Artefacts large

In this chapter we have largely relied on computational chemistry, in particular on density-functional theory. Quantum mechanical calculations of a macroscopic piece of metal with various species adsorbed on it are as yet impossible, but it is possible to obtain realistic results on simplified systems. One approach is to simulate the metal by a cluster of 3-30 atoms on which the molecule adsorbs and then describe all the involved orbitals. Many calculations have been performed on this basis with many useful results. Obviously, the cluster must be sufficiently large that the results do not represent an artefact of the particular cluster size chosen, which can be verified by varying the cluster size. 

If thick samples are placed in the specimen chamber for analysis, the particles are slowed down and eventually stopped in the sample, so the calculation of the X-ray yield and their absorption is more complicated. Some objects may be too large to be placed in the specimen chamber, in which case the external beam technique is employed. The particle beam passes through a window at the end of the beam-line into the air where an object of any size (e.g. an archaeological artefact) may be analyzed. 

For the analysis of large objects which cannot be placed within the irradiation chamber it is possible take the particle beam into the ambient air through a thin window at the end of the beam line. In this way any type of object can be analysed -for example paintings and archaeological artefacts. 

Simultaneous fits of a series of related complexes, as mentioned above. In this case, a handful of parameters can be used to reproduce the experimental behaviour of a large quantity of compounds. This reduces the possibilities of numerical artefacts, and also the propagation of experimental errors into the final parameters. 

It should first be noted that the measurement of emission anisotropy is difficult, and instrumental artefacts such as large cone angles of the incident and/or observation beams, imperfect or misaligned polarizers, re-absorption of fluorescence, optical rotation, birefringence, etc., might be partly responsible for the difference between the fundamental and limiting anisotropies. 

Fig. 3. (a) 300 MHz proton spectrum of the molecule shown, and (b) 500-fold expanded NOE difference spectrum obtained using reference deconvolution to enforce identical reference lineshapes (2 Hz Lorentzian) for the chloroform signals in the irradiated and unirradiated spectra. Note the complete absence of the usual up-down difference artefacts in trace (b), even for the sharp singlet methyl signals the remaining artefacts are largely homodecoupler... 

---