Methane and Carbon Dioxide Hydrate Formation and Dissociation in Presence of a Pure Quartz Porous Framework Impregnated with CuSn12 Metallic Powder: An Experimental Report

Hydrate formation and dissociation processes were carried out in the presence of a pure quartz porous medium impregnated with a metallic powder made with a CuSn12 alloy. Experiments were firstly made in the absence of that powder; then, different concentrations were added to the porous medium: 4.23 wt.%, 18.01 wt.%, and 30.66 wt.%. Then, the hydrate dissociation values were compared with those present in the literature. The porous medium was found to act as an inhibitor in the presence of carbon dioxide, while it did not alter methane hydrate, whose formation proceeded similarly to the ideal trend. The addition of CuSn12 promoted the process significantly. In particular, in concentrations of up to 18.01 wt.%, CO2 hydrate formed at milder conditions until it moved below the ideal equilibrium curve. For methane, the addition of 30.66 wt.% of powder significantly reduced the pressure required to form hydrate, but in every case, dissociation values remained below the ideal equilibrium curve.


Introduction
Natural gas hydrate (NGH) is an ice-like solid compound composed by a crystalline structure based on water molecules, which contains natural gas molecules [1]. Hydrate compounds were discovered during the 18th century when, in 1778, Sr. Joseph Priestly made cold experiments in his laboratory in Birmingham and discovered that if temperature was kept close to the ice point, but in any case higher than this value, a mixture containing water and vitriolic air (SO 2 ) was able to pass from the liquid to the solid phase. Starting from its discovery, the history of gas hydrate addressed three different periods. During the first period, hydrate compounds were simply considered a scientific curiosity; thus, scientific production remained particularly contained. Since 1934, due to its capability of causing gas pipelines blockage, gas hydrate started being studied with the aim of preventing its formation and avoiding the consequent issues for the natural gas industry. Finally, in the mid-1960s, scientific interest on gas hydrate rose due to the possibility of making it a new potential energy source. Just in this period, the first natural reservoirs were discovered and the first estimations about the amount of natural gas contained in those deposits were made. To date, these estimates still show some differences between them, but in all cases, the quantity of natural gas present in hydrate was found to be enough to produce more than twice the energy that can still be produced from all conventional energy sources put together [2,3]. NGH reservoirs are mainly sited in offshore sediments (≈97%) and in permafrost (≈3%). The most significant sites were found in the South China The present work deals with gas hydrate formation, with methane and carbon dioxide as guest molecules, in pure quartz sand and with the addition of a metallic CuSn12 powder. Tests were realized in a lab-scale reactor, and different powder concentrations were used. Experiments were carried out in order to establish whether a specific metallic powder is able to intervene on hydrate formation and dissociation conditions and, in a positive case, define its effect.
Hydrate compounds were firstly formed and then dissociated in order to produce the respective equilibrium values, which have been consequently compared with values present in the literature about CH 4 and CO 2 hydrate equilibrium. To the best of our knowledge, in the literature, there is currently a substantial lack of equilibrium values in the presence of that compound, while similar information could be extremely useful to better understand the interaction mechanism of metallic powders with hydrate formation and for potential practical applications. In particular, the present paper aimed to initially investigate the potentialities of realizing tanks and cylinders, having an internal highporous metallic lattice, via 3D print technologies, in order to improve gas storage efficiency, thus increasing the energy density per unit of volume.

