Kinetics and mechanism of a trans-tetraazamacrocyclic chromium(III

0 downloads 0 Views 305KB Size Report
concentration. The mechanism of the reaction has been discussed. A relatively inert intermediate chromium(V) species was detected based on characteristic ...
Transition Metal Chemistry 29: 855–860, 2004.  2004 Kluwer Academic Publishers. Printed in the Netherlands.

855

Kinetics and mechanism of a trans-tetraazamacrocyclic chromium(III) complex oxidation by hexacyanoferrate(III) in strongly alkaline media Anna Katafias*, Olga Impert, Przemysław Kita and Grzegorz Wrzeszcz Department of Chemistry, N. Copernicus University, 87-100 Torun´, Poland Received 06 April 2004; accepted 05 May 2004

Abstract Oxidation of the trans-[Cr(cyca)(OH)2 ]þ complex, where cyca ¼ meso-5,5,7,12,12,14-hexamethyl-1,4,8,11-tetraazacyclotetradecane, by [Fe(CN)6 ]3 ion in strongly alkaline media, leading to [CrV O(cycaox )]3þ ion, has been studied using electronic and e.p.r. spectroscopy. The kinetics of the CrIII ! CrIV transformation have been studied using a large excess of the reductant and OH ion over the oxidant. The reaction is a second order process: first order in [CrIII ] and [FeIII ] at constant [OH ]. The second order rate constant is higher than linearly dependent on the OH concentration. The mechanism of the reaction has been discussed. A relatively inert intermediate chromium(V) species was detected based on characteristic bands in the visible region and the e.p.r. signal at giso ¼ 1:987 for the systems where an excess of oxidant was used. The hyperfine structure of the main e.p.r. signal is consistent with the d1 -electron interactions with four equivalent nitrogen nuclei and [CrV ¼ O(cycaox )]3þ formula, where cycaox ¼ oxidized cyca, can be postulated for the intermediate CrV complex.

cyca ligand

Introduction Controversy about the role of chromium as one of the essential elements in biological systems, e.g. [1–4] acting at oxidation state +3 and on the other hand the well documented mutagenic and carcinogenic properties of chromium(VI), e.g. [4–6] has a large impact on research in recent years. Thermodynamic trends and kinetics of the thermal CrIII Ð CrVI transformation in aqueous solutions strongly depends on pH [7]. Chromium species at intermediate oxidation states +4 and +5 are unstable within the whole pH range. Recently some results on the structure and reactivity of chromium(IV) and chromium(V) complexes have been published [8–11]. Previously [12], we applied an electron transfer between a tetraazamacrocyclic chromium(III) complex and [Fe(CN)6 ]3 ion proceeding in strongly alkaline solution for generation of a long lived chromium(V) species stabilized by the macrocyclic ligand. Kinetics of the CrIII ! CrIV stage have been studied using cis-[Cr(cycb)(OH)2 ]þ where cycb ¼ rac-5,5,7,12,12,14hexamethyl-1,4,8,11-tetraazacyklotetradecane, and chromate(III) as the electron donors. In this work the other geometrical isomer, trans-[Cr(cyca)(OH)2 ]þ complex, where cyca ¼ meso-5,5,7,12,12,14-hexamethyl1,4,8,11-tetraazacyclotetradecane, is the starting chromium(III) species in the redox process. The structure of the macrocyclic ligand is shown in Scheme 1.

* Author for correspondence

Scheme 1 The aim of this project is the examination of structural effects on chromium(V) stability in solution and on the kinetics and mechanism of the CrIII ! CIV stage which is the rate limiting step in the CrIII ! CrV transformation [12]. E.p.r. and electronic spectroscopy have been applied.

Experimental Materials Trans-[Cr(cyca)(H2 O)2 ]Br3  3H2 O was prepared and characterized as reported previously [13]. K3 [Fe(CN)6 ], KOH and KBr (POCH Gliwice) were of pro analysi grade quality and were used without further purification. Solutions were prepared using water redistilled from alkaline KMnO4 and purged with Ar to minimalize contamination by CO2 .

