We report on elastic neutron scattering results showing that the intermediate phase of quartz existing in a 1.3 K range between the α and β phases is in fact incommensurate. Six satellites are observed around most Bragg peaks along the three [100] directions of reciprocal space. The modulation corresponds to the superposition of sinusoidal plane waves as no higher order harmonics are observed. With decreasing temperature the modulation vector decreases from 0.033 to 0.029 reciprocal units. Premonitory diffuse scattering is observed in the β phase a few degrees above Ti = 847.2 K, the incommensurate transition temperature. The results are discussed in relation with the current theories of incommensurate phases and are in many respect analogous to the situation of NaNO2

No abstract

2014 Des mesures détaillées de dilatation thermique, de module d'élasticité et de biréfringence, en phase 03B2, près de la transition (1er ordre) 03B1-03B2 du quartz révèlent une anomalie qui suggère l'existence d'une transition du 2e ordre à une température légèrement supérieure à celle de la transition 03B1-03B2. Abstract. 2014 Detailed measurements of thermal expansion, elastic compliance and birefringence in the 03B2 phase of quartz around the 03B1-03B2 first order transition show an anomaly which suggests the existence of a second order transition slightly above the usual 03B1-03B2 transition.

%e discuss the concept of an order parameter q for the a-P transition of quartz following Landau's theory. Some physical properties, among them the second-harmonic-generation coeAicient are found to vary as q, while others vary as g'. A measurement of d» from room temperature to the transition is described. Assuming that d» -q one obtains the temperature variation of the order parameter; our data are in good agreement with Landau's theory. They also agree with the temperature variation of other properties, particularly birefringence, thermal dilatation, and some elastic constants. Our data can also be fitted to a formula q = q~+ C{tt*)~, with a critical exponent P = 1/3 from room temperature to the transition, or with P = 1/6 and q* = 0 from 520 to 573. 2'C the transition temperature. W'e show that this is not fortuitous but that one can always find these exponents in making a four-or three-parameter fit to Landau's formula. Thus, in general, there is considerable ambiguity when one tries to fit critical exponents to a first-order transition.