Promoting and Inhibitory E ﬀ ects of Hydrophilic / Hydrophobic Modiﬁed Aluminum Oxide Nanoparticles on Carbon Dioxide Hydrate Formation

: Hydrate-based CO 2 capture from large emission sources is considered a promising process for greenhouse gas mitigation. The addition of nanoparticles may promote or inhibit the formation of hydrates. In this work, CO 2 hydrate formation experiments were performed in a dual-cell high-pressure reactor. Non-modiﬁed, hydrophilic modiﬁed and hydrophobic modiﬁed aluminum oxide (Al 2 O 3 ) nanoparticles at di ﬀ erent concentrations were added to assess their promoting or inhibitory e ﬀ ects on CO 2 hydrate formation. The equilibrium temperature and pressure, induction time, and total gas consumption during CO 2 hydrate formation were measured. The results show that the presence of Al 2 O 3 nanoparticles exerts little e ﬀ ect on the phase equilibrium of CO 2 hydrates. Under the experimental conditions, the addition of all Al 2 O 3 nanoparticles imposes an inhibitory e ﬀ ect on the ﬁnal gas consumption except for the 0.01 wt% addition of hydrophilic modiﬁed Al 2 O 3 nanoparticles. The induction time required for the nucleation of CO 2 hydrates mainly ranges from 70 to 90 min in the presence of Al 2 O 3 nanoparticles. Compared to the absence of nanoparticles, the addition of non-modiﬁed and hydrophilic modiﬁed Al 2 O 3 nanoparticle reduces the induction time. However, the hydrophobic modiﬁed Al 2 O 3 nanoparticles extend the induction time. applied to CO 2 geological sequestration to prevent hydrate formation in pipes. These experimental data also indicate that nanoparticle surface modiﬁcation could be applied to e ﬃ ciently promote hydrate formation. Nanoparticle surface modiﬁcation provides a theoretical basis for hydrate-based CO 2 capture. However, the inﬂuence of nanoparticles on hydrate formation is very complicated and requires the comprehensive consideration of a variety of inﬂuencing factors. This study only considers macro-level processes, while the micro-mechanism remains unclear, which requires follow-up e ﬀ orts.


Introduction
Gas hydrates or clathrate hydrates, are non-stoichiometric compounds composed of water and gas molecules under specified pressures and low temperatures (generally above the freezing point of water). In the hydrate structure, water molecules form hydrate cages via hydrogen bonds in which gas molecules such as CH 4 and CO 2 are trapped [1][2][3]. One volume unit of CO 2 hydrates ideally contains 175 standard volume units of CO 2 gas and generates pure water upon disassociation [4]. Therefore, in recent years, researchers have studied CO 2 hydrates as a means of gas separation and seawater desalination [5]. CO 2 is a primary greenhouse gas generated by large fossil fuel end users, such as coal burning power plants, steel works, and chemical plants. CO 2 capture and storage (CCS) has attracted increasing attention [6][7][8]. Traditional CO 2 capture methods mainly include the chemical absorption method, adsorption method, and membrane separation method [9,10], but they experience the problems of a high energy consumption and small adsorption amount. The hydrate-based CO 2 capture process has the advantages of environmental friendliness, a simple process, and low energy consumption, and is considered a CO 2 capture method with great application prospects [11,12].
