Labile cluster nucleation hypothesis

Pure water exists without guests, but with many transient, labile ring structures of pentamers and hexamers. 

Clusters of dissolved species combine to form unit cells. To form si coordination numbers of 20 and 24 are required for 512 and 51262 cavities, while sll requires coordination numbers of 20 and 28 for the 512 and 51264 cavities. Nucleation is facilitated if labile clusters are available with both types of coordination numbers for either si (e.g., CH4 + CO2 mixtures) or sll(e.g., CH4+C3H8 or most unprocessed natural gases). If the liquid phase has clusters of only one coordination number, nucleation is inhibited until the clusters can transform to the other size, by making and breaking hydrogen bonds. 

An activation barrier is associated with the cluster transformation. If the dissolved gas is methane, the barrier for transforming the cluster coordination number from 20 (for the 512) to 24 (for the 51262) is high, both because the guest cannot lend much stability to the larger cavity (see Section 2.1.3.2) and because the 51262 cavities outnumber the 512 in si by a factor of 3. Transformation of methane-water clusters from 

Initial condition Pressure and temperature in hydrate forming region, but no gas molecules dissolved in water. 

Labile clusters Upon dissolution of gas in water, labile clusters form immediately. 

---

A different model presented by Christiansen and Sloan is based on the fact that water molecules form labile water clusters around dissolved gas molecules. The number of water molecules in each water cluster shell depends on the size of the dissolved gas molecules, e.g. 20 for methane and 24 for ethane or 28 for propane. The clusters of the dissolved species combine to form unit cells. The formation rate of a particular hydrate structure depends on the availability of labile clusters with required coordination numbers. With a mixture of methane and propane dissolved in the liquid water phase, hydrates should form more rapidly than if either methane or propane alone are dissolved in the water phase. This cluster nucleation hypothesis is based on the assumption that the guest molecule has to be dissolved in the liquid phase before getting encased into a hydrate lattice. 

In the hypothesis, Points 5 and 8 above (alternative structures) have come under criticism, first by Skovborg et al. (1992) and then by Natarajan et al. (1994). However, Skovborg noted that alternating structures may account for some of his nucleation data. A further criticism of the labile cluster hypothesis is that the energy barrier for agglomeration of clusters is far larger than cluster disintegration (Radhakrishnan and Trout, 2002). 

Molecular simulation methods have been applied to investigate the nucleation mechanism of gas hydrates in the bulk water phase (Baez and Clancy, 1994), and more recently at the water-hydrocarbon interface (Radhakrishnan and Trout, 2002 Moon et al., 2003). The recent simulations performed at the water-hydrocarbon interface provide support for a local structuring nucleation hypothesis, rather than the previously described labile cluster model. 

A hypothesis picture of hydrate growth at a crystal is shown in Figure 3.21, modified from Elwell and Scheel (1975). This conceptual picture for crystal growth may be combined with either the labile cluster or local structuring hypotheses for nucleation. 

We find that the nucleation proceeds via "the local structuring mechanism/ " i.e., a thermal fluctuation causing the local ordering of CO2 molecules leads to the nucleation of the clathrate, and not by the labile cluster hypothesis, one current conceptual picture. The local ordering of the guest molecules induces ordering of the host molecules at the nearest- and next-to-nearest-neighbor shells, which are captured by a three-body host-host order parameter, f these thermodynamic fluctuations lead to the formation of the critical nucleus. 

Similar MC calculations were used by Trout s group to study the carbon dioxide-liquid water interface at 220 K and 4 MPa near the phase boundary of a carbon dioxide hydrate (273 K and 4MPa). Nucleation was achieved by seeding the system with a cluster of carbon dioxide hydrate. It was found that a small cluster with diameter <9.6 A dissolved into the solution readily. A hydrate crystal started to grow, however, when a hydrate cluster twice that size (19.3 A) was implanted into the system. The crystal eventually spanned the whole system (Figure 22). Thus the critical nucleus size for hydrate nucleation is estimated to be about 19 A consisting of approximately 200 water molecules. This is a considerably smaller number than that estimated from the local harmonic model of around 600 molecules. The theoretical results refuted the labile cluster hypothesis.This hypothesis speculates the agglomeration... 

---