Structure insensitivity experimental examples

HSEA method has been found to generate conformations that are frequently consistent with the NMR dataA Early HSEA-based predictions of the conformations of the blood group antigenic determinants remain examples of the success of the HSEA method when applied to molecules that appear to exist predominantly in one conformation in solution. In contrast, based on somewhat limited data from nuclear Overhauser effect (NOE) intensities, HSEA calculations predicted that sucrose preferred a single rigid conformation in solution. More recent experimental data demonstrated that sucrose is more flexible than had been believed. This example clearly illustrates the pitfalls associated with fitting structures to insufficient and insensitive experimental data. 

In a catalytic reaction, all steps do not equally depend on the surface structure. So, for example, the rate of simple desorption processes is often not markedly affected by the structure of the surface. In catalysis, therefore, reactions are classified into "structure sensitive" and "structure insensitive" [5], usually on the basis of the variation of reactivity with particle size. As an example, the electrocatalytic oxygen reduction at platinum (which is of importance for fuel cells) will be mentioned, where a decrease of specific activity with increasing particle size was reported [6,7]. In a theoretical analysis [8], the kinetics was treated on the (111), (10 0), and (211) facets of several transition metals, and the results were combined with simple models for the geometries of catalytic nanoparticles. Thus, the experimentally observed trend could be well reproduced. 

Planar magnetic dipole arrays exhibit a different type of structural stability. Experimental observations and computer checks show that the domain patterns generated by pivoted dipoles are insensitive to variations of the individual magnetic moments and perturbations of the underlying lattices. The net result is that the domain structures are robust under changes of scale, but vulnerable to qualitative shifts in the strength of multipolarities. These examples indicate that various levels of structural stability and instability can coexist in complex systems. 

This can be illustrated by a pair of hypothetical examples [22]. Suppose that we have four observations of a covalent bond length let them have e.s.d. s of 0.002, 0.008, 0.009 and 0.010 A, respectively, and assume, for the moment, that these e.s.d. s are reliable. Now, most bond lengths are hard parameters, i.e. considerable energy is required to distort them from their equilibrium values. They are therefore relatively insensitive to changes in the crystal-field environment. Thus, it is likely that the true value of the /th bond length (i.e. its value in the /th crystal structure, if we could measure it without experimental error u, in Equation 4.1), is virtually identical to that of the yth. Differences between the //, may therefore be assumed negligible compared with experimental errors, and (4.1) may be simplified to ... 

As was discussed, infrared and raman spectra for organometallic systems can typically be computed to within 5% of the experiment. Unlike adsorption energy predictions, structure and vibrational frequencies are fairly insensitive to differences in the DFT methods (local vs. nonlocal spin density). Even some of the earliest reported local-spin-density approximation (LDA) DFT calculations which ignored adsorbate and surface relaxation predicted frequencies to within 10 percent of the measured values. For example, Ushio et al. have shown that LDA calculations for formate on small Nia clusters (frozen at its bulk atomic positions) provide very good agreement with experimental HREELS studies on Ni(lll) [72]. Unlike adsorption energy predictions, structure and vibrational frequencies are fairly insensitive to gradient-corrections. 

---