856

E.p.r. spectra were recorded with an EPR Bruker Physik 418S reflection type spectrometer in X-band (ca. 9.5 GHz) with a 100 kHz modulation of a steady magnetic field. The microwave frequency was monitored with an 18 GHz microwave counter (Marconi Instrument). The magnetic field was measured with an automatic MJ-110 R n.m.r.-type magnetometer (Radiopan). All measurements were recorded for the reaction mixture of 8  103 M chromium(III) and 0.016–0.8 M iron(III) oxidant, at 0.1 M KOH, at 295 K. The flat quartz cell was used. WinSim e.p.r. simulation software [14] was used to estimate the spectral parameters. The simulations do not include contributions from the minor peaks that arise from the 53 Cr hyperfine coupling. However, the small error introduced this way is comparable to the experimental errors involved in fitting procedures and can be ignored. In the trans-[CrIII (cyca)(OH)2 ]þ – [Fe(CN)6 ]3 reaction mixture, chromium(V) species are formed relatively fast and were detected 3–15 min. after preparation of the reaction solutions; the larger excess of the oxidant the faster chromium(V) formation. Using a 100-fold excess of FeIII gives chromium(V) species immediately, although even a 3-fold excess of FeIII is enough to form slowly detectable amount of CrV . In the first case, the initial e.p.r. signal shows its maximum intensity and decayed after ca. 1 h. In the second case, intensities of the chromium(V) e.p.r. signals increase, reach a maximum and then decay during 45 min. The highest stability of chromium(V) species, up to several hours, was achieved at room temperature by using 6.25:1 molar ratio of FeIII to CrIII , cf. Figure 1. Kinetic measurements The rate of the reaction was followed spectrophotometrically using a HP 8453 diode-array instrument,

equipped with a Peltier HP 89090A Temperature Controller. Spectral changes due to FeIII to FeII reduction are presented in Figure 2. The reaction was studied under pseudo-first order conditions: [CrIII ] ¼ (0.5–2)  102 M , [OH ] ¼ 0.1– III 4 2.0 M , [Fe ] ¼ 4  10 M , in a few cases also 3  104 and 5  104 M . Oxidation of chromium(III) was studied at 288–318 K at constant ionic strength 2.0 M (Kþ , OH /Br ). The time scale of the measurements was 6–3900 s ð4t1=2 Þ. Absorbance-time data at k ¼ 303 and 420 nm were fitted to the exponential form of the AðtÞ first order dependence. The relative standard errors of the pseudofirst order rate constants were 1–2%.

Results and discussion The starting complex, trans-[Cr(cyca)(H2 O)2 ]3þ , is a medium strength Brnsted acid, pK1 ¼ 2:4, pK2 ¼ 6:9 [Mønsted, unpublished data] and in strongly alkaline media (pH > 13) is practically completely converted into the trans-dihydroxo conjugate base without any observable decomposition or isomerisation, as has been established before [13]. Exceptional inertness of the complex in alkaline solutions makes it a very convenient reductant in studies on an electron transfer. Detailed X-ray structure studies were performed for [Cr(cyca)(H2 O)Cl](NO3 )2 and the trans geometry was established without any doubt. The coordinated Cl and H2 O ligands are trans with respect to the CrN4 plane and the five-membered rings adopt the gauche ðdkÞ and the six-membered rings the chair (pp) conformation [15]. The analogous structure is proposed for the diaqua complex [13]. In comparison with the cycb ligand the cyca one is much more rigid, however it fits very well the chromium(III) center. One can expect that oxidation of chromium(III) leading to a decrease of the ionic radius will cause weaker bonding of CrIV (CrV ) by the cyca than

Absorbance

E.p.r. measurements

1

Absorbance

08

09 07 05 270

275

280

285

290

λ ,nm

06 04 02 0 270

Fig. 1. Experimental (top) and simulated (bottom) e.p.r. spectra of chromium(V) complexes generated via oxidation of the trans[Cr(cyca)(OH)2 ]þ complex by the [Fe(CN)6 ]3 ion. Conditions: [CrIII ] ¼ 8  103 M , [FeIII ] ¼ 5  102 M , [OH ] ¼ 1  101 M , T ¼ 295 K, reaction time, 155 min; microwave frequency, 9.491 GHz.

