Wettability-Induced Preferential Bubble Nucleation of a Gas from a Two-Gas Dissolved Liquid System

Abstract

This research investigates wettability-induced, preferential, pressure-driven bubble nucleation of gases from a multi-gas dissolved liquid system in hydrophilic and hydrophobic glass vials. The hydrophobic glass surfaces were prepared using (heptadecafluoro-1,1,2,2-tetrahydrodecyl)triethoxysilane (HT). Degassed deionized water in a vial, placed inside a pressure cell, was saturated with a precisely controlled mixture of CO₂ and CH₄ gases at either 6000 mbar or 3000 mbar for 24 h. To initiate the pressure-driven bubble nucleation process, a 500 mbar step-down pressure was applied to the pressure cell every 15 min until bubble nucleation was observed. CH₄ and CO₂ volume fractions were measured using micro-gas chromatography (Micro-GC), while a digital microscope was employed to observe the bubble nucleation process. No bubble nucleation was observed in the case of the hydrophilic vial even when the system pressure was brought to atmospheric pressure. In the case of the hydrophobic vial, the average onset bubble nucleation pressures were 4800 mbar and 2000 mbar for 6000 mbar and 3000 mbar saturation pressures, respectively. The average feed gas concentrations during saturation were 84.44 ± 0.14% and 15.44 ± 0.2% of CH₄ and CO₂, respectively, while at the onset pressure for bubble nucleation, the concentrations shifted to 85.24 ± 0.48% and 13.12 ± 0.52% of CH₄ and CO₂, respectively, when the saturation pressure was 6000 mbar. The average feed gas concentrations during saturation were 85.12 ± 0.28% and 14.67 ± 0.1% of CH₄ and CO₂, respectively, and the average concentrations of CH₄ and CO₂ gases at onset pressure for bubble nucleation were 86.06 ± 1.21% and 12.03 ± 1.03%, respectively, when the saturation pressure was 3000 mbar. The increase in CH₄ concentration is attributed to its preferential separation during the bubble nucleation process.

1. Introduction

Separation of gases from liquids is critical in many natural and industrial processes, such as degassing of magma, gas exchange in lungs, swamp gas formation, oceanic CO₂ outgassing, oil and gas production, separation and refining, bio-gas generation, removal of gases from nuclear reactor coolant systems to prevent corrosion and explosions, and carbon capture and storage (CCS) [1,2,3,4,5,6,7,8,9,10,11]. In natural gas and landfill gas, methane (CH₄) is the major component, while carbon dioxide (CO₂) is typically found as an impurity. The presence of CO₂ leads to the development of pipeline corrosion and a reduction in the energy content of the natural gas [12,13,14]. Therefore, it is important to limit CO₂ to increase the commercial value of natural gas. The preferential separation of CO₂ has been achieved through various technologies, including amine absorption; adsorption by porous materials (e.g., activated carbon, zeolite, and metal–organic frameworks (MOFs)); cryogenic distillation; membrane separation; and reversible absorption [3,12,14]. Pressure swing adsorption (PSA) technology is a promising method of gas separation due to its easier control, low operating and capital investment costs, and higher energy efficiency. MOFs are of great interest due to their high porosity and well-defined pore sizes. MOFs can also be regenerated under milder heating conditions compared to zeolite membranes [12,15,16]. Since the 1980s, the usage of polymeric membranes also gained attention due to a reduction in their production costs, decrease in the equipment size, low energy requirement, ease of operation and minimal environmental impact [3].

Preferential gas separation plays a crucial role in managing greenhouse gas emissions, as CH₄, which according to the US Environmental Protection Agency (EPA), has a global warming potential (GWP) 27 to 30 times greater than CO₂ over a 100-year period, often coexists with CO₂ in natural systems such as oceans and wetlands [17,18]. The preferential release or capture of CH₄ over CO₂ is therefore necessary not only for reducing the overall greenhouse effect but also for enabling energy recovery from CH₄-rich emissions, due to its high calorific value. Effective separation strategies that favor CH₄ separation can significantly enhance the sustainability of gas utilization processes while minimizing the environmental impact of the high-GWP gas.