Experimental Apparatus
The lab-scale experimental apparatus used in this work to perform hydrate formation and dissociation has been already used in previous research, and a detailed description of it is present elsewhere in the literature [21,24]. The reactor consists of a 316SS cylindrical chamber, with an internal volume equal to 949 cm 3 (diameter 7.3 cm and height 22.1 cm). Such material is chosen following its peculiar corrosion resistance also in severe applications [25][26][27]. More detailed information about geometry is provided in Table 1. The extreme surfaces are closed with two flanges, and an equal number of spirometallic gaskets (model DN8U PN 10/40 316-FG C8 OR) were inserted to avoid any gas leakage. Gas injection may occur both from the upper and the lower flange. Usually, when replacement tests are performed, methane is injected from the bottom, while carbon dioxide is injected from the top. Here, both gases have been alternatively inserted from the bottom in order to guarantee a better diffusion of gaseous molecules inside sand pores.
Temperature was monitored with four Type K thermocouples, having class accuracy 1 and positioned at four different depths, as shown in Figure 1.
Thermocouples have been named as the respective depths: T02, T07, T11, and T16. The temperature-monitoring system was established according to Wang [28][29][30]. Temperature was controlled and varied from the external by inserting the reactor in a thermostatic bath directly connected with a chiller, model GC-LT. Figure 2 shows the 316SS reactor and the thermostatic bath internally equipped with a double copper coil to allow heat exchange between the bath and the refrigerating fluid (glycol), together with a scheme of the completely assembled experimental apparatus. Pressure was measured by using a digital manometer, model MAN-SD, having an accuracy of ±0.5%. All sensors were connected to the Labview software with a data acquisition system manufactured by a National Instrument used for monitoring and recording data from sensors.  [28][29][30]. Temperature was controlled and varied from the external by inserting the reactor in a thermostatic bath directly connected with a chiller, model GC-LT. Figure 2 shows the 316SS reactor and the thermostatic bath internally equipped with a double copper coil to allow heat exchange between the bath and the refrigerating fluid (glycol), together with a scheme of the completely assembled experimental apparatus.

Materials
Ultra-High Purity (UHP) methane and carbon dioxide, with a purity respectively equal to 99.997% and 99.999%, were used for gas hydrate formation in addition to pure demineralized water and sand. This latter compound consists of pure quartz spheres, having diameter equal to 100 µm. The grain porosity is 34% and was measured with a porosimeter, model Thermo scientific Pascal 140. More in depth, the reactor was filled with 744 cm 3 of sand and 236 cm 3 of water; the remaining space was kept free for gas injection. In addition to those compounds, also a metallic CuSn12 powder was used, whose specifications are described in the next section.  [28][29][30]. Temperature was controlled and varied from the external by inserting the reactor in a thermostatic bath directly connected with a chiller, model GC-LT. Figure 2 shows the 316SS reactor and the thermostatic bath internally equipped with a double copper coil to allow heat exchange between the bath and the refrigerating fluid (glycol), together with a scheme of the completely assembled experimental apparatus. Pressure was measured by using a digital manometer, model MAN-SD, having an accuracy of ±0.5%. All sensors were connected to the Labview software with a data acquisition system manufactured by a National Instrument used for monitoring and recording data from sensors.

Materials
Ultra-High Purity (UHP) methane and carbon dioxide, with a purity respectively equal to 99.997% and 99.999%, were used for gas hydrate formation in addition to pure demineralized water and sand. This latter compound consists of pure quartz spheres, having diameter equal to 100 μm. The grain porosity is 34% and was measured with a porosimeter, model Thermo scientific Pascal 140. More in depth, the reactor was filled with 744 cm 3 of sand and 236 cm 3 of water; the remaining space was kept free for gas injection. In addition to those compounds, also a metallic CuSn12 powder was used, whose specifications are described in the next section.

CuSn12 Powder
Copper-Tin powder alloy produced by a gas-atomization process is considered in this work. The alloy's nominal chemical composition (expressed in weight %) is 12% Sn and 88% Cu. The morphology characterization of the CuSn12 powder was carried out by means of high-resolution electronic scanning microscope (FE-SEM Zeiss LEO-1530). The shape of the powder appears generally spherical ( Figure 3) with a size in the range between 5 and 20 μm (the average diameter is approximately equal to 11 μm). Moreover, ICP-MS analyses were carried out to detect the possible release of ions in water; results proved the absolute absence of any release of ions.

CuSn12 Powder
Copper-Tin powder alloy produced by a gas-atomization process is considered in this work. The alloy's nominal chemical composition (expressed in weight %) is 12% Sn and 88% Cu. The morphology characterization of the CuSn12 powder was carried out by means of high-resolution electronic scanning microscope (FE-SEM Zeiss LEO-1530). The shape of the powder appears generally spherical ( Figure 3) with a size in the range between 5 and 20 µm (the average diameter is approximately equal to 11 µm). Moreover, ICP-MS analyses were carried out to detect the possible release of ions in water; results proved the absolute absence of any release of ions.  Previous experiments led us to assert an inhibiting action of this compound on CO2 hydrates.
In the presence of CO2 as a guest compound, the results found in the literature were different from those observed during the experiments: Cu and CuO were defined as ki- Previous experiments led us to assert an inhibiting action of this compound on CO 2 hydrates.
In the presence of CO 2 as a guest compound, the results found in the literature were different from those observed during the experiments: Cu and CuO were defined as kinetic promoters. However, it might be due to the presence of CH 4 in the initial mixture. In conclusion, Cu and CuO particles are known to act as kinetic promoters for methane hydrates, while the experimental evidence led us to affirm an opposite action on carbon dioxide hydrates.
The chemical composition, method of production, and current application of CuSn12 powder explained why this compound was selected for the present research [33].