295

320

345

370

395

420

445

470

λ , nm

Fig. 2. Spectral changes during oxidation of trans-[Cr(cyca)(OH)2 ]þ complex by [Fe(CN)6 ]3 ion; [CrIII ] ¼ 1  102 M , [FeIII ] ¼ 4  104 M , [OH ] ¼ 1:0 M , I ¼ 2:0 M , T ¼ 298 K, I=1 cm, scans every 10 s.

857

Fig. 3. Spectral changes during oxidation of trans-[Cr(cyca)(OH)2 ]þ complex by [Fe(CN)6 ]3 ion; [CrIII ] ¼ 1:2  102 M , [FeIII ] ¼ 6  102 M , [OH ] ¼ 0:1 M , I ¼ 2:0 M , T ¼ 296 K, I=1 cm, scans every 120 s.

by the cycb ligand because of the mentioned differences in flexibility of the macrocycles. Oxidation of the chromium(III) macrocyclic complex by an excess of hexacyanoferrate(III) is a very composite process leading to oxidation and decomposition of the macrocyclic ligand followed by CrO2 4 ion formation. Oxidation of the coordinated macrocyclic ligand was studied in detail before [12] for the cis isomer. Several steps were distinguished: (i) formation from 2 to 4 double bonds; (ii) incorporation of double bonded oxygen atoms in the macrocyclic ligand; (iii) the ligand decomposition. The overall reaction is a slow process proceeding in several hours time scale. Formation of a fairly inert chromium complex at the intermediate oxidation state of the central ion is observed in e.p.r. and electronic spectra, cf. Figures 1 and 3. Only part of the visible spectrum is available because of a strong [Fe(CN)6 ]3 ion absorbance at k < 500 nm. Figure 3 shows the systematic absorbance increase within 480–800 nm due to accumulation of chromium(V) with a very characteristic band formation at 739 nm followed by an absorbance decrease. Finally CrO2 ions are detected. This observation correlates 4 with an increase and then a decay of the e.p.r. signals which can be attributed to the d1 electronic configuration of the central ion of the intermediate. An observed complicated character of absorbance changes presented within the whole 480–800 nm spectral range, cf. Figure 3, results from the presence of many absorbing species with concentrations changing during the reaction course. Based on the data published very recently [11] for the series of chromium complexes at different oxidation states, the molar absorption coefficient for k  500 nm decreases in the following order: e500 (CrIV ) > e500 (CrV ) > e500 (CrIII ) > e500 (CrO2 4 ). Additionally, oxidation of the coordinated macrocyclic ligand and further chromium(V) decomposition to CrO2 4 anion makes absorbance changes very complex. The initial e.p.r. spectrum is multicomponent and complicated. However, during the course of the reaction only one main species is stabilized. It is probably due to