Membrane separation uses selective barriers to separate gases based on differences in solubility and diffusion rates. It is widely used in carbon capture and natural gas processing due to its energy efficiency and high selectivity. Bernardo et al. (2009) reviewed polymeric membranes for CO₂ separation from flue gases, highlighting the solution-diffusion mechanism and the permeability–selectivity trade-off based on the Robeson upper bound. For example, if the membrane material performs above the Robeson upper bound, it is considered to exhibit exceptional performance, potentially overcoming the typical trade-off between permeability and selectivity, making it highly desirable for gas separation applications [19]. Ali et al. (2019) highlighted the urgent need for effective CO₂ separation to address climate change. Although graphene-based membranes show promise due to their potential for high CO₂ permeability and selectivity, several challenges remain, including limited stability, sensitivity to impurities (e.g., SOx, NOx, trace metals), and performance under real-world conditions. Furthermore, current fabrication methods are costly and complex, with additional expenses for graphene pre-treatment, limiting their industrial viability [20].

Lake Kivu (Democratic Republic of Congo/Rwanda) and Lakes Monoun and Nyos (Cameroon) are meromictic lakes with high concentrations of dissolved gas(es) (CO₂ and CH₄) and CO₂, respectively, in their deep waters, posing risks of catastrophic limnic eruptions. The disproportionate accumulation of magmatic CO₂ gas in the bottom layers of the Lakes Monoun and Nyos resulted in the 1984 and 1986 gas disasters, respectively, causing the loss of more than one thousand people and three thousand livestock [21,22,23]. Preferential gas separation, i.e., selectively extracting or removing specific gases, is critical for mitigating eruption risks, producing energy, and minimizing environmental impacts.

Kling et al. (2005) presented 12 years of observations on gas recharge and lake stability at Lakes Nyos and Monoun, demonstrating that while degassing has successfully maintained stability and lowered gas concentrations, significant volumes of CO₂ still pose a serious threat. Their modeling predicts that the single pipe currently operating in Lake Monoun will soon be unable to keep pace with the natural rate of gas recharge. Similarly, in Lake Nyos, the efficiency of the single pipe is expected to decline as gas concentrations near the intake decrease, potentially extending the timeline for reaching safe levels by several decades. Despite partial progress, the remaining gas is still sufficient to cause deadly consequences if suddenly released. Given that many local residents have remained displaced since the Nyos disaster, the authors argue for urgent acceleration of degassing efforts. Their simulations suggest that installing one additional pipe in Monoun and four more in Nyos would significantly enhance safety and shorten the timeframe for hazard mitigation [24].

Kusakabe et al. (2008) described the long-term stratification of Lakes Nyos and Monoun, where stable chemoclines separate dilute surface waters from high-TDS, CO₂-rich bottom layers. Degassing interventions have induced changes in surface water chemistry and weakened stratification, particularly in Lake Monoun. While the immediate hazard has been mitigated, the potential for renewed gas accumulation remains, introducing uncertainty about the lakes’ long-term stability. The study emphasizes that Nyos-type lakes are rare, requiring specific geophysical and geochemical conditions like depth, volume, CO₂ input, and sustained stratification. However, similar hazard cycles may occur in other volcanic-crater lakes under changing conditions. Therefore, Kusakabe et al. (2008) advocated for continued monitoring of deep, gas-charged lakes, particularly those in volcanic regions, to identify evolving risks of limnic eruptions or gas release events triggered by phenomena such as crater wall collapse or density-driven instability [25].