Methods
Experiments were carried out in the presence of pure quartz porous sand. That porous medium was used to extend the gas hydrate formation to the whole volume and not only in that in correspondence with the gas-liquid interface. In addition, sand allowed us to distribute CuSn12 grains along the whole reactor. Conversely, the powder would settle on the bottom. Two tests were realized without the metallic powder previously described, while in the others, three different concentrations were used in order to investigate also the relation between hydrate formation and powder concentration. In particular, the following quantities were used: 50, 250, and 500 g. Throughout the manuscript, those quantities have been expressed as percentage of the whole amount of solid material present inside the reactor before hydrate formation or the sum of quartz sand and metallic powder. The three quantities indicated above correspond to 4.23 wt.%, 18.1 wt.%, and 30.66 wt.%. That choice was preferred to the others because the chemical additive was at the solid state and not at the liquid or gaseous, as additives are usually applied.
Metallic powder was firstly mixed with silica sand until obtaining a satisfactory mixture; then, the reactor was filled with it and immediately after with water. In Figure 4, the mixture of sand and powder is shown.  The difference in density and granulometry made the achievement of a completely homogeneous phase difficult: metallic powder diffused elsewhere; however, its concentration was not completely uniform and, as it is possible to see in Figure 4 (at right), it assumed higher concentrations in some contained portions.
The chiller was regulated in order to generate an internal temperature about 1-3 °C inside the reactor. Gas injection occurred slowly in order to increase pressure by maintaining a constant gradient. In tests carried out with methane, the initial pressure was fixed at 54-56 bar, while in tests where carbon dioxide was involved, it ranged from 38 to 40 bar. The presence of feasible conditions immediately led to hydrate production, whose formation was proved by a constant pressure decrease. Initially, the temperature increased significantly due to heat production caused by water cages formation; then, it constantly decreased until reaching the thermal balance with the system. As soon as the pressure stabilized, the formation process was considered over. Following that, the chiller was switched off, and thermal energy was provided from the external to increase the tem- The difference in density and granulometry made the achievement of a completely homogeneous phase difficult: metallic powder diffused elsewhere; however, its concentration was not completely uniform and, as it is possible to see in Figure 4 (at right), it assumed higher concentrations in some contained portions.
The chiller was regulated in order to generate an internal temperature about 1-3 • C inside the reactor. Gas injection occurred slowly in order to increase pressure by maintaining a constant gradient. In tests carried out with methane, the initial pressure was fixed at 54-56 bar, while in tests where carbon dioxide was involved, it ranged from 38 to 40 bar. The presence of feasible conditions immediately led to hydrate production, whose formation was proved by a constant pressure decrease. Initially, the temperature increased significantly due to heat production caused by water cages formation; then, it constantly decreased until reaching the thermal balance with the system. As soon as the pressure stabilized, the formation process was considered over. Following that, the chiller was switched off, and thermal energy was provided from the external to increase the temperature again and cause hydrate dissociation. The strategy adopted to increase the temperature during the dissociation phase was selected according to previous research, which proved that the experimental results produced in this way had high similarity with the equilibrium conditions widely described in the literature. In addition, further experiments were carried out where temperature increase was regulated from the external and a lower temperature gradient was established. The comparison between results reached in this way and those present in this article proved again the accuracy of the method here adopted. This latter phase was used to consider equilibrium values for both types of guest compounds in the presence of CuSn12 powder.