gradual interamolecular oxidation of the coordinated macrocycle found before. Similar behaviour was observed in the cis-[CrIII (cycb)(OH)2 ]þ – [Fe(CN)6 ]3 system [12]. There are also differences between trans[CrIII (cyca)(OH)2 ]þ and cis-[CrIII (cycb)(OH)2 ]þ oxidation processes and formation of chromium(V) species. The trans isomer reactivity in the redox process is higher and also Cr(V) complexes are less stabile. There is no steric hindrance, as in the cis isomer, and a six coordinated chromium(V) species at lower g-value [16] could also be formed (see below). At room temperature the main chromium(V) complex exhibits an isotropic e.p.r. signal due to chromium isotopes with I ¼ 0 (90.5%) at giso ¼ 1:987, cf. Figure 1. The signal is resolved into nine lines, because of electron-nuclei hyperfine interactions with four equivalent or nearly equivalent nitrogen ðI ¼ 1Þ 14 N nuclei, a(14 N) ¼ 2.7 G, with the expected intensity ratio being 1:5:15:30:45:30:15:5:1. Hyperfine structure reveals that the Fermi-contact interaction of four nitrogen nuclei occur to the same extent and correspond to a spin density of 0.0048 ¼ 2.7/557 on each nitrogen, where 557 G is the value for an isolated nitrogen atom [17]. A quartet satellite signal, due to hyperfine coupling by the less abundant (9.5%) 53 Cr isotope ðI ¼ 3=2Þ, is not observed because of low intensity. The e.p.r. parameters given above are very close to those of [CrV O(cycb)ox ]3þ [12] and perchloratooxochromium(V) porphyrins [18]. A somewhat higher giso -value for the studied main chromium(V) complex as compared with perchloratooxochromium(V) porphyrins is due to differences in energy of the d-d transitions. The e.p.r. data are consistent with [CrV O(cyca)ox ]3þ formula of the formed chromium(V). A minor chromium(V) complex, with the concentration of ca. 1/6 of the main chromium(V) complex, exhibits a broad isotropic e.p.r. signal at giso ¼ 1:970, cf. Figure 1. It appears during the course of reaction after ca. 1 h. The hyperfine structure is not resolved probably because of the additional hyperfine interaction with a sixth axial bromide anion [a(79 Br), a(81 Br) < 1 G]. Based on the presented results and the results of our previous work on the coordinated macrocycle oxidation [12], the following stages of the overall chromium(III) ! CrO2 4 transformation can be proposed: FeIII

trans-½CrIII ðcycaÞðOHÞ2 þ ! ½CrIV ðcycaÞO2þ slowly

ð1Þ

FeIII

½CrIV ðcycaÞO2þ ! ½CrV ðcycaÞO3þ fast

FeIII

! ½CrV ðcycaÞox O3þ

slowly

FeIII

!

very slowly

CrO2 4 þ cycaox ð2Þ

This reaction sequence is also consistent with the recently published results on the structure of the chromium(IV) and chromium(V) species obtained using X-ray absorption fine structure data for frozen aqueous

858 solutions [11]. It has been established that CrIV and CrV complexes are five coordinated. The slowest rate of the electron transfer is expected if it is accompanied by substantial structure changes, thus the CrIII ! CrIV and CrV ! CrVI are the slow steps. The presence of the oxoligand in the coordination sphere of CrIV is deduced from the kinetic results discussed below. Detailed kinetic studies were performed under conditions which allow one to observe only the first oxidation stage (Equation 1) – electron transfer from the CrIII center of the trans-[Cr(cyca)(OH)2 ]þ ion resulting in formation of the reactive chromium(IV) complex. Applying the large molar excess of the chromium(III) ion over the oxidant, we have eliminated all complications caused by subsequent reactions depicted in Equation (2). Absorbance-time data were processed within 270– 480 nm. A factor analysis of the data presented in Figure 2 proves that the iron(III) oxidant disappears in a well-defined first order manner consistent with the production of iron(II), without spectral interference from chromium complexes. The absorbance-time data for an excess of [CrIII ] and [OH ] over [FeIII ] fit very well a first order rate expression. The obtained pseudofirst order rate constants ðkobs Þ are practically independent of the chosen wavelength and of the initial concentration of the oxidant. Values of kobs at different temperatures, at different CrIII and OH concentrations are collected in Table 1. Some important dependences of the kobs are illustrated in Figures 4 and 5 based on the data shown in Table 1. The rate of iron(III) disappearance, which is almost equal to the rate of the chromium(IV) formation under applied conditions, is described by an overall second order rate expression – first order in [FeIII ] and in [CrIII ]: d½FeIII  ¼ kOH ½CrIII ½FeIII  dt

ð3Þ

6 1.2

3

Fig. 4. Dependence of the kobs on [CrIII ]; [FeIII ] ¼ 4  104 M , [OH ] ¼ 0:6 M , I ¼ 2:0 M , T ¼ 308 K.

