Experimental background and applications

After illustrating the rather fascinating structural and rheological properties of confined fluids we conclude our discussion of MC simulations of continuous model systems (i.e., models in which fluid molecules move along continuous trajectories in space) with yet another example of the imique behavior of confined fluids. For pedagogic reasons we selected a topic that is standard in physical chemistry textbooks [26, 199-203] as far as bulk fluids are concerned, namely the Joule-Thomson effect. 

During the expansion the gas does not exchange heat with its environment. However, it exchanges work because of the expansion against the nonzero pressure P2. It is then a simple matter to demoiLstrate that the gas expands isenthalpically [26, 199-203]. This makes it convenient to discuss the Joule-Thomson process quantitatively in terms of a Joule-Thomson coefficient 

From a fundamental perspective the Joulc -Thotnson effect is important because it can be linked directly to the nature of intermolecular forces between gas molecules [205]. Consider, for example, a classic ideal gas as the simplest case in which molecules do not interact by definition. For this model it is simple to show that as a consequence of the absence of any intermolecular interactions a Joule-Thomson effect does not exist, that is, 5 = 0 [200, 201]. If, on the other hand, the ideal gas is treated quantum mechanically, it can be demonstrated [206] that a Joule-Thomson effect exists (S 0) despite the lack of intermolecular interactions. 

The origin of the nonvauishing Joule-Thomson effect is the effective repulsive (Fermions) and attractive (Bosons) potential exerted on the gas molecules, which arises from the different ways in which quantum states can be occupied in sy.stems obeying Fermi-Dirac and Boso-Einstein statistics, respectively [17]. In other words, the effective fields are a consequence of whether Pauli s antisymmetry principle, which is relativistic in nature [207], is applicable. Thus, a weakly degenerate Fermi gas will always heat up ((5 0), whereas a weakly degenerate Bose gas will cool down (5 0) during a Joule-Thomson expansion. These conclusions remain valid even if the ideal quantum gas is treated relativistically, which is required to understand 

Unfortunately, previous work is almost exclusively concerned with the inversion temperature in the limit of vanishing gas density, Ti y (0). The inversion temperature can be linked to the second virial coefficient, which can be measured [210] or computed from rigorous statistical physical expressions [211] with moderate effort. Currently, only the fairly recent study of Heyes and Llaguno is concerned with the density dependence of the inversion temperature from a molecular (i.c., statistical physical) perspective [212]. These authors compute the inversion temperature from isothermal isobaric molecular dynamics simulations of the LJ (12,6) fluid over a wide range of densities and analyze their results through various equations of state. 

---