Results
As previously explained, methane and carbon dioxide hydrate formation were firstly tested in the presence of pure silica sand and then in the presence of three different concentration of CuSn12 powder: 4.23 wt.%, 18.1 wt.%, and 30.66 wt%. Hydrate compounds were formed and then left free to dissociate in order to produce equilibrium values. A comparison between the results here produced and equilibrium values present elsewhere in the literature was provided .
This section has been divided into three parts: firstly, methane hydrate formation tests are described and represented in pressure-temperature diagrams; then, a similar section is dedicated to tests involving carbon dioxide. Finally, the results produced with these two compounds were put together and compared in order to verify if the use of this CuSn12 powder produced similar effects in both cases or not. Table 2 shows equilibrium values for methane hydrate both in the absence and in the presence of the CuSn12 powder previously described. According to the method present in the literature, those equilibrium values were taken during hydrate dissociation. In fact, the dissociation phase followed an almost linear trend and was less affected by external parameters, unlike hydrate formation, whose temperature trend suffered from the formation of new hydrate formation nuclei, which caused local and delayed temperature peaks as well as the tendency of the whole system to reach thermal balance with the surrounding experimental apparatus. Then, Figures 5-12 describe the pressure and temperature trends over time and the pressure-temperature trend observed during tests involving methane. For each different concentration analyzed, one test is shown. In the first type of diagram, values measured by all thermocouples are shown in order to understand where hydrate formation mainly occurred and also to verify if secondary hydrate nuclei appeared during the hydrate massive growth phase. In diagrams where pressure was described as a function of temperature, this latter parameter was calculated as the average of values measured by each thermocouple.    The temperature trends over time revealed that hydrate formation mainly affected the lowest portion of the volume available for the process: the most significant variations in temperature, associated to the exothermic nature of such reactions, were observed with thermocouples T16 and T11. However, also, the remaining volume was involved, and several secondary peaks in temperature were observed with T07 and T02. In the first experiment shown, carried out without the CuSn12 powder, the first peak, or the main temperature variation immediately after hydrate formation occurs, was not verified as soon as the thermodynamic conditions made hydrate formation feasible; instead, it occurred with a certain delay in time. That aspect is commonly due to the stochastic behavior of the process. The corresponding diagram describing the pressure trends over time showed that any decrease occurred before such temperature variation. In the same experiment, a secondary peak in temperature was observed by all thermocouples, which made the pressure drop more intense. In the remaining part of the process, only thermocouple T11 measured further temperature peaks, which were less intense than the previous one. In every case, a variation in pressure reduction gradient was observed in the correspondence of all those peaks.                The temperature trends over time revealed that hydrate formation mainly affected the lowest portion of the volume available for the process: the most significant variations in temperature, associated to the exothermic nature of such reactions, were observed with thermocouples T16 and T11. However, also, the remaining volume was involved, and several secondary peaks in temperature were observed with T07 and T02. In the first experiment shown, carried out without the CuSn12 powder, the first peak, or the main temperature variation immediately after hydrate formation occurs, was not verified as soon as Differently from it, in the last test, where the CuSn12 powder was inserted inside the reactor, with a concentration equal to 30.66 wt.%, secondary and delayed peaks in temperature were observed in regions associated to T02 and T07, or near the top. As expected, those peaks were immediately followed by a more intense pressure drop than before.

