Electronucleation for Rapid and Controlled ... - ACS Publications

Jun 14, 2016 - ... The University of Texas at Austin, Austin, Texas 78712, United States ... Arjang ShahriariPalash V. AcharyaKatherine CarpenterVaibh...
0 downloads 0 Views 938KB Size
Letter pubs.acs.org/JPCL

Electronucleation for Rapid and Controlled Formation of Hydrates Katherine Carpenter and Vaibhav Bahadur* Department of Mechanical Engineering, The University of Texas at Austin, Austin, Texas 78712, United States S Supporting Information *

ABSTRACT: Nucleation of hydrates involves very long induction times (hours to days), which is a challenge for applications requiring rapid hydrate formation. This study introduces and analyzes the use of electric fields to accelerate and control hydrate nucleation. Experiments with tetrahydrofuran (THF) hydrates reveal that the induction time can be reduced by 100×, by applying an electrical potential across the precursor solution. The induction time rapidly decreases with increasing voltages and is on the order of a few minutes at 100 V. It is seen that voltage-induced current flow in the solution is responsible for electronucleation. Very low currents (microamperes) are sufficient for electronucleation. Nucleation promotion can be attributed to phenomena associated with bubble formation due to chemical reactions at the electrodes. Overall, this study lays the foundation for the control and promotion of nucleation by electric fields, and enables possibilities for instantaneous nucleation.

H

accompanying chemical reactions at the electrodes will provide nucleation triggers to overcome the energy barrier for hydrate formation. This hypothesis originates from previous work on electrofreezing,11 that is, the use of electric fields to elevate the nucleation temperature of supercooled water. The objective of this work is to explore nucleation promotion via electric fields. The influence of the applied voltage and current on nucleation promotion is quantified and analyzed to identify the mechanisms underlying electronucleation. Additionally, instantaneous nucleation by electronucleation is explored; this would benefit applications that require “nucleation on demand”, which is not currently feasible. Electronucleation of tetrahydrofuran (THF) hydrates was studied in this work. THF (molecular formula C4H8O) forms structure II hydrates4 (C4H10O2) with water from a mixture of THF and water (stoichiometric molar ratio between THF and water is 1:17) at atmospheric pressure and at temperatures12 below 4.4 °C. THF hydrates are commonly used12−18 as model hydrates for methane hydrates (which form under challenging high pressure conditions of ∼75 atm). The ratio of THF to water in this study was 1:15, the relative excess of THF ensures that hydrate formation is not accompanied by ice formation. Figure 1a shows a schematic of the experimental setup. A 15 L cold bath (cooled by a 50/50 mixture of ethylene glycol and water) provided isothermal conditions. Hydrate formation was studied in glass tubes (length, 76 mm; inner diameter, 10.5 mm) fitted with a custom-made rubber stopper. This stopper prevents evaporation and has provisions to insert and hold the electrodes and a thermocouple. Stainless steel electrodes (diameter, 1.6 mm) and a Type-T thermocouple were inserted into the tube, as shown in Figure 1b. The distance between the bottom of the stopper and the electrode tip/thermocouple was 50 mm. The spacing between the electrodes and thermocouple

ydrates are clathrate structures consisting of a lattice of water molecules that trap a hydrocarbon molecule. Hydrates have generated recent research interest, with applications focused on natural gas extraction from hydrates, hydrate-based desalination, hydrogen storage, and carbon sequestration.1−5 Many experimental efforts have studied the kinetics of hydrate formation, often under experimental conditions involving high pressures (>75 atm) and low temperatures (−10 to 0 °C). A formidable challenge associated with the synthesis of hydrates is the high induction (wait) time for hydrate nucleation. As per classical thermodynamics, induction time is the time interval to form the first hydrate “seed” that is large enough for spontaneous further growth.6 Hydrate nucleation times can range from hours to days.4 This is an impediment for applications that require rapid hydrate formation (e.g., desalination of water via hydrate formation). Presently, mechanical agitation and surfactants are the two avenues to promote natural gas hydrate nucleation. Multiple studies4 have found that mechanical stirring can induce nucleation in time frames ranging from tens of minutes to a few hours. The faster kinetics is attributed to increased liquid− gas area for hydrate nucleation. The use of surfactants is another common technique to induce hydrate nucleation within timeframes of a few hours.4 Multiple studies have investigated the role of surfactant type and concentration on hydrate formation.7−10 Sodium dodecyl sulfate (SDS) has been highlighted for its ability to accelerate hydrate formation in nonstirred systems, by inducing a morphological change in the hydrate, which ensures that gas−water contact is maintained.9,10 It is noted that both the above techniques have limitations associated with performance, cost, and environmental issues. This study reports an alternate concept to control and rapidly induce hydrate nucleation. The concept involves electronucleation of hydrates by applying a potential difference across two electrodes immersed in the hydrate former solution. The underlying hypothesis is that the current flow and the © XXXX American Chemical Society