Fig. 5. Dependence of the kobs on [OH ]; [CrIII ] ¼ 1:0  102 M ; [FeIII ] ¼ 4:0  104 M ; I ¼ 2:0 M .

where kOH  [CrIII ] ¼ kobs for a large excess of [Cr] as is seen from Figure 4. Figure 5 shows a higher order than linear dependence of the Pseudo-first order rate constant ðkobs Þ on [OH ]. An analogous dependence was found before [13] for base hydrolysis of [Cr(cyc)(OH)X]þ type complexes. The strong acceleration of reaction (1) by OH ions, cf. Figure 5, can be rationalized in terms of a rapid preequilibrium, which involves hydroxyl ion and

Table 1. The pseudo-first order rate constants for the CrIII fi CrIV step (errors are expressed as the standard deviations); [CrIII] = 1.0 Æ 10)2 M ; [FeIII] = 4.0 Æ 10)4 M ; I = 2.0 M (K+, OH) Br)) 102 Æ kobs, s)1 288 K

298 K

308 K

318 K

0.1 0.2 0.4

0.070 ± 0.005 0.110 ± 0.001 0.280 ± 0.001

0.190 ± 0.003 0.270 ± 0.005 0.63 ± 0.04

0.39 ± 0.02 0.63 ± 0.01 1.37 ± 0.01

0.82 ± 0.04 1.24 ± 0.02 2.74 ± 0.01

0.6

0.440 ± 0.003

1.01 ± 0.01

1.17 2.29 3.20 4.46 2.30 2.28

± ± ± ± ± ±

0.01a 0.02 0.02b 0.01c 0.01d 0.01e

4.63 ± 0.02

0.8 1.0 1.2 1.4 1.6 1.8 2.0

0.700 ± 0.003 0.99 ± 0.01 1.51 ± 0.01 2.04 ± 0.01 2.73 ± 0.01 3.76 ± 0.02 5.56 ± 0.02

1.60 2.20 3.18 4.39 6.18 8.39 11.4

3.50 4.52 7.37 9.35 13.5 18.9 24.6

± ± ± ± ± ± ±

0.01 0.03 0.03 0.04 0.1 0.1 0.2

6.86 10.0 14.6 20.2 25.9 34.9 47.5

[KOH], M

a

± ± ± ± ± ± ±

0.01 0.01 0.01 0.01 0.02 0.03 0.1

± ± ± ± ± ± ±

[CrIII] = 0.5 Æ 10)2 M ; b [CrIII] = 1.5 Æ 10)2 M ; c [CrIII] = 2.0 Æ 10)2 M ; d [FeIII] = 3.0 Æ 10)4 M ; e [FeIII] = 5.0 Æ 10)4 M .

0.1 0.03 0.1 0.2 0.1 0.3 0.2

859 

b½OH  III Table 2. Results of the least squares fitting of the kOH on [OH)] data to equation kOH ¼ 1þc½OH fi CrIV step; I = 2.0 M (K+,  for the Cr  OH)/Br))

T/K

b, M )1 s)1

10c

288 298 308 318

0.66 1.53 3.42 6.93

)3.82 )3.67 )3.64 )3.55

± ± ± ±

0.01 0.05 0.16 0.18

kcb, M )1s)1 ± ± ± ±

0.03 0.05 0.08 0.05

5.6 11.5 25.2 47.8

DH#/DH0, kJ Æ mol)1 DS#/DS0, J Æ K)1 Æ mol)1 kcb298, M )1s)1/K298

51.9 ± 1.3 )50.0 ± 4.0 12.0

K 0.236 0.266 0.272 0.290 4.9 ± 1.0 5.2 ± 3.2 0.257

The activation parameters calculated from the kcb and the standard enthalpy and entropy calculated from the K (Scheme 2).

