Abstract
Recent considerations by Wada make it appropriate to extend a previous discussion of Grüneisen ``numbers'' for polymers and other molecular solids. Wada hypothesized that the proper Grüneisen constant for polymers is γG=αV/βCvb , where Cvb is the interchain contribution to the heat capacity at constant volume; α, β, and V are the volumetric thermal expansivity, the compressibility, and the volume. To the extent that a polymer can be treated as a vibrational lattice, the hypothesis appears to be consistent with γG=Σγiε(xi)/Σε(xi) , averaged over Einstein oscillator functions ε(xi) with xi=hvi/kT. At low temperatures, ε(xi) is much larger for the low‐frequency modes, so that they tend to determine γG below the Debye −θ. Since γi=−∂ lnνi/∂ lnV and since low νi are likely to be more sensitive to changes in V, it is expected that γG will be larger for molecular solids than for metallic, ionic, or covalent crystals. Earlier predictions and Wada's calculations agree that γG≈4 might be typical for polymers and suggest that ∂γ/∂T>0. The correlation, Eαl2≈15 N/m2°K2, between modulus E and linear expansivity αl led to the prediction, now verified, that there should be relations between the harmonic and anharmonic moduli. The anharmonic coefficients in the relation ΔV/V0=a1p+a2p2+a3p3+⋯are a2=C1a12 and a3=C3a13, where for metals C2=−2.5±0.5, and for polymers C2=−4.0±0.1 and C3=8.8±0.2. A phenomenological theory based on a ``bundle of tubes'' model is developed which is in good agreement with data and according to which γG=−C2 and G/dT∼αC2 . The relation of γG to intermolecular potential functions also is discussed and some qualtitatively encouraging results are obtained.