a further vindication of the “solid solution” model. In a paper published in Nature entitled, “A complex clathrate hydrate structure showing bimodal hydration,” Udachin and Ripmeester showed that a new hydrate structure formed which consists of alternating stacks of structure H and structure II hydrates. This structure is just one example of the many possible hydrate structures that can be formed from stacked layers.
Figure 1.5 National Research Council Canada building at 100 Sussex Drive, Ottawa, the home of much of the clathrate hydrate research 1990–2015. Source: Reproduced with permission from the National Research Council.
From phase equilibrium studies, dimethyl ether was known to form two hydrates, one having the CS‐II structure, the second was thought to be isostructural with bromine hydrate. However, it, was discovered that the second hydrate had a novel, dense trigonal structure. Unique among the simple hydrate structures, it did not have any pentagonal dodecahedral cages, although some novel cage geometries were observed.
In collaboration with Satoshi Takeya, a major advance was made in the diffraction of powdered hydrate samples (2009). Direct methods for the analysis of powder X‐ray diffraction (PXRD) patterns, dealing with molecular fragments rather than atoms, reduced the number of parameters to be refined and were able to give detailed structural information, including cage occupancies.
The NMR facilities were renewed and improved and Igor Moudrakovski, joining initially as a post‐doctoral fellow (1992), was able to develop a number of unique capabilities suitable for clathrate hydrate work. One of these, MRI, was used to image hydrate layers on ice particles, which, on melting leave intact water droplets inside a hydrate coat. Another new technique, NMR spectroscopy with hyperpolarized 129Xe was developed to give unprecedented signal enhancement to allow the time‐dependent studies of structural and compositional changes. Applications to hydrate formation kinetics showed evidence for a precursor structure before nucleation of crystalline hydrate with the primacy of a small cage environment. Also, a transient structure II hydrate of Xe was identified. Time‐resolved MRI studies of hydrate formation kinetics revealed that the formation process is locally very inhomogeneous even though the spatially averaged kinetics in the bulk system may look homogeneous. Nucleation and crystal growth are simultaneous rather than sequential processes, with multiple nucleation sites forming during the crystal growth process.
The Clathrate Group joined a Canada–Japan project (JAPEX/JNOC/GSC) on the exploration of natural gas hydrate deposits in the Mallik site in the Canadian Arctic. The NRC group contributed laboratory analysis of recovered hydrate samples from the Mallik 2L‐38 site: 13C NMR, Raman, diffraction, calorimetry, gas, and saturation analyses were conducted on samples. Hailong Lu joined the group, first as a visiting scientist (2002), bringing his expertise on the geochemical analysis of natural gas hydrates. One outcome was the establishment of a protocol for the characterization of hydrate from natural sources, including the determination of structure and composition. Collaboration with members of a number of hydrate cruises followed. Samples from many worldwide hydrate locations were shipped to NRC in liquid nitrogen containers. One highlight, published in 2007, was the discovery of naturally occurring structure H hydrate on the Cascadia margin, offshore Vancouver Island. The synthetic version was first reported 20 years before by the hydrate group. The structure is capable of trapping large hydrocarbon molecules and is far more stable than CS‐I hydrate. The findings were published in Nature entitled, “Complex gas hydrate from the Cascadia Margin.”
The van der Waals–Platteeuw theory was developed to describe hydrate formation from water and small non‐polar guests. Potentials to describe guest–host interactions in this case tend to be of the Lennard–Jones or Kihara type. Differences due to guest chemistry tend to be hidden because of the limited number of potential parameters which ultimately are obtained by fitting to experimental observables. This procedure effectively captures all guest–host interactions (e.g. van der Waals interactions, dipolar interactions, hydrogen bonding interaction, etc.) even though only the non‐directional van der Waals interactions are explicitly taken into account when developing the approach. At the NRC, efforts to look for explicit effects of guest chemistry on hydrate properties were spread over many years and involved a number of techniques and contributions from Davidson, Alavi, Udachin, Ratcliffe, Moudrakovski, and Ripmeester, among others.
A most evident guest observable is its water solubility, a property enabled by hydrogen bonding between the water and the guest species. One can then ask: Does hydrogen bonding just disappear once the guest is incorporated from the bulk aqueous solution into the hydrate lattice? In the 1960s Don Davidson, using dielectric relaxation to study water motions in hydrates noted that water reorientational dynamics for hydrates of non‐polar guests were very similar to those for ice Ih. On the other hand, water dynamics for hydrates of water‐soluble guests were much faster with smaller activation energies. He attributed this to transient host–guest hydrogen bonding. For THF, this would be hydrogen bonding of the oxygen atom in the rapidly rotating guest with the water molecules forming the cage wall. Bjerrum defect injection was proposed as a suitable mechanism for this process, as discussed previously by Onsager and Runnels for water molecule reorientation in ice. From the relatively small differences in guest reorientational activation energies for hydrogen‐bonded versus non‐hydrogen‐bonded guests, it can be seen that the interactions tend to be weak, e.g. for THF ∼3.9 kJ mol−1 as compared to cyclopentane with ∼2.8 kJ mol−1. Insight from molecular dynamics computational studies gave the dependence of the defect concentration for different guest types on temperature. One effect of guest–host hydrogen bonding is the transfer of electron density from the host lattice to guest–host units, thus affecting lattice stability. Another is that in a guest–host O⋯HO hydrogen bond, the O⋯O distance is less than the sum of the van der Waals radii. This makes some of the guests appear to be too large for the cage in which they reside, according to calculations of the free van der Waals volume in the cage, e.g. CO2 in the small cage of CS‐I CO2 hydrate. Whereas the oxygen in guests with ether or ketone functions are hydrogen bond acceptors, guests with OH, NH, or NH2 functions can also be hydrogen bond donors, giving rise to complex guest–host hydrogen‐bonding geometries.
Stronger hydrogen bond formation, such as for methanol in the double CS‐II THF‐methanol hydrate and for t‐butylamine in the CS‐II t‐butylamine + H2S binary hydrate, can lead to displaced or vacant water positions in the hydrate lattice. It can be surmised that a sufficient number of such defects can destabilize the hydrate lattice.
Since water‐soluble molecules affect the activity of aqueous solutions, it is clear that water‐soluble guests will also act as inhibitors, as accounted for by a modified van der Waals–Platteeuw equation. Thus, it is important to recognize that hydrate instability may have two origins – one from strong liquid water–guest interactions accounted for by the altered activity, the other by insufficiently strong guest–hydrate cage interactions as accounted for by the magnitude of potential function parameters. Note that the first of these effects can be “turned off” by eliminating the liquid aqueous phase and producing hydrate from an ice–guest molecule reaction. Clathrate hydrates of formaldehyde and ammonia were made this way. On the other hand, so far it has not been possible to produce a binary methanol hydrate. Another interesting observation was the catalyst‐like behavior when small quantities of methanol or ammonia were added to the reaction of methane and ice. These molecules, while being excluded from bulk ice, function as catalysts by disrupting the ice surface by hydrogen bonding to surface water molecules. This greatly enhances the rate of clathrate hydrate formation. This is not a true “catalytic” effect since a small amount of the methanol or ammonia may be incorporated into the hydrate phase.
Halogen–water interactions have proven to give chlorine and bromine hydrates unusual properties. Although the chlorine van der Waals diameter is far too large to fit into the CS‐I hydrate small D cage, cage occupancies of ∼30% have been measured. Compositional analysis by Cady has shown that chlorine