In the process of CO 2 sequestration, hydrate formation will cause serious problems such as pipeline blockage [13,14]. The addition of inhibitors may effectively prevent hydrate formation [15]. There are three kinds of commonly used hydrate inhibitors: thermodynamic hydrate inhibitors (THIs), kinetic hydrate inhibitors (KHIs), and anti-agglomerant inhibitors (AAs). Among them, KHIs and AAs are low-dose hydrate inhibitors. THIs shift the hydrate phase equilibrium curve to higher pressures and lower temperatures [16][17][18]. Nanoparticles can be added to drilling fluids to maintain the wellbore stability. Nanoparticles have a dual function in terms of hydrate formation as a new additive, which mainly affects the kinetics of hydrate formation. The main influencing factors of nanoparticles are induction time, hydrate formation rate, gas consumption, storage capacity, and so on [19][20][21]. The suspensions of 0.1 wt%, 0.2 wt%, and 0.3 wt% of different nanoparticles (Al 2 O 3 , SiO 2 , Cu, and Ag) were selected. During dissolution, the 0.3 wt% SiO 2 nanoparticles exhibit the most positive effect on gas consumption [20]. Three nanofluids of CuO, Al 2 O 3 , and SiO 2 with three concentrations show the promising effects on the reduction of the average induction time of CH 4 and CO 2 hydrate formation [21]. Moreover, certain nanoparticles have been found to impose a slight inhibitory effect on the thermodynamics of hydrates. Compared to deionized water at a given temperature, the formation pressure of CO 2 hydrate increases in suspensions of graphite nanoparticles. However, this inhibitory effect is independent of the concentration of graphite nanoparticles [22]. In addition, zinc oxide nanoparticles slightly inhibit the thermodynamics of CO 2 hydrate formation but increase gas consumption [23]. Different types of Al 2 O 3 nanoparticles exhibit the dual effect of inhibiting and promoting the induction time of CO 2 hydrate formation, and may constrain gas consumption [24]. Therefore, the effects of nanoparticles on hydrate formation need to be considered from the aspects of thermodynamics and kinetics.
Moreover, at a certain concentration and particle size range, hydrophilic SiO 2 nanoparticles inhibit CH 4 hydrate formation and growth. They increase the induction time required for hydrate formation and reduce gas consumption [25]. The dry water prepared from hydrophobic nanometer-sized silica exerts a thermodynamic promoting effect on hydrate formation, and relevant model results are in good agreement with experimental data [26]. Compared to water baselines, the addition of both hydrophobic and hydrophilic multiwall nanotubes (MWNTs) improves the growth rate of methane hydrates, and the growth rate varies with the change in MWNT concentration [27]. According to the above study [25][26][27], compared to non-modified nanoparticles, hydrophilic and hydrophobic modified nanoparticles have a significant effect on the formation of hydrate. The hydrophilic and hydrophobic properties of modified nanoparticles may be used to obtain the better hydrate promoters or hydrate inhibitors.
In this work, non-modified nanoparticles, hydrophilic modified nanoparticles, and hydrophobic modified nanoparticles of Al 2 O 3 with an average diameter of 30 nm were applied to study their effect on CO 2 hydrate formation. The effects on the equilibrium of the hydrate phase, the induction time required for hydrate formation, and the final gas consumption were mainly measured and analyzed to evaluate the promoting or inhibitory effect of these Al 2 O 3 nanoparticle additives. These Al 2 O 3 nanoparticles could be applied to engineering practice according to their promoting or inhibitory effect. Figure 1 shows a schematic diagram of the experimental setup adopted to study the effect of Al 2 O 3 nanoparticle additives on CO 2 hydrate formation. The setup mainly consists of a high-pressure dual-cell reactor, a circulator bath and a data acquisition system. The custom-made reactor (Shandong ShiYi Science and Technology Co. Ltd., Dongying, China) has an effective volume of 240 cm 3 , and a high-pressure resistant glass window is installed on the front side of each cell of the reactor. The reactor is provisioned with a pressure sensor (UNIK5000, 0-25 MPa, ±0.2% FS BSL, GE, Connecticut, USA) and a temperature sensor (PT100, −50-100 • C, ±(0.15 + 0.002 |t|), Tianjin Jinming Instrument Co. Ltd., Tianjin, China). The data acquisition system records the temperature and pressure in the reactor every 30 s and stores the data in a data processor (a personal computer). A cylindrical rotor is placed in each cell of the reactor and controlled by a magnetic stirrer (RCX-1100D, Tokyo Rikakikai Co., Ltd., Tokyo, Japan). A circulator bath (VIVO RT4, Julabo Technology Co., Ltd., Seelbach, Germany) is employed to control the reaction temperature. Carbon dioxide gas with a purity of 99.9% (Dalian Special Gases Co., Ltd., Dalian, China) was used for gas hydrate formation in this work. Al 2 O 3 nanoparticles (Xuancheng Jingrui New Material Co., Ltd., Xuancheng, China) were employed as the additives. The characteristic parameters of the experimental materials are listed in Table 1. In addition, an ultrasonic homogenizer (Scientz-IID, Ningbo Scientz Biotechnology Co., Ltd., Ningbo, China) was adopted to disperse the nanoparticles in water via ultrasonication. The preparation requires 10 min of ultrasonic dispersion for each nanofluid sample. Self-prepared deionized water was obtained with an ultrapure water polishing system (Aquapro2S, Aquapro International Company LLC, Delaware, USA), and the resistivity was measured to be 18.2 MΩ cm −1 . The micromorphology of the Al 2 O 3 nanoparticles was observed by a scanning electron microscope (SU8220, Hitachi, Tokyo, Japan) and transmission electron microscope (HT7700 EXALENS, Hitachi, Tokyo, Japan). The contact angle of water on the Al 2 O 3 nanoparticle powder surface and the surface tension between the nanoparticle suspension and air were measured with the DropMeter™ system (DropMeter A-100, MAIST Vision Inspection and Measurement Co. Ltd., Ningbo, China). The surface functional groups of the Al 2 O 3 nanoparticles were characterized with an advanced Fourier transform infrared spectrometer (Nicolet 6700, Thermo Fisher Scientific, Waltham, MA, USA).

Phase Equilibrium Experiments
In this work, the phase equilibrium of the CO 2 hydrates was determined by the constant volume hydrate formation method. Before each test, the Al 2 O 3 nanoparticle suspensions with a certain concentration were prepared with the ultrasonic homogenizer, and the reactor was cleaned with deionized water three times. Then, the high-pressure reactor was filled with 120 cm 3 of the prepared suspension and vacuumed for 10 min with a vacuum pump to a negative pressure of 0.10 MPa.
Moreover, the circulator bath was set to 288.15 K, at which temperature hydrates could not be formed. Then, CO 2 was injected into the reactor to achieve the expected pressure. After the temperature and pressure had stabilized, the magnetic stirrer mixed the gas and suspension at a rotating speed of 400 rpm, and the circulator bath was set to 274.15 K and cooled the reactor at a cooling rate of approximately 0.20 K/min. The pressure and temperature were automatically recorded every 30 s. After the temperature and pressure in the reactor had stabilized, the circulator was stopped, and the temperature naturally rose. After reaching a certain temperature in the rising process, the hydrates began decomposing, and the observed hydrate decomposition caused a sharp rise in pressure. The starting point of CO 2 hydrate decomposition was considered as the equilibrium point of the CO 2 hydrate phase in this experiment. The collected temperature and pressure data were used to draw p-T curves.

Hydrate Formation Experiments
Different Al 2 O 3 nanoparticle suspensions with concentrations of 0.005, 0.01, 0.05, 0.1, 0.3, and 0.5 wt% were selected as the additives. The methods of nanofluid addition and reactor vacuuming were the same as those in the phase equilibrium experiments. To maintain the same initial conditions after gas injection, the system temperature (292.26 K) was maintained above the hydrate formation temperature. Then, CO 2 was injected into the reactor to achieve a pressure of 4.50 MPa, and the magnetic stirrer was controlled then at a stirring rate of 400 rpm. Moreover, the temperature of the whole system was gradually reduced to 275.15 K at a cooling rate of 0.2 K/min. During the experiment, hydrate formation and its formation phenomenon could be observed through the visual window. The hydrate formation test lasted 6 h. Each group of experiments was repeated three times to obtain accurate results.

