Lanthanide induced shift NMR experiments : a computational model study

The results of the lanthanide induced shift (LIS) experiment to determine lanthanide-substrate (Ln-S) array structures is viewed by two different comparisons: external and internal. The external comparison is assessed only by precision and is therefore usually quite straightforward. In addition to being easily assessed, the external comparison does not require any arbitrary decision about the "true" nature of the lanthanide-substrate array. Unfortunately, the internal comparison is not as simple as the external comparison because of the inclusion of the all elusive concept of accuracy. By the very nature of accuracy, the "true" Ln-S array must be determined by some method external to the LIS experiment. However, no consistent determination exists and the definition of an accurate Ln-S array is then left to the individual experimenter. The manifold decisions about the Ln-S array are manifested by different computational models. The theoretical considerations and the relative precision involved in these different computational models are examined. As was previously indicated, there is no statement made as to the accuracy involved in these computational models. However, the method of least adjustable parameters is advocated because it produces Ln-S arrays of comparable precision and is applicable to a large number of substrates. Finally, some specific applications of the method of least adjustable parameters are presented as well as some reporting recommendations so that the complex internal comparison can be facilitated.