Received: May 27, 2016 Accepted: June 14, 2016

2465

DOI: 10.1021/acs.jpclett.6b01166 J. Phys. Chem. Lett. 2016, 7, 2465−2469

Letter

The Journal of Physical Chemistry Letters

Figure 1. (a) Schematic of experimental setup. (b) Details of the electrodes inside the tube. The cathode was coated with a dielectric layer (CYTOP), which had pinholes to allow limited current flow.

was approximately 1.5 mm. The electrodes were connected to a DC power supply and an ammeter. A custom holder was fabricated to hold and immerse multiple tubes in the bath. Within each tube, the cathode was dip coated with a dielectric layer (∼5 μm thick CYTOP from Asahi Chemicals). Pinholes were then introduced in the dielectric layer using a sharp blade tip. The objective behind the dielectric layer with pinholes is to restrict and control the amount of current flow. In the absence of a dielectric layer, there will be significant current flow through the solution, resulting in Joule heating. At the same time, a certain amount of current is required to trigger nucleation, and the pinholes enable current flow pathways. Each cathode had 40 pinholes distributed uniformly near the lower end; the pinholes were 0.5 mm long and 0.1 mm wide. Additional details regarding electrode fabrication and cleaning processes are provided in the Supporting Information. Experiments were conducted with multiple tubes simultaneously to obtain statistically meaningful results. Each tube containing 5 mL solution (ratio of THF:water was 1:15) was agitated to ensure mixing, and then degassed in a sonication bath to remove air bubbles (which can act as potential nucleation sites). The tubes were then immersed in the bath set at 5 °C. After the tubes cooled down to 5 °C, the bath temperature was reduced to −5 °C, and the tubes cooled down to −5 °C within 20 min. At this point, the electrical voltage (10, 25, 50, or 100 V) to promote hydrate nucleation was applied. The induction time was measured from this point onward to the time when hydrate nucleation is detected. The experiment was allowed to run until all tubes nucleated, or for a maximum duration of 3 h. The onset of nucleation was detected by two measurements. The first indicator is a temperature spike registered by the thermocouple. At the onset of nucleation, the water−THF system releases heat (heat of formation of hydrate) and reaches the equilibrium temperature for hydrate formation (∼4 °C). This sudden temperature spike (Figure 2) is an indication of hydrate nucleation. It should be noted that this technique has been used by Dai et al.19 to infer the nucleation of THF hydrates; additionally, it has also been used by multiple researchers20,21 to infer ice nucleation. Importantly, the temperature spike to 4 °C indicates that hydrates are being formed instead of ice. The temperature spike corresponding to ice nucleation would have been to 0 °C. The Supporting Information has details of additional related experiments that confirm that THF hydrates are being formed and has an image of a THF hydrate plug. It should be noted that an ungrounded thermocouple was used to prevent the applied voltage from influencing the thermocouple.

Figure 2. Temperature spike and current decrease at the onset of hydrate nucleation at an electronucleation voltage of 25 V.

