Microscopic structures and dynamics of high- and low-density liquidtrans-1,2-dichloroethylene

Abstract

We present a study of the dynamics and structural changes for trans-1,2-dichloroethylene between high-and low-density liquids using neutron-scattering techniques ͑diffraction, small-angle neutron scattering, and time of flight spectroscopy͒ and molecular-dynamics simulations. We show that changes in the short-range ordering of molecules goes along with a change in the molecular dynamics: both structure and dynamics of the highdensity liquid are more cooperative than those of the low-density liquid. The microscopic mechanism underlying the cooperative motions in the high-density liquid has been found to be related to the backscattering of molecules due to a strong correlation of molecular ordering. Classical thermodynamics establishes the existence of one unique liquid state for any material, i.e., there is only a liquid phase characterized by its density given the thermodybamic coordinates pressure and temperature. Nevertheless, recent experimental results and molecular-dynamics ͑MD͒ simulations suggest that, even for one-component systems, several liquid phases can appear with an associated liquid-liquid phase transition ͑LLPT͒. A noticeable number of cases has been found for atomic liquids, the best-known example concerning liquid phosphorus, 1,2 where the LLPT appears as a transition between thermodynamically stable phases with strong structural changes. 3 As far as molecular liquids are concerned, the number of experimental evidences for LLPT is still rather scarce and comprises only a limited number of compounds such as triphenyl phosphite 4 and n-butanol. 5 According to the so-called two order parameter theories that propose an explanation for the LLPT, liquids must be described not only by their density but also by an additional order parameter accounting for changes in the molecular arrangement. [6][7][8] The LLPT can end in a liquid-liquid critical point between a high-density liquid ͑HDL͒ and a low-density liquid ͑LDL͒. However, changes in the dynamics with an associated change in structural features can also be explained by a singularity-free scenario. 9 In the latter case, changes in both dynamics and structure from a HDL to a LDL also take place at the point where the isobaric heat capacity C P has a maximum, but no critical point or LLPT are observed at nonzero temperature.An early work on trans-1,2-dichloroethylene ͑T melt = 223 K͒ suggested the existence of a LLPT at T t = 247 K Ͼ T melt based on a small jump in density ͑less than 0.06%͒ as well as in the compressibility and a clear discontinuity on the spin-lattice relaxation time T 1 . 10,11 The observed changes in T 1 were tentatively related to a lack of freedom of the molecular rotation in the HDL, not present in the LDL. The change in the dynamics of this substance between both liquids was thereafter also supported by discontinuities in the viscosity measurements and the slope of the rotational relaxation time 12 and by the absorbance, frequency, and linewidth of several infrared vibrational spectroscopy bands. 13 Concerning structural rela...

The liquid structure of tetrachloroethene: Molecular dynamics simulations and reverse Monte Carlo modeling with interatomic potentials

“…The first coordination shell itself has a complex structure: a shoulder appears that represents molecules closest to each other, around 3.9-4.1 Å, with a distinctive preference for parallel molecular axes and planes. Similar preference for parallel molecular axes and planes in case of trans-1,2-dichloro-ethylene was found in the work of Rovira-Esteva 47 .…”

Fitting in a complexχ2landscape using an optimized hypersurface sampling

“…More complex problems have already successfully been studied with this algorithm such as model selection using Quasielastic Neutron Scattering data [25], non-functional fits in the case of dielectric spectroscopy [26] or finding the molecular structure from diffraction data with a model defined by as many as 27 parameters [27]. In the last case, the proper initialization of parameters to use a LM algorithm would have been a difficult task, made easy by the use of the presented algorithm.…”

Characterizing ordering in liquids: An information theoretic approach

“…In order to calculate the MCN we have ordered neighbour molecules by their distance to the central molecule and then we have numbered them to calculate it. This has two main advantages: first it eliminates trivial effects of density when comparing a liquid at different temperatures [28]. On the other hand, the structure of a liquid changes fast at short distances and slower for molecules far away from the central molecule.…”

“…Although it is not the goal of this work, we will now shortly describe a general method to study the details of molecular arranging in a liquid for the sake of clarity of the later discussion. This general method is suitable to deal with small molecules such as Carbon Tetrachloride as well as with molecules as large as proteins [26] (more details can be found in our previous works [15,27,28]). For doing that it is necessary to attach an axis set to each molecule.…”