Our previous studies on pressure-driven bubble nucleation in aqueous systems demonstrated that surface hydrophobicity and the magnitude of step-down pressure significantly affect the onset pressure for bubble nucleation, applicable to both highly and sparingly soluble gases in liquid(s) [26]. For example, our study on the influence of wettability on CO₂ bubble nucleation revealed that the degree of hydrophobicity of the solid surface strongly influences the onset pressure for bubble nucleation [27]. Another study on the influence of step-down pressure on bubble nucleation revealed that not only wettability but also the step-down pressure affects bubble nucleation of sparingly soluble gases (CH₄ and N₂) in water [28]. Our analytical method to estimate supersaturation demonstrated that the supersaturation levels generated in the CO₂–water system are significantly higher than those in CH₄–water and N₂–water systems [29]. A subsequent study investigating the effect of wettability on pressure-driven bubble nucleation in a gas-supersaturated oil–water system suggested that surface/interfacial tensions and vapor pressures of the fluids are the primary driving forces behind bubble formation on hydrophobic surfaces in the aqueous phase exposed to higher step-down pressure, while lower step-down pressure did not result in bubble nucleation [30]. However, it is important to note that all our aforementioned investigations were limited to single-gas supersaturation; therefore, the influence of wettability on pressure-driven multi-gas bubble nucleation warrants further investigation. In the present study, we examine wettability-induced, pressure-driven preferential bubble nucleation within a multi-gas dissolved liquid system to determine whether process conditions can be engineered to selectively favor the separation of one gas over another. Additionally, we analyzed the impact of saturation pressure on bubble nucleation dynamics in this two-gas/liquid system.

2. Materials and Methods

2.1. Materials

Untreated (hydrophilic) and treated (hydrophobic) glass substrates viz. glass slides (make: Fisherfinest, Pittsburgh, PA, USA; dimensions: 25 mm × 75 mm) and glass vials (make: VWR, Radnor, PA, USA; capacity: 3.7 mL; dimensions: 12 mm diameter × 25 mm height) were used for contact angle and roughness measurements and bubble nucleation experiments. (Heptadecafluoro-1,1,2,2-tetrahydrodecyl)triethoxysilane (HT) (Source: Gelest, Morrisville, PA, USA; Product code: SIH5841.2) was used for the preparation of hydrophobic surfaces. For the reaction vessels, 50 mL polypropylene centrifuge tubes (make: Fisher Scientific, USA) were chosen, as they do not react with the chemicals utilized in the surface chemical treatment process. Isopropyl alcohol (IPA; 99%) and ethanol (absolute), procured from Pharmco-Aaper, Brookfield, CT, USA, were used as solvents for the wettability alteration process. Glacial acetic acid, obtained from Pharmco-Aaper, USA, was used as a catalyst for preparing hydrophobic surfaces.

2.2. Wettability Alteration Procedure

Hydrophobic surfaces were prepared using HT chemical, following the procedure outlined in our previous study on pressure-driven bubble nucleation [27]. The summarized procedure is as follows. A solution consisting of 95 wt.% IPA and 5 wt.% water was prepared in a polypropylene tube for surface chemical modification. Subsequently, 10 wt.% of HT was added to the solution, followed by the addition of 20 μL of 0.02 wt.% glacial acetic acid to facilitate hydrolysis of the silane (HT). The solution was mixed thoroughly in a shaking water bath (make: Benchmark Scientific, Edison, NJ, USA; model: SB-12L) for 30 min to ensure complete hydrolysis.

A glass substrate, rinsed with IPA, was then immersed in the prepared solution within the polypropylene tube and subjected to shaking at 150 rpm for 60 min to facilitate the surface chemical reaction. After the reaction, the glass substrate was thoroughly washed with ethanol to remove any unreacted or undesired residues. The treated glass samples were then cured at 105 °C for 30 min in a hot-air oven (make: Thermo Scientific, Waltham, WA, USA).

The untreated glass substrates were inherently hydrophilic, and to remove any undesired polar and non-polar compounds adsorbed on these surfaces, they were rinsed with n-hexane and ethanol.

2.3. Static Contact Angle and Roughness Measurements

The wettability of the glass substrates was quantified by static contact angle measurements, which represent the angle between the tangent to the liquid–fluid interface and the tangent to the solid surface at the three-phase contact line [31,32]. The static air–water contact angles of the untreated and treated surfaces were measured by placing a 5 μL water droplet on a glass slide (flat surface) using a goniometer. The roughness of the hydrophobic surfaces was measured by the AFM analysis.

