Model-Driven Approaches


It was pointed out in Chapter 2 that the general theory of pressure driven flow in a capillary, throughout the intermediate range where mean free path length and tube diameter are comparable, has so far proved intractable. For just the same reasons we have no complete theory of thermal transpiration in a capillary to span the range of cube diameters between Knudsen streaming and Maxwell s calculation. However, it is possible to approach this problem through the dusty gas model, which once again demonstrates its power by predicting both thermal transpiration and thermal diffusion, throughout the Intermediate range of pore sizes, and for mixtures of any number of species rather than just a pure gas. Furthermore, as we shall see, the results reduce to equations (A.1.2) and (A.1.8) above in the simple limiting cases. A good account of this aspect of the dusty gas model has been given by Wong and Denny [68].  [c.182]

DSI is discussed in Part C (Chapter 17), since the approach usually requires an initial evaporation of solvent from a solution by moderate heating in a gas stream so as to leave the solute (the analytical sample). The resulting residual sample is then heated strongly to vaporize it. Typically, a solution is placed onto a heat-resistant wire or onto a graphite probe, and then the solvent is allowed to evaporate or is encouraged to do so by application of heat, directly or indirectly. The residual solid on its metal or graphite support is placed just below the plasma flame, which is allowed to stabilize for a short time. The probe and sample are then driven into the high-temperature flame, which causes vaporization, fragmentation, and ionization (Figure 16.2). Because the heat capacity of the flame is relatively small, the sample holder and sample should have as low a thermal mass as possible so as not to interfere with the operation of the flame. With the direct-insertion method, samples appear transiently in the flame therefore, if a wide range of elements is to be examined, the mass spectrometer should be one that can span a wide m/z range in the short space of time the sample takes to pass through the flame (quadrupole, time-of-flight). Further details of the DSI technique are discussed in Part C (Chapter 17).  [c.105]


See pages that mention the term Model-Driven Approaches : [c.439]   
See chapters in:

Chemoinformatics  -> Model-Driven Approaches