Abstract-The initial assignments by Mahler and Pettit of relative configurations of the two racemic diastereomeric tricarbonyl-(rrans,rrans-3,5-heptadien-2-ol)iron complexes and to the two raceniic diastereomeric tricarbonyl-(~runs-3,5-hexadien-2-ol)iron complexes are consistent with NMR coupling constants, conformational analysis, previously reported Eu (fod)8 lanthanide-induced shift data and new Yb(dpm), lanthanide-induced shift data. A detailed discussion shows that, contrary to the claim by Foreman, the lanthanide-induced shift data alone provide no new information regarding the relative configurations of these alcohols.ASSIGNMENT of the configurations RR or SS and RS or SR to the two racemic diastereomeric tricarbonyl-(trans,trans-3,5-heptadien-2-ol)iron complexes (1 and 2) as well as to the two diastereomeric tricarbonyl-(trans-3,5-hexadien-2-ol)iron complexes (3 and 4) was made from mechanistic arguments by Mahler and Pettit in 1963.' The relative stereochemistries of these four compounds and others were confirmed by Lillya and Clinton, using first a detailed conformational analysis of the diastereomeric pairsza and then a careful interpretation of the rates of solvolysis of the 3,Sdinitrobenzoate esters of these diastereomers.2b The selfconsistency of the Lillya-Clinton and Pettit-Mahler reports left little doubt about the correctness of the stereochemical descriptions. Subsequently, Foreman recognized that the lanthanide-induced NMR proton shifts (LIS) provided another possible avenue for the assessment of the relative stereochemistries of compounds 3 and 4., Foreman concluded from LIS spectra of the two tricarbonyl-(trans-3,5-hexadien-2-ol)iron complexes (3 and 4) that the previously accepted stereochemistries 'may perhaps not be correct.' The tentative Foreman correction has been mentioned in several reviews4 and is cited as a reasonable use of the LIS technique. The Pettit-Mahler assignment of the diastereomeric constitution of compound 1 has now been verified by a single crystal X-ray diffraction study.5 We have reexamined the Foreman LIS account, adduced new LIS * Author to whom correspondence should be addressed.
557Printed in Northern Ireland. data for compounds 1 and 2, and conclude that interpretation of the LIS data alone does not allow assignment either of configuration or of conformation for theseWe report the details of our experience with the LIS interpretation here.Foreman accepted the Lillya arguments about conformational analysis and reduced the strategy for assigning the descriptors R R and RS to determining whether the alcohol group was y-ex0 or y-endo in compounds 3 and 4. The y-exo is that which features a dihedral angle between the -OH group and the butadiene residue of go", and a trans disposition of the alcohol and the Fe(CO), group. The y-endo conformation'" has the aliphatic group rotated c. 120" counterclockwise about C-2-C-3.Both conformations are illustrated below. Foreman incorporated the LIS data by fixing an array of atoms including the lanthanide ion, and then assessing ...