2.4. Bubble Nucleation Experimental Facility

The schematic of the bubble nucleation experimental setup is shown in Figure 1. The multi-gas bubble nucleation experimental facility consists of three ultra-high purity (UHP) gas cylinders (i.e., helium, CO₂, and CH₄). The CO₂ and CH₄ gas cylinders were connected to two different microfluidic pumps (make: Dolomite, Hertfordshire, UK; models: 1600935 and 4900010; max. pressure: 10,000 mbar; resolution: 1 mbar) through a pressure relief valve (set pressure: 125 psi). One check valve was installed after the cylinder regulator of each gas to prevent any backflow of gas or gas mixture. The Flow Control Center software (make: Dolomite; Version: 1.59) was used to control the pressures of both pumps. The output lines of both pumps were connected to a tee connector to mix and flow both gases simultaneously. The mixed-gas flow line, i.e., the output line of the tee connector, was connected to an inline flow sight pressure cell. The pressure cell contains a hydrophobic vial containing a 5 mm height of degassed DI water.

The output of the pressure cell was connected to an ultra-high-sensitivity zero-flow back pressure regulator (BPR) (make: Equilibar, Fletcher, NC, USA; part no: ZF0; max. pressure: 3000 psi). The pilot pressure of the BPR was supplied from the nitrogen gas cylinder. The outlet of the BPR was connected to a micro-gas chromatograph (make: Agilent, Shanghai, China; model: 490 Micro GC) to measure the individual gas composition. The micro-GC has a Thermal Conductivity Detector (TCD) that measures gas compositions in terms of volume percent, and the PoraPLOT Q and CP-Sil 5 CB columns were used to detect CO₂ and CH₄ gases, respectively. The micro-GC was also connected to a UHP helium gas cylinder, which was used as a carrier gas source.

The OpenLab Chromatography Data System (CDS) EZChrom Edition (make: Agilent) software (Version: A.04.07) was used to process the chromatogram and compute the composition of each gas. Bubble nucleation pictures were taken using a digital microscope (make: Dino-Lite, Torrance, CA, USA; model: AM73915MZTL; magnification range: 10–140×; resolution: 5.0-megapixel, software version: dnc2_1.5.23.A_U) to observe and record the bubble nucleation phenomena. The Extended Dynamic Range (EDR) camera feature was used to capture pictures. All bubble nucleation experiments were conducted at room temperature (~22 °C).

2.5. Procedure to Measure Gas Concentrations in the Absence of Water

To understand the behavior of the gas mixture at different pressures, three standardized experiments were conducted in the pressure cell: the first with an empty hydrophobic vial, the second with an empty hydrophilic vial (both containing no water), and the third with the pressure cell without a hydrophobic vial. The gas concentrations were analyzed at five different pressures (i.e., 6000, 5500, 5000, 4500, and 4000 mbar), starting at 6000 mbar and applying a 500 mbar step-down pressure.

2.6. Bubble Nucleation Experiments

To conduct bubble nucleation experiments with hydrophobic vials containing CO₂- and CH₄-saturated water, the following procedure was followed. The hydrophobic glass vial was filled with degassed deionized water to a height of 5 mm and placed inside the inline flow sight pressure cell. Deionized water was degassed through thermal degasification, which involved boiling to expel dissolved gases and then allowing it to cool to room temperature in a sealed container [33]. A pressure of 6000 mbar (87 psig) was applied to both microfluidic pumps carrying CO₂ and CH_4, and the gas mixture was injected into the pressure cell for flushing out the trapped air for 30 min.