CO 2 Consumption
The consumption of CO 2 during hydrate formation is calculated using the real gas law [23,28]: where P and T are the pressure and the temperature of the reactor, respectively. R is the universal gas constant with a value of 8.314 J/(mol·K), and Z is the compressibility factor, which is obtained with REFPROP (refrigerant physical property software). Subscripts 0 and t in Equation (1) refer to the initial conditions and the conditions at time t, respectively. Equation (2) was applied for the evaluation of the volume of CO 2 inside the reactor at time t [29]: where V reactor and V w 0 are the reactor volume and initial volume of water, respectively. The V RW t and V H t values of the water reacted and hydrates formed at time t, respectively, can be obtained using the following equation [29]: It is generally believed that CO 2 forms structure I hydrates, and the theoretical hydration number of CO 2 hydrates is M = 5.75 [30,31], while v L w is the molar volume of water, which is defined as follows [32]: v MT w is the volume of the empty hydrate lattice and it is defined as follows [33]: The gas consumption rate r (mol·min −1 ) during hydrate formation is a very important for industrial applications and can be obtained using the following equation [34]: where n i−1 i and n i−1 i are the mole numbers of gas in the gas phase at time intervals t i−1 and t i+1 , respectively.

Characterization Results of the Al 2 O 3 Nanoparticles
The surface functional groups of the Al 2 O 3 nanoparticles were characterized by Fourier transform infrared spectroscopy (FTIR). Figure 2 shows the FTIR spectra of the three kinds of Al 2 O 3 nanoparticles.
The curves all reveal wide absorption bands in the wavenumber range from 400 to 1000 cm −1 , which is the characteristic absorption band of Al 2 O 3 . The peaks observed at 2850 and 2918 cm −1 are caused by the symmetric and asymmetric stretching, respectively, of the ethyl group (-CH 2 −), and the absorption peak at 2971 cm −1 is caused by the methyl group (-CH 3 ) [35]. The embedding of such groups can be considered the successful modification of Al 2 O 3 nanoparticles. The absorption peaks at 3443 and 1632 cm −1 correspond to the stretching and bending vibration absorption peaks of O-H, respectively. The absorption peaks of the hydrophilic modified Al 2 O 3 nanoparticles are much higher at these two wavenumbers than those of the other two kinds of Al 2 O 3 nanoparticles, indicating that the amount of hydroxyl groups on the surface is large [36]. In particular, the hydrophilic modified Al 2 O 3 nanoparticles are much more hydrophilic than the other two kinds of Al 2 O 3 nanoparticles.  As shown in the figures, the agglomeration of the Al 2 O 3 nanoparticles is obvious. Therefore, the shape of the Al 2 O 3 nanoparticles cannot be accurately determined. The nanoparticles used in this work are prepared via the electrospray method and can be regarded as spherical particles.

Effect of the Al 2 O 3 Nanoparticle Additives on the Hydrate Phase Equilibrium
No clear conclusion on the effect of nanofluids on the phase equilibrium curve of hydrate formation has been reported in the literature. Park et al. [37] found that there is a positive effect on the CO 2 hydrate phase equilibrium in the presence of multi-wall carbon nanotube (MWCNT) nanofluids. MWCNT nanofluids shift the equilibrium curve of CO 2 hydrates to lower pressures and higher temperatures. Yu et al. [22] indicated that graphite nanoparticles negatively affect the formation of CO 2 hydrates and that the concentration of graphite nanoparticles does not influence the phase equilibrium curve of CO 2 hydrate formation.
In this work, we evaluate the equilibrium pressure and temperature of the formation and dissociation of CO 2 hydrates in the presence of modified Al 2 O 3 nanoparticles at a concentration of 0.1 wt%. Figure 5 shows the phase equilibrium pressures and temperatures of the CO 2 hydrates formed in deionized water, non-modified Al 2 O 3 suspension, hydrophilic modified Al 2 O 3 suspension, and hydrophobic modified Al 2 O 3 suspension. We found that compared to deionized water, the Al 2 O 3 nanoparticle additives impose little effect on the phase equilibrium pressure and temperature of the CO 2 hydrates. It can be inferred that the nanoparticle additives slightly impact the thermodynamics of hydrate formation. This effect is similar to that of surfactants on gas hydrate formation.

