X-ray crystallographic...... acetate adducts

Abstract

Distances seem to be evident from the plot. With the results of the magnetic work, the above conclusion can now be examined. The parameter most appropriate for comparison is the magnitude of 2J. If the interaction is direct in nature (through either a ?-bond or a ?-bond), a closer approach of the two copper atoms would lead to an enhanced interaction (larger 2J value). In other words, a shorter Cu–Cu distance would be associated with a larger 2J value.

The variable temperature susceptibility data are available for all the adducts except those of pyrazine and urea. The urea adduct was prepared according to the method described in the literature, and the measurements were made in the range 150–500 K by the Faraday method described in Chapter II. The corrected molar susceptibilities are given in Table 6.2. The values for g are 2.16 and 60 × 10?? c.g.s.e.m.u, respectively, and the 2J value was obtained by the best-fit method (Chapter II). The agreement between the measured and calculated susceptibilities is shown in Fig. 6.2. The room temperature magnetic moment, 1.39 B.M., obtained here agrees exactly with the earlier value. Measurements on the pyrazine adduct could not be made since the ligand could not be obtained.

For the other adducts whose susceptibility data have been reported, the 2J values have been recalculated by the best-fit method to provide a uniform treatment. The recalculation is considered necessary for the following reasons. As mentioned in Chapter II, different methods have been used by other investigators for calculating the 2J value from the susceptibility data. Secondly, in the absence of ESR data, the g-values employed are as high as 2.30. A survey of g-values over a large number of dimeric copper(II) carboxylates revealed that the values lie in the range 2.16 ± 0.03. Accordingly, in the present recalculation, the g-value, 2.16, was assumed whenever the ESR data were not available. The recalculated 2J values, along with the published ones, are given in Table 6.3.

A plot of Cu–Cu distance vs the |2J| values is shown in Fig. 6.3. The data appear to fall into certain groupings according to the axial ligands. The points corresponding to the adducts of acetic acid, methanol, DMP, and water (numbers 2, 4, 5, and 6, respectively) form one group (Group I), while those of urea, pyridine, TEDA, and thiocyanate (numbers 7, 8, 9, and 10, respectively) form another group (Group II). The points corresponding to the adducts of quinoline and ?-picoline are closer to Group II and hence may be considered along with them.

The first group contains only oxygen donor ligands, while the second one consists of nitrogen donors, with the exception of urea. One would expect urea to bind similarly to DMP and water, but experimentally it binds at a shorter distance than either of them. The stronger coordination of urea is perhaps due to the greater negative charge on the carbonyl oxygen, resulting from the mesomeric effect of the two amide nitrogens. But for urea, the grouping appears to be quite logical, since oxygen and nitrogen will have differing tendencies for coordination to copper.

When the ligands are considered separately in this manner, it appears that there is a tendency for 2J to decrease as the Cu–Cu distance increases. This indicates that the interaction is direct in nature. It should, however, be pointed out that the order of differences in the 2J values of nitrogen donors is rather small. In the case of oxygen donors, however, the trend is clearly seen. More data are hence needed to strengthen the conclusion.

Considering both Cu–Cu and Cu–L distances, and also keeping in mind that the differences in those distances are not always statistically significant, the ligands could be arranged in the following order of coordinating ability:

Group I: H?O < CH?OH < DMP < CH?COOH One would normally expect DMP to follow water, but the experimentally observed position, lower in the series, is perhaps due to the comparatively bulky nature of DMP.

Group II: NCS < pyridine < urea < TEDA This order is obtained as expected.

The correlation between Cu–Cu and Cu–L distances was previously taken to indicate the predominance of ?-type interaction. This conclusion needs careful consideration, especially in view of the support for the ?-bonded model from the ESR studies. If ?-bond is regarded as a ?-type of interaction, then the above results can also be interpreted in terms of the trans influence of the axial ligand, taking into account both the ?- and ?-interactive effects.

In the case of nitrogen donor ligands, metal-to-ligand ?-bonding is possible in addition to ?-bonding, since they possess suitable vacant delocalized orbitals for M–L back bonding. If L is a strong ?-accepting ligand, then it will compete with the metal–metal bonding for the d-orbital density and thereby weaken the M–M bond if the latter is of ?-type. The net effect would again be an inverse relation between Cu–Cu and Cu–L distances.

On the other hand, in the case of oxygen donor ligands, only ?-bonding with the metal is likely, and hence the trans effect can almost exclusively be regarded as due to the ?-bonding ability of the axial ligand. Hence one would expect that the oxygen donors would not show the inverse relation (between Cu–Cu and Cu–L) in a ?-bonded model. However, from the available structural results, it appears that both oxygen and nitrogen donors show similar variation in Cu–Cu and Cu–L, thus tempting one to conclude in favor of a ?-type of interaction. Further, it may be pointed out here that the end-on overlap of the orbitals will be greater than the lateral overlap of the d²?² orbitals. In view of this, at an internuclear separation of ca. 2.60 Å found in copper acetate adducts, the ?-overlap is likely to be more prominent than the ?-overlap.

As mentioned in Chapter I, apart from direct interaction, superexchange through the bridging ?-system of the carboxylate group has also been proposed for copper acetate. However, if the superexchange process plays any dominant role, it is unlikely that there would be any trend in the variation of one parameter with respect to another, especially the variation of Cu–Cu with respect to Cu–L observed here. Further discussions utilizing the role of superexchange are not possible by the approach adopted here.

In conclusion, the above discussion shows that in copper acetate adducts, an unequivocal interpretation of the experimental results is quite difficult, and a certain degree of intuitive reasoning becomes necessary at some stage or the other. It is perhaps for this reason that the conclusions drawn by one group of workers are often at variance with those of another. Although a few investigators may make it appear that the question has been settled, the author believes that more work is needed, and the field will continue to be a fertile one in the years to come