Then the back pressure was set at 90 psig using the nitrogen gas cylinder regulator, slightly higher than the pump’s pressure to prevent gas flow, to saturate the water with CO₂ and CH₄. The required saturation time was estimated using the one-dimensional bounded-diffusion equation [34]. When CO₂ and CH₄ flowed simultaneously at a 6000 mbar pressure, it was observed that the average concentrations (vol.%) of both gases were 15.3 ± 0.2% and 84.6 ± 0.2%, respectively. The estimated saturation time for the gas mixture was approximately 16.9 h, for 99.99% saturation in water. To ensure complete saturation, a 24 h saturation time was maintained for the bubble nucleation experiments.

After complete gas saturation, the two 2-way inline valves were closed. To initiate the dissolved gas separation process, a step-down pressure of 500 mbar (7.25 psig) was applied every 15 min to the pressure cell by reducing the pressure from the nitrogen gas cylinder regulator. Similar gas separation experiments were conducted using hydrophilic vials to study bubble nucleation, followed by measurement of gas concentrations. After conducting five experiments at a 6000 mbar saturation pressure, similar experiments were conducted at a 3000 mbar (43.5 psi) saturation pressure (five replicates) to study the influence of different initial saturation pressures on liberated-gas compositions.

3. Results and Discussion

3.1. Static Contact Angle and Surface Roughness Measurements

The average air–water contact angle and the standard deviation of five repetitions on hydrophobic glass samples were 114.6° ± 0.8°. It should be noted that the untreated flat surface was inherently hydrophilic, and in our previous study, we found that the average air–water contact angle was 33.9° ± 0.4° [27]. The average surface roughness and the corresponding standard deviation values for the hydrophobic flat and vial substrates were 2.4 ± 0.2 and 7.4 ± 1 nm, respectively, showing that the surfaces remained smooth after the silane treatment [27].

3.2. Measurement of Gas Concentrations in the Absence of Water

When the pressure cell did not contain a vial, the average gas concentrations (vol.%) of CH₄ and CO₂ at five different pressures (6000, 5500, 5000, 4500 and 4000 mbar) were 84.35 ± 0.51% and 14.50 ± 0.31%, respectively. When the pressure cell contained a hydrophilic vial without water, the average gas concentrations of CH₄ and CO₂ at these five different pressures were 84.10 ± 0.61% and 14.95 ± 0.39%, respectively. Similarly, when the pressure cell contained a hydrophobic vial without water, the average gas concentrations of CH₄ and CO₂ at the same five different pressures were 84.13 ± 0.50% and 15.04 ± 0.35%, respectively. These results indicate that the gas concentrations were within a similar range when the vial did not contain water.

3.3. Bubble Nucleation Experiments

3.3.1. Bubble Nucleation in Hydrophilic Vials (Saturation Pressure: 6000 mbar)

The average feed gas concentrations (vol.%) in the hydrophilic vials before starting the gas saturation were 84.63 ± 0.39% and 15.34 ± 0.40% of CH₄ and CO₂, respectively. When a step-down pressure of 500 mbar (7.25 psig) was applied to the pressure cell after achieving the desired saturation, bubble nucleation was not observed in the hydrophilic vial, even when the system pressure dropped to atmospheric pressure (Figure 2). The average gas concentrations of CH₄ and CO₂ during the pressure reduction process were found to be 84.15 ± 0.64% and 14.40 ± 0.68%, respectively. From the above measurements, it was observed that the concentration of CH₄ decreased by ~0.5%, whereas the concentration of CO₂ decreased by ~1% compared to the corresponding initial gas concentrations.

