carbonyl)bis(triphenyl-phosphine)iridium and its adduct with tetracyanoethylene, respectively.
Recently, the calculation of electric field gradients (EFG) became feasible within the density functional theory (see for instance References 16–19 and references therein). Particularly, the zeroth-order regular approximation (ZORA) relativistic density functional method and its improved ZORA-4 variant have shown to provide reasonable EFGs for various centers (halogens, IIIA metals) even with relatively small basis sets.16–22 The calculations were carried out using the ADF 2002.
02 package.23 No experience exists, however, for the application of these methods to predict EFGs for heavier transition metal centers. The present Mössbauer data provide an opportunity for test calculations, therefore,
we applied the ZORA and ZORA-4 methods to evaluate the Ir quadrupole splittings in IrCl(CO)(PH3)2 and IrCl(CO)(PH3)2( 2-TCNE).
The EFG calculations were also carried out at both LDA and GGA levels of theory, but using an all-electron (ZORA/TZP) basis set. The ZORA methodology was used to include the scalar relativistic (SR) and spin-orbit coupling (SO) effects. The ZORA-4 scheme,23 which incorporates the small component density effects in the relativistic treatment, was also applied for the EFG calculations.
The EFG components were calculated in principal axis system (|Vzz|>|Vyy|>|Vxx|, by convention) and they
were used to estimate the quadrupole splitting ( EQ) for the 193Ir I=3/2 level according to:
1 3 2
+ 2
= zz
Q eQV
E
where e is the electron charge, Q is the quadrupole moment of the 193Ir nucleus in its I=3/2 excited state (Q=0.751.10–28 m2), Vzz is the largets EFG component and is the asymmetry parameter of the EFG ( =(Vxx–Vyy)/Vzz). In order to get EQ in mm.s–1, one has to express eQVzz in the same units, which can be done by:
eQVzz [mm/s] = (eQVzz [Joule] c[mm/s])/E [Joule]
where c is the speed of light in vacuum, and E is the energy of the 193Ir Mössbauer -ray (for I=3/2 I=1/2 transition, E =73.0 keV or 1.1696.10-14 Joule). The calculated results are listed in Table 3.
Note first that the EQ values calculated for the IrCl(CO)(PH3)2( 2-TCNE) complex are all reasonably close to the observed quadrupole splitting of CCTI-( 2- TCNE), especially the ZORA predictions appear to be in very good agreement with experiment. It is also apparent from Table 3 that the calculated values for this complex are insensitive to the level of density functional theory.
The situation is rather different for the IrCl(CO)(PH3)2 complex, for which the calculated EQ values range in a notably broader interval (3.8–5.1 mm.s–1 from ZORA and 2.1–3.3 mm.s–1 from ZORA-4 calculations). For a given geometry, the LDA and GGA calculations give quite similar results, however, the predictions seem to be quite sensitive to the reference geometry. The ZORA values fall much closer to the measured EQ=6.52 mm.s–1 splitting than the corresponding ZORA-4 results.
The experimental Mössbauer data indicate that the quadrupole splitting of the 193Ir decreases significantly upon the coordination of the TCNE ligand to CCTI. Our present calculations show that this tendency is reasonably reproduced with the ZORA scheme, which is not the case when using the ZORA-4 method. This finding is quite surprising in light of a recent study,16 which clearly concluded that the ZORA-4 methodology is superior to the standard ZORA approach. We have examined the quality of the basis set and the numerical integration we used in our calculations as possible sources for the above discrepancy by repeating some of our EFG calculations with the QZ4P basis set or a finer integration grid, but we found only minor effects.
Systematic studies on “real size” models ought to be carried out to reveal the applicability of the ZORA and ZORA-4 schemes for the calculation of quadrupole splittings of transition metal centers.
*
This work was supported by the Hungarian Science Foundation OTKA (T 034839).
References
1. A. L. BALCH, V. J. CATALANO, W. LEE, Inorg. Chem., 30 (1991) 3980.
2. M. RASINKANGAS, T. T. PAKKANEN, T. A. PAKKANEN, M. AHLGREN, J. ROUVINEN, J. Am. Chem. Soc., 115 (1993) 4901.
3. A. VÉRTES, M. GÁL, F. E. WAGNER, F. TUCZEK, P. GÜTLICH, Inorg. Chem., 32 (1993) 4478.
4. F. TUCZEK, M. GÁL, F. E. WAGNER, A. VÉRTES, Electrochem.
Soc. Proc., 95 (1995) 1092.