The second indicator of hydrate nucleation is a sudden decrease in the electrical conductivity of the solution. As the water and THF molecules rearrange in a clathrate structure, the ability to conduct electricity decreases; this translates to a sharp decrease in the current (Figure 2). It is important to note that the influence of current-flow-induced Joule heating was negligible. The prenucleation current was low at 37, 115, and 250 μA at 25, 50, and 100 V, respectively; this translates to very low heat generation. Even at 100 V, the temperature increase due to Joule heating was less than 0.2 °C, which does not adversely affect nucleation. A third indicator of nucleation was by visual observation. Hydrates are opaque, and formation could be observed in the tubes immersed in the bath from above. Figure 3 shows the fraction of tubes that did not nucleate hydrates versus time for applied voltages of 0, 10, 25, and 100 V. It should be noted that there were at least 10 experiments conducted at each voltage (details provided in Supporting Information). The results of the control experiments (no applied voltage) show that more than 95% tubes did not nucleate hydrates even after 180 min. In fact, the 0 V experiments were allowed to run for 15 h, with nucleation not observed in 85% of the tubes. In contrast, all the tubes at 100 V showed hydrate nucleation within 7 min. It is clearly seen that increased voltages reduce the induction time, with all tubes nucleating in less than 50 and 18 min at 25 and 50 V, respectively. Figure 3 clearly highlights the utility of electronucleation. The results indicate that electrical potentials can reduce the induction times by more than 100×. Furthermore, the induction times are voltage dependent; this opens up new 2466

DOI: 10.1021/acs.jpclett.6b01166 J. Phys. Chem. Lett. 2016, 7, 2465−2469

Letter

The Journal of Physical Chemistry Letters

Figure 4. Induction time for THF hydrate nucleation versus (a) voltage and (b) current in single-tube experiments.

pulses instead of a constant voltage could further lower the induction time. Overall, these results indicate possibilities for “nucleation on demand”, which would benefit many applications. The mechanisms underlying electronucleation can be explained by considering the energetics of the reaction, and the occurrence of nucleation promoting phenomena. First, the energy barrier for the hydrate formation reaction is lowered with the application of an electric field;11 this increases the likelihood of nucleation. Second, the current flow in the solution leads to localized electrolysis at the pinhole sites; this will generate hydrogen bubbles at the cathode. These bubbles act as nucleation sites to increase the likelihood of nucleation. Furthermore, these bubbles grow and eventually detach from the surface. The liquid convection and the pressure variations associated with bubble growth and detachment can provide the additional energy to initiate nucleation in an otherwise quiescent fluid. The bubble-related hypothesis was validated by separate visualization experiments to obtain qualitative comparisons of bubble formation at various voltages. Figure 5 shows the

Figure 3. Fraction of experiments without hydrate nucleation versus time at different voltages. It is seen that electronucleation can significantly accelerate hydrate nucleation. At 100 V (with current flow), all experiments successfully nucleated hydrates within 7 min (leftmost curve). In the absence of a voltage, 85% experiments did not nucleate hydrates even after 15 h.

avenues for control and promotion of nucleation. All these aspects position electronucleation as a powerful tool to enable rapid production of hydrates, in addition to mechanical stirring and the use of chemical promoters. Additional insights into electronucleation can be obtained by measuring the influence of an electric field on nucleation in the absence of current flow. Such experiments were conducted by not introducing pinholes in the dielectric layer. The pinholefree dielectric blocks current flow, but allows charge to build up at the electrode surface. Figure 3 shows the hydrate nucleation fraction versus time for the 100 V/no-current experiments. The 100 V/no-current experiments showed much fewer hydrate formation instances compared with the 100 V/current experiments. In fact, the 100 V/no-current experiments had fewer hydrate formation instances than even the 10 V/current experiments. These results suggest that an electric field by itself is not the primary cause of electronucleation and that current flow has an important role in promoting nucleation. These observations agree with previous results from Carpenter and Bahadur,11 which showed that although an electric field alone can promote ice nucleation, the combined effect of electric fields with current flow is much stronger at inducing ice nucleation. The results reported in Figure 3 are based on multiple experiments conducted simultaneously. To probe hydrate nucleation in greater detail, additional experiments with individual tubes were conducted at voltages of 25 and 100 V. This allowed for direct measurements of the currents associated with hydrate formation instead of measuring the total current in multiple tubes (arranged in parallel). Figure 4 shows the induction time versus current and voltage for these single tube experiments. It is seen that the induction time decreases with higher voltages and currents. The induction times at 100 V are considerably lower than those at 25 V (Figure 4a). Furthermore, the spread in measured induction times is also lower at 100 V. It should be the noted that the error bars in Figure 4 are much smaller (