the chromium(III) complex, before an electron transfer occurs. The trans-[CrIII (cyca)(OH)2 ]þ complex is a potential Brønsted acid in strongly alkaline media and the coordinated hydroxyl ligand or/and the amine nitrogen atoms can be deprotonated. This idea implies the decisive role of trans-[CrIII (cyca)(OH)2 ]þ deprotonation and a higher reactivity of the reactant in the form of a conjugate base in the redox process. The deprotonation inforces the p-antibonding character of the chromium(III) ‘t2g ’ electrons ( pseudo-Oh symmetry approximation) and increases the electron donor capability of the chromium(III) center. In the previously developed model [13] important is the assumption that the conjugate base formation is retarded by a competitive interaction between the trans-[CrIII (cyca)(OH)2 ]þ and Br ions. A simple stoichiometric reaction scheme for such a model is shown in Scheme 2,

b ¼

kcb K  kca 2

ð6bÞ

c ¼

K 1 2

ð6cÞ

The results of a least squares fitting of the kOH – [OH ] data to Equation (4) give close to zero values of the a parameter which fluctuates irregularly. A better fit – lower errors in the b and c parameters – was obtained for the limiting case of Equation (4) where a ¼ 0. It means, in terms of the presented model (Equation 6a) that the rate of the trans-[CrIII (cyca)(OH)2 ]þ oxidation is much slower than the oxidation rate for its conjugate base. Values of the b and c parameters and calculated from them values of the second order rate constant for the conjugate base of the reductant, kcb (Equation 6b) and the equilibrium constant K (Equation 6c) obtained at four temperatures are collected in Table 2. The discussed results are consistent with formation of the CrIV complex via an outer-sphere electron transfer from the conjugate base of the trans-[CrIII (cyca)(OH)2 ]þ according to Equation (7):

Scheme 2 where kca and kcb are the second order rate constants and K is the equilibrium constant for competitive outersphere complex formation between Br and OH . The conjugate base formed via proton transfer within the CrIII  OH ion pair is not explicitly shown in Scheme 2. Basing on the model presented in Scheme 2 a function operating with three independent parameters (Equations 4–6) describes the dependence of the second order rate constant (Equation 3) on the [OH ]: kOH ¼

kobs a þ b½OH  ¼ 1 þ c½OH  ½Cr

ð4Þ

which is equivalent to the rate expression 5: kOH ¼

kca cBr þ kcb KcOH cBr þ KcOH

where cBr þ cOH ¼ I, (I ¼ 2:0 M ) and a ¼ kca

ð5Þ I

being

ionic

strength ð6aÞ

½CrIII ðcycaÞOðOHÞ0 þ ½FeðCNÞ6 3 kcb

! ½CrIV ðcycaÞO2þ þ ½FeðCNÞ6 4 þ OH

ð7Þ

Reaction (7) is characterized by a negative entropy of activation and a relatively low enthalpy of activation, cf. Table 2. Previously [12], it was impossible to separate the kcb and the K from the b and the c parameters for the cis reductant because a close to – 0.5 value of the latter parameter gives a large error in K, even in the case of a small error in the c value. For that reason comparison of the reactivity of the isomers can be based only on the experimental a, b and c parameters. The following regularities are observed: (i) the b value for the trans reductor is substantially larger – over three times at 298 K; (ii) the reaction path via the conjugate acid is of kinetic importance only for the cis isomer ða > 0Þ; (iii) the rate of electron transfer increases faster with an increase of [OH ] for the cis reductant due to the more negative value of the c parameter. A faster electron transfer from the trans- than from the cis-chromium(III) complex can be correlated with a higher energy of the HOMO (dxy )

860 orbital of the former species. This conclusion is based on the assumption of pseudo-D4h local symmetry for both isomers, taking into account the average OAO0 , OAN and O0 AN donor atoms ligand field strength and locating the weaker ligand strength donors on the ‘Z’ axis.

Prof. F. Rozpłoch (Institute of Physics, N. Copernicus University) for his kind help in carrying out the e.p.r. measurements.