Methane Hydrate Formation Tests
The detection of nucleation sites with all thermocouples proved that methane well diffused in the whole reactor and, similarly, that hydrate formation mainly occurred in sand pores. The first information justified the choice of using a porous medium to produce an internal skeleton able to guarantee diffused hydrate formation. Conversely, only the gas-liquid interface would have been affected by the reaction. The sand pores created a lot of little gas-liquid interfaces inside them, in particular because methane was injected from the bottom; thus, it had pass through the entire medium before reaching the upper surface. Considering the time spent thoroughly mixing sand with CuSn12 powder, whose size is extremely lower than that of sand grains, and taking into account the results obtained in this experience (which will be described in the following lines), we believed that powder adhered to the sand grains, thus influencing hydrate formation both inside and outside sand pores.
Diagrams showing the relation existing between pressure and temperature provided useful information to understand whether that powder intervened or not in the formation process and, in positive cases, what effects it had. The first of those diagrams refers to hydrate formation carried out in the absence of CuSn12 powder. Here, the continuous black line, related to hydrate dissociation, remained completely below the equilibrium curve. It means that the particular silica-based porous medium used during experiments and previously described in Section 2 acted as a promoter for methane hydrate formation. More in-depth, such a promoting effect was observed to increase with temperature: in Figure 6, as in all other P-T diagrams, these two curves moved away from each other in the presence of higher temperature values. That deviation was also found in the presence of powder; however, in this latter case, further considerations were required.
Diagrams describing tests made with concentrations of CuSn12 powder of 4.23 and 18.01 wt.% led to results very similar to those reached in its absence: values of hydrate dissociation were slightly shifted to lower pressures and higher temperature regions, proving again that milder conditions were needed to form hydrate. However, no meaningful variation in the process condition was found. The addition of powder, with the concentrations reported above, did not affect hydrate formation, neither inhibiting nor promoting it. The horizontal trend present in Figure 10 is not particularly meaningful: once the maximum pressure value was reached, the dissociation diagram could only assume that horizontal trend. The values reported in Table 2 provided additional confirmation, as all the values were similar between each other, and in all cases, the observed promoting effect can be completely associated with sand properties.
A completely different behavior was observed in the last experiment, where the CuSn12 powder was used with a concentration equal to 30.66 wt.%. For temperature values higher than 3.8-4.0 • C, the pressure and temperature conditions required to form hydrate become significantly milder than in previous experiments. Thus, with this last concentration, the powder acted as a promoter for methane hydrate formation. As in previous tests, the distance between the ideal equilibrium curve and experimental values increased with temperature. Table 2 confirmed the tendency as soon described.
Considering that the promoting effect was observed only in the presence of a certain concentration, it cannot be motivated with the chemical composition of that powder or geometrical characteristics of its grains. As it is possible to see in Figure 13, such powder acted as a cementing element for sand grains, thus hindering methane escape. Considering that the promoting effect was observed only in the presence of a certain concentration, it cannot be motivated with the chemical composition of that powder or geometrical characteristics of its grains. As it is possible to see in Figure 13, such powder acted as a cementing element for sand grains, thus hindering methane escape. Figure 13. As soon as the test with 30.66 wt.% CuSn12 powder finished, the reactor was opened, and the medium was found to be much more compact than in the absence of powder.
In addition, that powder surely increased the heat transfer rate inside the porous medium, thus giving a further contribution to hydrate formation.

Carbon Dioxide Hydrate Formation Tests
As in the previous section, a table was inserted to describe the hydrate equilibrium values measured in all tests (see Table 3); then, one experiment for each different concentration of CuSn12 (or 0, 4.23, 18.1, and 30.66 wt.%) powder was described by showing the trend over time of its thermodynamic parameters and with a corresponding pressuretemperature diagram. Those diagrams are visible in Figures 14-21. All experiments were carried out by fixing an initial internal pressure in the range of 37-40 bar, thus lower than the corresponding range established for methane hydrate formation tests, due to the milder conditions required by systems involving carbon dioxide to form hydrate. Table 3. Equilibrium values for carbon dioxide hydrate in the presence of CuSn12 powder, analyzed with four different concentrations. Figure 13. As soon as the test with 30.66 wt.% CuSn12 powder finished, the reactor was opened, and the medium was found to be much more compact than in the absence of powder.
In addition, that powder surely increased the heat transfer rate inside the porous medium, thus giving a further contribution to hydrate formation.