3.3.2. Bubble Nucleation in Hydrophobic Vials (Saturation Pressure: 6000 mbar)

The average feed gas concentrations (vol.%) in the hydrophobic vials before starting the gas saturation were 84.57 ± 0.24 vol.% (168.01 ± 0.47 mol/m³) and 15.34 ± 0.23 vol.% (30.48 ± 0.46 mol/m³) of CH₄ and CO₂, respectively, which is similar to the gas concentrations when the vial did not contain water. Once the complete gas saturation in water was achieved, i.e., after 24 h, a supersaturation was created in the water by applying a step-down pressure of 500 mbar (7.25 psig), from the initial saturation pressure (6000 mbar or 87 psi) every 15 min. Bubble nucleation did take place for the 500 mbar step-down pressure in the hydrophobic vial, as can be seen in Figure 3. The average onset pressure for bubble nucleation in hydrophobic vials was 4875 ± 231 mbar (70.7 ± 3.4 psi). The average gas concentrations at the onset pressure for bubble nucleation for the CH₄ and the CO₂ gases were 85.50 ± 0.6 vol.% (169.86 ± 1.19 mol/m³) and 13.26 ± 0.54 vol.% (26.33 ± 1.08 mol/m³), respectively. Therefore, the CH₄ gas concentration increased by 0.93 vol.% (1.85 mol/m³) and the CO₂ gas concentration decreased by 2.08 vol.% (4.15 mol/m³) after bubble nucleation compared to those before saturating the liquid.

3.3.3. Bubble Nucleation in Hydrophobic Vials (Saturation Pressure: 3000 mbar)

The average CH₄ and CO₂ concentrations (vol.%) before starting the gas saturation were 85.12 ± 0.28 vol.% (69.38 ± 0.23 mol/m³) and 14.67 ± 0.10 vol.% (11.96 ± 0.08 mol/m³), respectively. After creating supersaturation by applying a 500 mbar step-down pressure from the initial saturation pressure (3000 mbar or 43.5 psi), the average onset pressure for bubble nucleation in hydrophobic vials was 2000 mbar (29 psi). As can be seen in Figure 4, bubble nucleation did take place for a 500 mbar step-down pressure in the hydrophobic vial. The average concentrations of CH₄ and CO₂ gases at onset pressure for bubble nucleation were 86.06 ± 1.21 vol.% (70.14 ± 0.98 mol/m³) and 12.03 ± 1.03 vol.% (9.80 ± 0.84 mol/m³), respectively. Similarly to in the 6000 mbar saturation pressure experiments, the CH₄ gas concentration increased by 0.94 vol.% (0.76 mol/m³) and the CO₂ gas concentration decreased by 2.64 vol.% (2.16 mol/m³) after bubble nucleation compared to gas concentrations before saturating the liquid. The details of the average CH₄ and CO₂ gas concentrations and the average onset nucleation pressures of 6000 mbar and 3000 mbar initial saturation pressure cases are summarized in

Table 1. Summary of average gas concentrations and average onset nucleation pressures in hydrophobic vials containing water.

At 6000 mbar Saturation Pressure
	Initial Gas Concentration	Final Gas Concentration	Onset Nucleation Pressure
	vol.% (mol/m3)	vol.% (mol/m3)	mbar
CH4	84.6 ± 0.2% (168.0 ± 0.5)	85.5 ± 0.6% (169.9 ± 1.2)	4875 ± 231
CO2	15.3 ± 0.2% (30.5 ± 0.5)	13.3 ± 0.5% (26.3 ± 1.1)
At 3000 mbar Saturation Pressure
	Initial Gas Concentration	Final Gas Concentration	Onset Nucleation Pressure
	vol.% (mol/m3)	vol.% (mol/m3)	mbar
CH4	85.1 ± 0.3% (69.4 ± 0.2)	86.1 ± 1.2% (70.1 ± 1.0)	2000 ± 0
CO2	14.7 ± 0.1% (12.0 ± 0.1)	12.0 ± 1.0% (9.8 ± 0.8)

.

From the above results, it is observed that the CH₄ concentration was increased by ~1 vol.%, whereas the gas concentration of CO₂ was decreased by ~2 vol.%, after bubble nucleation, compared to initial gas concentrations, irrespective of saturation pressure. However, with 6000 mbar saturation pressure, CH₄ concentration was increased by ~2 mol/m³, whereas the gas concentration of CO₂ was decreased by ~4 mol/m³. When the saturation pressure was 3000 mbar, the CH₄ concentration was increased by ~1 mol/m³, whereas the gas concentration of CO₂ was decreased by ~2 mol/m³. At 4875 mbar pressure, CH₄ is approximately 23-times less soluble than CO₂ [35,36]. Therefore, we hypothesize that during the bubble nucleation process, CH₄ gas was preferentially separated from the liquid, resulting in an increase in the CH₄ gas concentration in the headspace. Despite the potential preferential separation of CH₄ gas, the relatively small increase in CH₄ concentration could be due to the fact the original headspace gas concentration was still the initial gas concentration and only a small amount of CH₄ is added to the headspace gas during the bubble nucleation process.