Effect of the Non-Modified Al 2 O 3 Nanoparticle Additives on the Kinetics of Hydrate Formation
Many kinds of nanofluids have been applied to hydrate formation in previous studies. However, the mechanism of how nanofluids affect hydrate formation remains unclear. In this work, different non-modified Al 2 O 3 nanoparticle suspensions with concentrations of 0.005, 0.01, 0.05, 0.1, 0.3, and 0.5 wt% are selected as the additives. They are applied to test the inhibitory or promoting effect on CO 2 hydrate formation.
The induction time is an important index of the kinetics of hydrate formation [38]. From the induction time, it can be determined when hydrates begin to form and grow. However, there is no accurate definition of the meaning of the induction time, so in this work, the induction time is defined as the time from the beginning of the cooling process to the obvious rising point on the temperature curve (please refer to Figure 6) [39]. The results reveal that the non-modified Al 2 O 3 nanoparticle additives at the different concentrations impose an inhibitory effect on the final gas consumption of CO 2 hydrate formation compared to that in deionized water (as shown in Figure 7a). The final gas consumption in deionized water is 0.194 mol. The final gas consumption is 0.105 mol in the 0.01 wt% non-modified Al 2 O 3 nanoparticle suspension, which is 45.9% lower than that in deionized water. The final gas consumption is the lowest in the 0.3 wt% non-modified Al 2 O 3 nanoparticle suspension, which is 51.6% lower than that in deionized water. In terms of the final gas consumption, the inhibitory effect of the Al 2 O 3 nanoparticle additive concentration is obvious. These experimental observations agree well with previous studies on the promoting and inhibitory effects of oxide powder additives on gas hydrate formation. Nesterov et al. [24] found that the influence of Al 2 O 3 powders on hydrate formation mainly depends on the surface functional groups and the particle size. The decrease of carbonate carbon, C-H carbon and total carbon on the nanoparticles surface inhibited the formation kinetics of hydrate. The size of the particle aggregates in deionized water is closely related to the concentration of carbonate (carbonyl and C-H carbon on the particle surface) [24]. Within a certain particle size range, larger particle size leaded to less gas consumption. The possible reason was that smaller nanoparticles provided higher surface free energy and were more conducive to gas-liquid contact [40]. We also speculate that the non-modified Al 2 O 3 nanoparticle additives reduce the solubility of CO 2 in deionized water.  These pressure drops are similar in different non-modified Al 2 O 3 nanoparticle suspensions but lower than those in deionized water. The reduction in the pressure drop occurs because the formation of CO 2 hydrates in deionized water and the non-modified Al 2 O 3 nanoparticle additives constrain formation. Figure 7c shows the rate of gas consumption from the beginning to the end of the experiment (the duration of each formation experiment was 6 h in this work). The initial gas consumption rates are very high because of the high dissolution rate of CO 2 in water. The gas consumption rate in deionized water greatly increased at 134 min and reached a maximum value of 1.27 × 10 −3 mol/min at approximately 153 min. A sudden pressure decrease accordingly occurred at the same time (please refer to Figure 7b). It is inferred that at this time, CO 2 hydrates quickly formed and consumed a large amount of CO 2 . Based on the curves of the non-modified Al 2 O 3 nanoparticle additives, the gas consumption rate was not related to the concentration of non-modified Al 2 O 3 nanoparticles, and the gas consumption rate in the non-modified Al 2 O 3 nanoparticle suspensions was lower than that in deionized water. At approximately 80 min, the gas consumption rate became very low.