Carbon Dioxide Hydrate Formation Tests
As in the previous section, a table was inserted to describe the hydrate equilibrium values measured in all tests (see Table 3); then, one experiment for each different concentration of CuSn12 (or 0, 4.23, 18.1, and 30.66 wt.%) powder was described by showing the trend over time of its thermodynamic parameters and with a corresponding pressuretemperature diagram. Those diagrams are visible in Figures 14-21. All experiments were carried out by fixing an initial internal pressure in the range of 37-40 bar, thus lower than the corresponding range established for methane hydrate formation tests, due to the milder conditions required by systems involving carbon dioxide to form hydrate.      As for methane, also, carbon dioxide hydrate formation mainly affected the portion of internal volume corresponding to thermocouples T11 and T16, whose trend remained higher than other devices in all tests.
Secondary temperature peaks were observed in almost all experiments. Figure 14 shows the temperature and pressure trend over time measured in the absence of CuSn12 powder: thermocouple T11 described an initial peak more extended than the other devices. Then, T16 and T07 registered two secondary peaks, which occurred contemporarily. However, those events are independent from each other: thermocouple T11, which is inserted between T16 and T07, did not register any variation, proving that thermocouples revealed two different nucleation sites, which occurred at the same time.
Conversely to other tests, in the presence of 18.01 wt.% CuSn12 powder, the reaction started near the top of the reactor or in the correspondence of T02 and T07; then, the process involved the whole reactor, and the other two thermocouples started measuring temperature values higher than the others. In the same test, T11 also showed a secondary peak, which was associated to the formation of a new hydrate nucleation site. A similar trend was finally observed in the last tests, where the concentration powder was about 30.66 wt.%. In addition, here, secondary peaks were verified; however, they belonged to thermocouples T02 and T07. It means that hydrate formation firstly interested the lowest portion of the reactor and, in the following period, it involved the whole volume.     A first comparison between methane and carbon dioxide hydrate formation, based on pressure and temperature trend over time, did not reveal any substantial difference: hydrate formation was found to occur elsewhere, even if it mainly affected the regions corresponding to T11 and T16, and the formation of secondary nucleation sites occurred in almost all experiments, despite the quantity of CuSn12 powder used.
On the contrary, important differences were noticed in the diagrams describing the pressure trend as a function of temperature.     Despite what was observed for methane, the porous medium acted as an inhibitor for carbon dioxide hydrate formation. Figure 15 clearly shows that values describing hydrate dissociation remained above the ideal equilibrium curve during the whole test.
Additionally for this guest compound, the formation conditions became milder with increasing temperature values.          As for methane, also, carbon dioxide hydrate formation mainly affected the portion of internal volume corresponding to thermocouples T11 and T16, whose trend remained higher than other devices in all tests.
Secondary temperature peaks were observed in almost all experiments. Figure 14 shows the temperature and pressure trend over time measured in the absence of CuSn12 powder: thermocouple T11 described an initial peak more extended than the other devices. Then, T16 and T07 registered two secondary peaks, which occurred contemporarily. However, those events are independent from each other: thermocouple T11, which is inserted between T16 and T07, did not register any variation, proving that thermocouples revealed two different nucleation sites, which occurred at the same time.
Conversely to other tests, in the presence of 18.01 wt.% CuSn12 powder, the reaction started near the top of the reactor or in the correspondence of T02 and T07; then, the process involved the whole reactor, and the other two thermocouples started measuring temperature values higher than the others. In the same test, T11 also showed a secondary peak, which was associated to the formation of a new hydrate nucleation site. A similar trend was finally observed in the last tests, where the concentration powder was about 30.66 wt.%. In addition, here, secondary peaks were verified; however, they belonged to thermocouples T02 and T07. It means that hydrate formation firstly interested the lowest portion of the reactor and, in the following period, it involved the whole volume.
A first comparison between methane and carbon dioxide hydrate formation, based on pressure and temperature trend over time, did not reveal any substantial difference: hydrate formation was found to occur elsewhere, even if it mainly affected the regions corresponding to T11 and T16, and the formation of secondary nucleation sites occurred in almost all experiments, despite the quantity of CuSn12 powder used.
On the contrary, important differences were noticed in the diagrams describing the pressure trend as a function of temperature.
Despite what was observed for methane, the porous medium acted as an inhibitor for carbon dioxide hydrate formation. Figure 15 clearly shows that values describing hy- The same porous medium acted as a promoter for methane hydrate, while it inhibited carbon dioxide hydrate formation. The main reason of such a difference can be found in sand grains and, in particular, in their size.
Borchardt et al. [60] explained how hydrate growth in inner pores is completely different from that in bulk water: the lattice formed in micro and meso pores is generally smaller, due to confinement effects. For methane, nano hydrate structures were proved to grow in micro pores, with a stoichiometry about one molecule of methane trapped every two molecules of water involved [61]. Both guest compounds studied in this work usually form sI hydrate; however, the CO 2 molecule is larger than that of CH 4 , thus, it may encounter more difficulties in forming hydrate inside inner pores.
The addition of a small amount of CuSn12 powder to the porous medium did not cause any significant change: the hydrate dissociation values remained above the ideal equilibrium curve. However, in this latter case, the dissociation values were slightly shifted to lower pressures than in the absence of powder, thus proving a little promoting effect. This can be easily seen in Table 3, where the results measured during each test are shown and compared.
The introduction of higher quantities of powder inside the reactor led to completely different results. Hydrate formation and dissociation values became significantly milder, until moving below the ideal equilibrium curve. In particular, the dissociation values remained completely below it. The relation with temperature verified in previous experiments was observed also in those cases, where it become even more pronounced. Few differences were found between hydrate formation in the presence of 18.01 wt.% and 30.66 wt.% CuSn12 powder: Table 3 proved again that the greater usage of such powder, or 30.66 wt.%, led to milder dissociation values.
These two concentrations corresponded to the introduction of respectively 250 g and 500 g of powder inside the reactor. In the latter case, the usage of a double quantity of powder only produced a weak effect rather than further promoting action. Consequently, further additions were not considered useful to improve that promoting effect and were not taken into account.
As for methane hydrate, the results reached in this experience were not expected. Very few data about hydrate formation in the presence of metal alloys are currently present in the literature and, in all cases, copper is considered an inhibitor for the process [62,63].
Finally, based on previous research studies [64], a brief description of the formation process is here proposed. This process can be mainly distinguished in two phases: hydrate nucleation and massive growth phases. These two processes can be easily observed in the pressure-temperature diagrams, because they usually assume different trends. For instance, Figures 12, 15 and 19 well show it. During nucleation, pressure decreased slowly due to the contemporary formation and dissociation of hydrates. The main pressure drop occurred during the growth phase, during which the pressure decreased drastically even if the temperature did not show equally relevant variations. This phenomenon happened because the hydrate crystals inside the reactor reached the so-called "critical size" and the spontaneous dissociation of them finished, while they continued their growth. This latter aspect can be explained with the Labile Cluster Theory [1,64], which describes the formation of hydrate nuclei during the first phase, or hydrate nucleation.
According to hydrate nucleation, liquid water molecules initially absorb the gaseous molecules, form primordial clusters, which are composed by a guest molecule and 20-24 water molecules, and finally generate the first unstable 5 12 cages. These structures may continue their growth, via collision with other structures, or dissociate. When vertices are shared during collision, small cubic sI units are formed. Conversely, when faces are shared, small cubic sII cages are produced. In addition, these structures may dissociate or continue their growth. This latter phase continues until the hydrate nuclei reach the critical size; then, massive growth occurs. In the figures previously mentioned, the difference between nucleation and massive growth is well visible: nucleation is less expectable, because it depends on numerous variables and consequently consists of a stochastic process. Conversely, the massive growth appears with a drastic decrease in pressure, which may occur as soon as the equilibrium conditions are reached (obviously, the induction time must be considered, because the transition phase, during which the system goes through the so-called metastable region, cannot be neglected). Such characteristics are particularly helpful in defining the entity of the inhibiting/promoting effect associated to the presence of CuSn12 particles.

