Nucleation Curves of Carbon Dioxide Hydrate in the Absence of a Solid Wall

Abstract

Nucleation kinetics of clathrate hydrate is poorly understood because of the difficulties in determining the essential parameter, nucleation rate. Nucleation rate enables quantitative comparisons of the impacts of various additives and solid walls on nucleation. We made a setup that determines the nucleation curves of structure Iforming CO2 hydrate in quiescent quasi-free water droplets supported by a bulk of liquid perfluoromethyldecalin under isobaric conditions. The results were compared to the nucleation rates of CO2 hydrate in quiescent water that was in direct contact with stainless-steel walls. We assessed the convergence of the nucleation curves with the increasing numbers of nucleation data and compared our results to the nucleation rates of methane/propane mixed gas hydrate in quiescent quasi-free water droplets and the nucleation rates of gas hydrates of various guest types in the presence of solid walls reported in the literature. We found that (1) 400 nucleation events were sufficient to construct a reliable nucleation curve; (2) the addition of stainless-steel walls promoted the nucleation kinetics of CO2 hydrate, as it did to methane/propane mixed gas hydrate; (3) the kinetic parameter was significantly lower than the theoretically expected value, whereas the thermodynamic parameter was comparable to both the theoretically expected value and the experimentally determined value reported in the literature; and (4) CO2 hydrate nucleated over a substantially shallower supercooling range and had higher nucleation rates than those of methane/propane mixed gas hydrate, both in the presence and in the absence of a solid wall.

Direct Measurement of the Four-Phase Equilibrium Coexistence Vapor–Aqueous Solution–Ice–Gas Hydrate in Water–Carbon Dioxide System

“…The ice freezing process usually occurs at a higher rate than the formation of the sI methane hydrate or sII C 1 -C 3 hydrate [65]. However, CO 2 has a higher solubility in water compared to hydrocarbons, so the kinetics of carbon dioxide hydrate formation are faster and more favorable [69][70][71][72]. Therefore, the difference between the ice and gas hydrate formation rates becomes smaller in the case of carbon dioxide.…”

Nucleation of CO2 Hydrate in Quasi-Free Droplets of Dilute Electrolytes

“…The details of the experimental highpressure setup were described in the reference. 27 Briefly, the highpressure setup (shown in Figure 2) consisted of a CO 2 gas cylinder (99.5% purity for CO 2 gas, supplied by Linde Welding and Supplies), a two-stage vacuum pump (model VP10D, CPS Products, Inc.), a pressure transducer (model PM, Heise), a gas piston (model ZR-3), an ISCO syringe pump (model 260X, Syrixus), and a high-pressure chamber with the sample cell placed in it.…”

“…It is noted that the experimental results measured in pure water were the same as those in the reference. 27 We also noted that the concentration of these two salts was 1 mM, and around 120 data points were collected for deriving the nucleation curve for each salt.…”

“…After obtaining the raw data, the survival curve and nucleation curve were derived by using the analytical method described in the earlier publications. ,− The procedure could be concluded as follows: (1) obtaining the supercooling (Δ T ) of each droplet, where Δ T = T equilibrium – T f ; (2) constructing the survival curve of survival probability, F , as a function of Δ T , where F was calculated from the number of unnucleated droplets divided by the total number of droplets at a given time; (3) calculating ln F with respect to the lag time ( t ), where t was calculated from Δ T divided by the real sample cooling rate (derived from the calibration table); (4) fitting an exponential function curve with the form of A ·exp­( Bt ) + C to the curve ln F ( t ) and obtaining the slope of ln F at each t by analytically differentiating the fitted curve; (5) deriving the nucleation rate as a function of Δ T , where the nucleation rate was calculated from correcting each slope by a factor of −ln 2.…”