In Figure 7d, from left to right, the induction time of CO 2 hydrate formation at the different concentrations is 82, 73, 78, 69, 73, 69, and 72 min, respectively. It is found that the non-modified Al 2 O 3 nanoparticle suspensions with the different concentrations slightly reduce the induction time of CO 2 hydrate formation. This can be attributed to the additional nucleation sites provided by the nanoparticle additives. The lower the concentration of the non-modified Al 2 O 3 nanoparticle suspension, the fewer hydrate nucleation sites there are. From the perspective of heat transfer, the hydrates generated much heat at the beginning of formation, and the nanoparticles increased the thermal conductivity of the suspension so that the generated heat could be quickly released [41]. From the results, it is found that the addition of Al 2 O 3 nanoparticles shortens the induction time and reduces the final gas consumption. Hence it can be inferred that the presence of Al 2 O 3 nanoparticles exerts a strong inhibitory effect during the subsequent hydrate growth process.

Effect of the Hydrophilic/Hydrophobic Modified Al 2 O 3 Nanoparticle Additives on the Final Gas Consumption
In this work, the effect of the modified Al 2 O 3 nanoparticle additives on the final gas consumption is studied. The experimental results are shown in Figure 8. As shown in Figure 8, with increasing concentration of the hydrophilic modified Al 2 O 3 suspension, the final gas consumption increases at first and then decreases. In the 0.01 wt% hydrophilic modified Al 2 O 3 nanoparticle suspensions, the final gas consumption reaches a maximum value of 0.209 mol, which is 1.99 times that in the 0.01 wt% non-modified Al 2 O 3 nanoparticle suspensions. Similarly, the addition of 0.05 wt% hydrophilic nanoparticles enhances the final gas consumption up to 86.5% over the level attained in the 0.05 wt% non-modified Al 2 O 3 nanoparticle suspensions. As shown in Figures 3 and 5, the hydrophilic Al 2 O 3 nanoparticles exhibit a high hydrophilicity, which enables these nanoparticles to be uniformly and stably dispersed in deionized water. This stability enhances the liquid phase mixing and improves the mass transfer between the unfilled water cages and gas molecules [42]. However, with increasing hydrophilic modified Al 2 O 3 nanoparticle suspension concentration, the binding force between the hydroxyl groups on the nanoparticle surfaces and the water molecules is enhanced. The activity of water in the fluid weakens, and the final gas consumption decreases.
Compared to the non-modified Al 2 O 3 nanoparticles, the hydrophobic modified Al 2 O 3 nanoparticles also impose a promoting effect on the final gas consumption. The promoting effect reaches its peak when the hydrophobic modified Al 2 O 3 nanoparticle suspension concentration is 0.05 wt%, and the final gas consumption is 1.49 times of that in the 0.05 wt% non-modified Al 2 O 3 nanoparticle suspensions. The hydrophobicity of the hydrophobic modified Al 2 O 3 nanoparticles alters the arrangement of the water molecules and redirects them to positions favorable for hydrate formation [43]. This can also be ascribed to the heterogeneous nucleation of gas bubbles occurring on the hydrophobic surfaces [44]. At low concentrations, the hydrophobic modified Al 2 O 3 nanoparticle suspension concentration influences the final gas consumption little. However, when the concentration reaches 0.5 wt%, the agglomeration of the hydrophobic Al 2 O 3 nanoparticles is enhanced, and the promoting effect of their hydrophobicity on the final gas consumption is weakened.
In general, as the hydrophilic Al 2 O 3 nanoparticles are readily dispersed in water, the effect of these hydrophilic Al 2 O 3 nanoparticles on the final gas consumption is better than that of the hydrophobic Al 2 O 3 nanoparticles. At a low concentration, the mass transfer of dissolved CO 2 is improved due to the greater mixing provided by Brownian motion [27]. However, at a high concentration, the Brownian nanoparticles face a greater opposition to their free motion in the presence of a large number of similar nanoparticles [45]. Due to the notable inhibitory effect of the non-modified Al 2 O 3 nanoparticles, the final gas consumption in the presence of modified Al 2 O 3 nanoparticles is lower than that in deionized water.