5. F. TUCZEK, M. GÁL, F. E. WAGNER, A. VÉRTES, Fullerene Sci.
Technol., 5 (1997) 443.
6. J. P. COLLMAN, C. T. SEARS Jr., M. KUBATA, A. DAVISON, E. T. SHAWL, Inorg. Synth., 11 (1968) 102.
7. R. CRAMER, J. A. MCCLEVERTY, J. BRAY, Inorg. Synth., 15 (1974) 18.
8. A. L. BALCH, V. J. CATALANO, J. W. LEE, M. L. OLMSTEAD, S. R. PARKIN, J. Am. Chem. Soc., 113 (1991) 8953.
9. G. K. SHENOY, F. E. WAGNER, in: Mössbauer Isomer Shifts, North-Holland, 1978, p. 443.
10. R. S. RUOFF, D. BEACH, J. CUOMO, T. MCGUIRE, R. L. WHETTEN, F. DIEDERICH, J. Phys. Chem., 95 (1991) 3457.
11. R. TYCKO, R. C. HADDON, G. DABBAGH, S. H. GLARUM, D. C. DOUGLASS, A. M. MUJSCE, J. Phys. Chem., 95 (1991) 518.
12. R. C. HADDON, L. F. SCHNEEMEYER, J. V. WASZCZAK, S. H. GLARUM, R. TYCKO, G. DABBAGH, A. R. KORTAN, A. J. MULLER, A. M. MUJSCE, M. J. ROSSEINSKY, S. M. ZAHURAK, A. V. MAKHIJA, F. A. THIEL, K. RAGHAVACHARI, E. COCKAYNE, V. ELSER, Nature, 350 (1991) 46.
13. S. H. VOSKO, L. WILK, M. NUSAIR, Can. J. Phys., 58 (1980) 1200.
14. A. D. BECKE, Phys. Rev., A38 (1988) 3098.
15. J. P. PERDEW, Phys. Rev., B33 (1986) 8822.
16. E. VAN LENTHE, E. J. BAERENDS, J. Chem. Phys., 112 (2000) 8279.
17. P. DUFEK, P. BLAHA, K. SCHWARZ, Phys. Rev. Lett., 75 (1998) 3545.
18. R. H. HAVLIN, N. GODBOUT, R. SALZMANN, M. WOJDELSKI, W.
ARNOLD, C. E. SCHULZ, E. OLDFIELD, J. Am. Chem. Soc., 120 (1998) 3144.
19. J. ANTONY, B. HANSEN, L. HEMMINGSEN, R. BAUER, J. Phys. Chem., A104 (2000) 6047.
20. E. VAN LENTHE, E. J. BAERENDS, J. G. SNIJDERS, J. Chem. Phys., 99 (1993) 4597.
21. E. VAN LENTHE, E. J. BAERENDS, J. G. SNIJDERS, J. Chem. Phys., 101 (1994) 9783.
22. E. VAN LENTHE, A. E. EHLERS, E. J. BAERENDS, J. Chem. Phys., 110 (1999) 8943.
23. E. J. BAERENDS, J. A. AUTSCHBACH, A. BÉRCES, C. BO, P. M.
BOERRIGTER, L. CAVALLO, D. P. CHONG, L. DENG, R. M.
DICKSON, D. E. ELLIS, L. FAN, T. H. FISCHER, C. FONSECA
GUERRA, S. J. A. VAN GISBERGEN, J. A. GROENEVELD, O. V. GRITSENKO, M. GRÜNING, F. E. HARRIS, P. VAN DEN HOEK, H. JACOBSEN, G. VAN KESSEL, F. KOOTSTRA, E. VAN LENTHE, V. P. OSINGA, S. PATCHKOVSKII, P. H. T. PHILIPSEN, D. POST, C. C. PYE, W. RAVENEK, P. ROS, P. R. T. SCHIPPER, G. SCHRECKENBACH, J. G. SNIJDERS, M. SOLA, M. SWART, D. SWERHONE, G. TE VELDE, P. VERNOOIJS, L. VERSLUIS, O. VISSER, E. VAN WEZENBEEK, G. WIESENEKKER, S. K. WOLFF, T. K. WOO, T. ZIEGLER, ADF2002.02, SCM, in: Theoretical Chemistry, Vrije Universiteit, Amsterdam, The Netherlands, 2002.
