The high-resolution solid-state 13C NMR (nuclear magnetic resonance) spectra are reported of the iron carbonyl complexes Fe3(CO)8(PhCCPh)2(I) and Fe3(CO)12(II), 13C-enriched at the carbonyl carbons. The spectra were recorded with slow magic-angle spinning, and the intensities of the spinning side bands analysed by a computerized adaptation of the method of Herzfeld & Berger to give values for the principal components of the carbonyl carbon shielding tensors. Complex (I) includes two asymmetric double-bridging carbonyl groups, and the 13C shielding-tensor components assigned to the C-O bond axis (sigma33) are intermediate between the values obtained for symmetric double-bridging carbonyl groups and linear M-CO terminal carbonyls. It is proposed that the sigma33 shielding component may therefore be used to monitor the degree of asymmetric bridging of a carbonyl group. All 13C resonances of complex (II) exhibit a degree of asymmetry in the shielding tensor and possible reasons for this are discussed. Three methods of estimating errors in the derived shielding-tensor components are considered. One method uses calculated variances based upon differences between the experimental and calculated side-band intensities; this gives unreasonably large errors in the cases of axial symmetry in the shielding tensor. A second method explores different shielding components for values of Phi2 (the sum of the squares of the differences between the experimental and calculated side-band intensity ratios) which exceed the best minimum by an arbitrary percentage. The third method considers errors due to the signal:noise ratios in the experimental spectra. The relative merits of these three approaches of error estimation are examined.
Solid-State 13C NMR Spectra of Iron Carbonyl Complexes: Shielding Tensor and Error Analyses
GOBETTO, Roberto
1989-01-01
Abstract
The high-resolution solid-state 13C NMR (nuclear magnetic resonance) spectra are reported of the iron carbonyl complexes Fe3(CO)8(PhCCPh)2(I) and Fe3(CO)12(II), 13C-enriched at the carbonyl carbons. The spectra were recorded with slow magic-angle spinning, and the intensities of the spinning side bands analysed by a computerized adaptation of the method of Herzfeld & Berger to give values for the principal components of the carbonyl carbon shielding tensors. Complex (I) includes two asymmetric double-bridging carbonyl groups, and the 13C shielding-tensor components assigned to the C-O bond axis (sigma33) are intermediate between the values obtained for symmetric double-bridging carbonyl groups and linear M-CO terminal carbonyls. It is proposed that the sigma33 shielding component may therefore be used to monitor the degree of asymmetric bridging of a carbonyl group. All 13C resonances of complex (II) exhibit a degree of asymmetry in the shielding tensor and possible reasons for this are discussed. Three methods of estimating errors in the derived shielding-tensor components are considered. One method uses calculated variances based upon differences between the experimental and calculated side-band intensities; this gives unreasonably large errors in the cases of axial symmetry in the shielding tensor. A second method explores different shielding components for values of Phi2 (the sum of the squares of the differences between the experimental and calculated side-band intensity ratios) which exceed the best minimum by an arbitrary percentage. The third method considers errors due to the signal:noise ratios in the experimental spectra. The relative merits of these three approaches of error estimation are examined.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.