References Conclusions Electron transfer from the conjugate base of the chromium(III) complex to the [Fe(CN)6 ]3 ion is the slow rate limiting step in formation of a relatively inert chromium(V) species. The observed characteristic strong acceleration of the redox process with the increase of OH ion concentration results from competitive ion-pair formation between the positively charged chromium(III) reactants and the dominating anions of the solution. The same form of the kobs on [OH ] dependence is valid as well for the ligand substitution [13] as for the electron transfer process of the studied chromium(III) macrocyclic complexes. Higher reactivity of the trans- than the cis-chromium(III) isomer in the redox process contrasts with its higher inertness in ligand substitutions [13] and is probably due to a slower electron transfer from the conjugate base of the cis complex. The e.p.r. data are consistent with [CrV O(cycaox )]3þ formula of the most inert chromium(V) complex. The chromium(V) species formed via oxidation of the trans isomer are less inert than the analogous CrV complexes obtained from the cis isomer. It can be caused first of all by discussed differences in steric hindrances of the organic ligands. Further oxidation of CrV to CrO2 4 is much easier for the cyca complex because of weaker bonding of the central ion by the rigid cyca than the flexible cycb macrocyclic ligand. Acknowledgements The authors would like to express their graditude to Dr O. Mnsted ([rsted Institute of Copenhagen University) for the samples of the starting complex and to

1. M.L. Bishop, J.L. Duben-Engelkirk and E.P. Fody, Clinical Chemistry, 4th edit., Lippincott Williams and Wilkins, Philadelphia, 2000, p. 328 and 566. 2. J.J.R. Frausto da Silva and R.J.P. Williams, The Biological Chemistry of the Elements, 2nd edit., Oxford University Press, New York, 2001, p. 468. 3. R.M. Roat-Malone, Bioinorganic Chemistry, Wiley-Interscience, New Jersey, 2002, p. 278. 4. A. Levina, R. Codd, C.T. Dillon and P.A. Lay, Prog. Inorg. Chem., 51, 145 (2003). 5. M. Cies´ lak-Golonka, Wiad. Chem., 45, 403 (1991); G. Gałuszka and M. Cies´ lak-Golonka, ibid., 53, 379 (1999). 6. J.B. Vincent, Acc. Chem. Res., 33, 503 (2000). 7. D.F. Shriver and P.W. Atkins, Inorganic Chemistry, 3rd edit., Oxford University Press, New York, 1999, Appendix 2. 8. R. Codd, C.T. Dillon, A. Levina and P.A. Lay, Coord. Chem. Rev., 216–217, 537 (2001). 9. A. Levina, R. Codd, C.T. Dillon and P.A. Lay, Prog. Inorg. Chem., 51, 145 (2003). 10. A. Levina, A.M. Bailey, G. Champion and P.A. Lay, J. Am. Chem. Soc., 122, 6208 (2000). 11. A. Levina, R. Codd, G.J. Foran, T.W. Hambley, T. Maschmeyer, A.F. Masters and P.A. Lay, Inorg. Chem., 43, 1046 (2004). 12. J. Chatłas, O. Impert, A. Katafias, P. Kita, G. Wrzeszcz, J. Eriksen, O. Mnsted and A. Mills, Transition Met. Chem., in press, TMCH 5932. 13. E. Madej, O. Mnsted and P. Kita, J. Chem. Soc., Dalton Trans., 2361 (2002). 14. D.R. Duling, J. Magn. Reson., B104, 105 (1994). The software is available from the Internet (http://alfred.niehs.nih.gov/pest.html). 15. R. Temple, D.A. House and W.T. Robinson, Acta Cryst., C40, 1789 (1984). 16. R.P. Farrell and P.A. Lay, Comm. Inorg. Chem., 13, 135 (1992). 17. B.A. Goodman and J.B. Raynor, Adv. Inorg. Chem. Radiochem., 13, 135 (1970). 18. H. Fujii, T. Yoshimura and H. Kamada, Inorg. Chem., 36, 1122 (1997).

TMCH 5960