5.3 The Memory Effect
129
5.3 The Memory Effect
A long-standing mystery in nucleation of clathrate hydrate is the so-called memory
effect [27]. The memory effect refers to observations that clathrate hydrates nucleate
more easily (within a shorter timeframe at a given constant subcooling or at a smaller
subcooling during a linear cooling ramp) in a dissociated water than in a fresh water
that has no history of clathrate hydrate formation. The memory effect has important
practical implications in a recovery of methane gas from methane-bearing sediments.
Since methane hydrate is 85 mol% water, and water is heavy, it is not economical
to transport the mined methane in the clathrate form. It is thus preferable to decompose the methane hydrate into methane gas and water by depressurization and only
transport the recovered methane gas in a pipeline. Transportation of freed methane
gas in the presence of dissociated water is susceptible to reformation of methane
hydrate if any section of the pipeline is inside the thermodynamically stable zone of
the phase diagram of methane hydrate, and the memory effect would increase the
risk by rendering nucleation of methane hydrate easier.
5.3.1 Historical Perspective
The early studies on the kinetics of clathrate hydrate formation already encountered
the memory effect although the term had not yet been coined back then [31]. It was
found that the sample history had some influence on the induction times of clathrate
hydrates but not on the growth after the nucleation [31].
Takeya et al. then reported an interesting study that the memory effect was
observed for the nucleation of carbon dioxide hydrate in the water produced by
melting of ice that had no history of carbon dioxide hydrate formation [62]. The
memory effect was destroyed at 298 K that corresponded to 25 K of superheating
[62].
Rodger reported a molecular dynamic simulation study which showed that mostly
ice-like clusters, not clathrate-like structures, had formed after dissociation of
methane hydrate [63]. He suggested that the memory effect was due to the suppressed
diffusion of methane gas in the dissociated water, which would leave the dissociated
water supersaturated for longer with the methane gas. Such persisting supersaturation of methane gas would render the nucleation of methane hydrate easier when the
dissociated water is brought back inside the thermodynamically stable zone of the
phase diagram of methane hydrate [63].
A distinct attribute of the memory effect is that it varied among an ensemble of
samples that had the same thermal history. Ohmura et al. studied the memory effect
of HCFC-141b hydrate from the dissociated water at the superheating of 0.5, 1.0,
and 1.5 K [64]. Notably, they reported that the level of memory preservation differed
from sample to sample despite the same thermal history of the samples [64]. Zeng
et al. suggested that the memory may be imprinted on a solid surface that is present
129
5.3 The Memory Effect
A long-standing mystery in nucleation of clathrate hydrate is the so-called memory
effect [27]. The memory effect refers to observations that clathrate hydrates nucleate
more easily (within a shorter timeframe at a given constant subcooling or at a smaller
subcooling during a linear cooling ramp) in a dissociated water than in a fresh water
that has no history of clathrate hydrate formation. The memory effect has important
practical implications in a recovery of methane gas from methane-bearing sediments.
Since methane hydrate is 85 mol% water, and water is heavy, it is not economical
to transport the mined methane in the clathrate form. It is thus preferable to decompose the methane hydrate into methane gas and water by depressurization and only
transport the recovered methane gas in a pipeline. Transportation of freed methane
gas in the presence of dissociated water is susceptible to reformation of methane
hydrate if any section of the pipeline is inside the thermodynamically stable zone of
the phase diagram of methane hydrate, and the memory effect would increase the
risk by rendering nucleation of methane hydrate easier.
5.3.1 Historical Perspective
The early studies on the kinetics of clathrate hydrate formation already encountered
the memory effect although the term had not yet been coined back then [31]. It was
found that the sample history had some influence on the induction times of clathrate
hydrates but not on the growth after the nucleation [31].
Takeya et al. then reported an interesting study that the memory effect was
observed for the nucleation of carbon dioxide hydrate in the water produced by
melting of ice that had no history of carbon dioxide hydrate formation [62]. The
memory effect was destroyed at 298 K that corresponded to 25 K of superheating
[62].
Rodger reported a molecular dynamic simulation study which showed that mostly
ice-like clusters, not clathrate-like structures, had formed after dissociation of
methane hydrate [63]. He suggested that the memory effect was due to the suppressed
diffusion of methane gas in the dissociated water, which would leave the dissociated
water supersaturated for longer with the methane gas. Such persisting supersaturation of methane gas would render the nucleation of methane hydrate easier when the
dissociated water is brought back inside the thermodynamically stable zone of the
phase diagram of methane hydrate [63].
A distinct attribute of the memory effect is that it varied among an ensemble of
samples that had the same thermal history. Ohmura et al. studied the memory effect
of HCFC-141b hydrate from the dissociated water at the superheating of 0.5, 1.0,
and 1.5 K [64]. Notably, they reported that the level of memory preservation differed
from sample to sample despite the same thermal history of the samples [64]. Zeng
et al. suggested that the memory may be imprinted on a solid surface that is present