It should be noted that the bubble nucleation observed in this study under a 500 mbar step-down pressure (from saturation pressures of 3000 or 6000 mbar) does not rely on homogeneous nucleation; instead, it is governed by heterogeneous nucleation. As reported by Hemmingsen (1970, 1975, 1977) [37,38,39], classical homogeneous-nucleation theory predicts that large pressure drops, up to 1500 bar, would be required to form bubbles in pure aqueous solutions. In contrast, our experiments were conducted in hydrophobic glass vials, where bubble formation is dominated by heterogeneous nucleation, which dramatically reduces the nucleation energy barrier. In heterogeneous nucleation, bubbles originate from pre-existing gas cavities, allowing nucleation to occur at significantly lower supersaturation levels than those predicted for homogeneous systems. Therefore, the onset of bubble nucleation at modest pressure reductions in our experiments is fully consistent with previously reported observations in the literature.

The pressure reductions at which bubble nucleation occurred in this study (4875 mbar for 6000 mbar saturation pressure and 2000 mbar for 3000 mbar saturation pressure, i.e., ΔP ≈ 1.0–1.13 bar) are consistent with heterogeneous nucleation on hydrophobic surfaces. Mahmood and Kwak [40] reported that extremely hydrophobic surfaces with a liquid-side contact angle of approximately 160° allow bubble nucleation by CO₂ at a relatively low supersaturation (≈0.68 bar). The hydrophobic glass vials used in our experiments exhibit a lower contact angle (≈114.6°). It should be noted that the maximum theoretical air–water contact angle on an ideally smooth surface is approximately 120° [41,42]. A decrease in contact angle is known to increase the nucleation barrier, thereby requiring higher supersaturation according to heterogeneous-nucleation theory and the trends described in the molecular cluster model. In addition, heterogeneous nucleation from pre-existing gas cavities further facilitates bubble formation within the supersaturation range applied in our experiments. Therefore, the onset of bubble nucleation under this modest step-down pressure is consistent with the expected behavior for moderately hydrophobic surfaces.

4. Conclusions

In this work, we investigated the effects of wettability and step-down pressure on preferential bubble nucleation of a gas from a two-gas dissolved liquid system. Degassed deionized water in hydrophilic and hydrophobic vials was saturated with a CO₂ and CH₄ gas mixture of known composition at two different saturation pressures (6000 and 3000 mbar). After achieving the required supersaturation, a 500 mbar step-down pressure was applied to induce bubble nucleation and measure the liberated-gas concentrations. The results demonstrated a strong influence of wettability on the onset pressure for bubble nucleation. The 500 mbar step-down pressure successfully generated the required level of supersaturation to initiate bubble nucleation in the hydrophobic vials. However, bubble nucleation was not observed in the hydrophilic vial even when the system pressure was reduced to atmospheric pressure. In the hydrophobic vial, the CH₄ gas concentration increased by approximately 1%, while the CO₂ concentration decreased by about 2% compared to the initial gas concentrations. In contrast, in the hydrophilic vial, the CH₄ concentration decreased by approximately 0.5%, whereas the CO₂ concentration decreased by about 1% relative to the initial gas concentrations. Due to its lower solubility in water compared to CO₂, CH₄ was likely released from the bulk solution during the pressure reduction process, whereas CO₂ from the headspace may have re-entered the liquid phase. This interplay between gas release and re-dissolution might have contributed to the observed increase in the CH₄ concentration during the gas separation process. Future work will focus on the development of a method for in situ compositional analysis of the nucleated bubble during the pressure reduction.