A friend asks me this. If considering the equation: [itex]∫\frac{dQ}{T}[/itex], then it is technically feasible to work out some forms of expressions with measurable physical quantities like temperature and specific heat, therefore it is possible to work out a precise value for entropy change. But is there a more economic way? I think Claussius entropy is too phenomenological to be directly observed in experiments, and the Boltzmann definition is not suitable for experiments. While above is about the entropy change, my friend also asks how to determine the entropy of a system, for example a tank of CO2. If a perfect crystal has zero entropy, does that meran in order to calculate the entropy we have to construct possible quasi-static processes from perfect crystals to the present compound and work out the entropy change, which seems to be very uneconomic?
There is a good description in a book by Thess - search for "Thermodynamics for the unsatisfied" at amazon.com, and check "search inside this book". In the table of contents, click on page 85, "determination of the entropy of simple systems".