Effect of the Hydrophilic/Hydrophobic Modified Al 2 O 3 Nanoparticle Additives on the Induction Time
The acquisition of the induction time has been explicitly explained in the previous section, and the experimental results are shown in Figure 9. The induction time after the addition of the above non-modified and hydrophilic modified Al 2 O 3 nanoparticle additives is shorter than that without nanoparticle addition. However, the hydrophobic modified Al 2 O 3 nanoparticles extend the induction time. The hydrophilic Al 2 O 3 nanoparticles exhibit a better nucleation promoting effect than do the hydrophobic Al 2 O 3 nanoparticles. In the six different hydrophilic Al 2 O 3 nanoparticle suspensions, the induction times for hydrate formation are similar at nanoparticle concentrations of 0.01 and 0.05 wt%. The induction time for CO 2 hydrate formation is 82 min in deionized water. Due to the increasing hydrate nucleation sites provided by the nanoparticles, the hydrophilic and non-modified Al 2 O 3 nanoparticles promote the induction period of hydrate formation. When the amount added of the hydrophobic Al 2 O 3 nanoparticles ranges from 0.005 to 0.1 wt%, the induction time gradually increases, and the inhibitory effect on hydrate nucleation gradually increases to a peak at 0.1 wt%, but the maximum induction time is 6 min longer than the minimum value. It could be considered that the hydrophobic Al 2 O 3 nanoparticle suspension concentration exerts little effect on the induction time. Because the agglomeration of nanoparticles on the dispersed liquid surface is not conducive to the gas-liquid contact, the effect of the hydrophobic Al 2 O 3 nanoparticles on the induction time is inhibited.
In summary, all the experimental results reveal an inhibitory effect on the final gas consumption except for the 0.01 wt% addition of hydrophilic modified Al 2 O 3 nanoparticles. Moreover, the 0.01 wt% addition of hydrophilic modified Al 2 O 3 nanoparticles exhibits a promising effect on the reduction of induction time. Due to Brownian motion, the mass transfer of the hydrophilic modified Al 2 O 3 nanoparticles is enhanced at low concentration. Moreover, the hydrophilic groups on the surface of the nanoparticles make nanoparticles easy to disperse, and nanoparticles provide more sites for hydrate nucleation. In addition, the promoting or inhibitory effect of Al 2 O 3 nanoparticles is related to the experimental conditions and has a high degree of randomness. Further work is needed to reveal the microscopic mechanism of the effect of hydrophilic and hydrophobic nanoparticles on hydrate formation.

Conclusions
In this study, non-modified nanoparticles, hydrophilic modified nanoparticles, and hydrophobic modified nanoparticles of Al 2 O 3 with an average diameter of 30 nm are employed to study their effects on the formation of CO 2 hydrates. The results demonstrate that the presence of Al 2 O 3 nanoparticles imposes little effect on the phase equilibrium of CO 2 hydrates. Under the experimental conditions, all the experimental results reveal an inhibitory effect on the final gas consumption except for the 0.01 wt% addition of hydrophilic modified Al 2 O 3 nanoparticles. The promoting or inhibitory effect depends on the particle size and surface functional groups of the Al 2 O 3 nanoparticles. When considering the influence of the hydrophilicity and hydrophobicity on the final gas consumption, we find that both exert promoting effects. The reasons are that hydrophilicity enables the Al 2 O 3 nanoparticles to be more easily dispersed in deionized water and hydrophobicity alters the arrangement of water molecules on the hydrophobic surface. In terms of the induction time for CO 2 hydrate formation, the hydrophilic and non-modified Al 2 O 3 nanoparticles reduce the induction time over the hydrophobic Al 2 O 3 nanoparticles.
Based on this study, non-modified Al 2 O 3 nanoparticles could be applied to CO 2 geological sequestration to prevent hydrate formation in pipes. These experimental data also indicate that nanoparticle surface modification could be applied to efficiently promote hydrate formation. Nanoparticle surface modification provides a theoretical basis for hydrate-based CO 2 capture. However, the influence of nanoparticles on hydrate formation is very complicated and requires the comprehensive consideration of a variety of influencing factors. This study only considers macro-level processes, while the micro-mechanism remains unclear, which requires follow-up efforts.