We have performed small-angle Rayleigh light-scattering measurements in polymer solutions under various externally applied temperature gradients. Our experiments confirm the existence of an enhancement of the concentration fluctuations. The nonequilibrium concentration fluctuations are proportional to ͑=T͒ 2 ͞k 4 , where =T is the applied temperature gradient, and k is the wave number of the fluctuations. The measured strengths of the nonequilibrium fluctuations in the dilute and semidilute concentration regime agree with the strengths calculated from fluctuating hydrodynamics.[ S0031-9007(98) PACS numbers: 61.25. Hq, 05.70.Ln, 66.10.Cb During the past decade, considerable theoretical and experimental effort has been devoted to the study of fluctuations in fluids in nonequilibrium steady states, in particular, when subjected to a constant temperature gradient =T . It is now known that the correlation functions become spatially long ranged [1][2][3], because microscopic detailed balance is violated. These nonequilibrium correlation functions can be observed by small-angle Rayleigh light scattering. Experiments have been carried out with simple fluids, in which case excellent agreement with the theory has been found [4,5].In fluid mixtures, the situation is more complicated because the temperature gradient induces a concentration gradient through the Soret effect. Hence, in addition to nonequilibrium entropy and viscous fluctuations, which are also present in one-component fluids, nonequilibrium concentration fluctuations are present in mixtures [6]. All of these nonequilibrium fluctuations have intensities proportional to ͑=T ͒ 2 ͞k 4 , where k is the wave number of the fluctuations. The existence of these nonequilibrium fluctuations and their dependence on ͑=T ͒ 2 ͞k 4 have been checked experimentally, but the quantitative values of the observed strengths of the nonequilibrium concentration fluctuations were not in agreement with the theory [5,7,8].Dilute polymer solutions appear to be a more suitable system to check the theory for nonequilibrium concentration fluctuations, because the expression for the dynamic correlation function C͑k, t͒ as a function of time is simpler. When the polymer concentration is below the overlap concentration, viscoelastic and entanglement effects are negligible, and ordinary fluctuating hydrodynamics should be applicable to describe the nonequilibrium fluctuations. Furthermore, the collective mass-diffusion coefficient is so small and the Rayleigh ratio so large that, in practice, the concentration fluctuations are dominant. Therefore, for polymer solutions at sufficiently low concentrations, the equations in Refs. [5,6] can be simplified, yielding the following for fluctuations with a wave vector perpendicular to =T :where C 0 is the ratio of the signal to the local oscillator background, which corresponds to the amplitude of the experimental equilibrium correlation function; D is the collective mass-diffusion coefficient, and A c is the nonequilibrium enhancement. A c may be ...

We have repeated and extended turbidity measurements on a new sample of the ionic mixture triethyl n-hexyl ammonium triethyl n-hexyl boride (N2226B2226) in diphenyl ether in a temperature range above its consolute point. Previous measurements on this system by some of us indicated mean-field critical behavior over the range of reduced temperatures 10−4<t<10−1. Our new measurements do not reproduce those reported before. The transmission of the new sample is substantially lower, and the turbidities up to a factor of 3 higher than those of the sample used in the previous work. The correlation-length amplitude is approximately 40% larger than that found for the previous sample. Unlike the earlier results, the new data do not display a substantial range of mean-field critical behavior. In our investigation of this irreproducibility, we have carefully reviewed the earlier experiment, and we report here an effect not previously noted: a time dependence of the turbidity of the earlier sample. It is likely that the discrepancy is due to impurity effects, but we do not understand their origin.