Direct Comparison between Results Reached with Methane and Carbon Dioxide
In this concluding section, the dissociation values measured during all the experiments, both in the presence of methane or carbon dioxide as a guest compound, were shown in a single diagram in order to provide a direct comparison among those values and, if possible, suggest potential applications.
In that figure, the two dotted lines have been used to indicate the ideal methane (the higher) and carbon dioxide hydrate equilibrium (the lower). Warm colors were used to describe experiments involving methane, while cold colors described tests made with carbon dioxide. Figure 22 shows all the results produced in this work, together with ideal equilibrium curves related to both compounds and those already shown in previous diagrams.
As previously explained, in all the tests, methane hydrate formation occurred at milder conditions than equilibrium, while a high dosage of CuSn12 powder was necessary to reach a similar result in the presence of carbon dioxide. In Figure 22, different indicators were used to represent tests made with a certain powder concentration. The quartz-based porous medium used drastically reduced the distance existing between the formation conditions of hydrate containing these two gaseous compounds. In the presence of temperatures lower than 4.5-4.0 • C, CO 2 hydrate required more severe conditions than CH 4 hydrate. The introduction of 4.23 wt.% CuSn12 powder did not cause significant variations, while for higher concentrations, values associated to CO 2 hydrate formation drastically dropped below the ideal equilibrium curve. Similar variations in tests involving methane were found only with 30.66 wt.% powder, and according to CO 2 -based tests, they become more and more evident with temperature increase. helpful in defining the entity of the inhibiting/promoting effect associated to the presence of CuSn12 particles.

Direct Comparison between Results Reached with Methane and Carbon Dioxide
In this concluding section, the dissociation values measured during all the experiments, both in the presence of methane or carbon dioxide as a guest compound, were shown in a single diagram in order to provide a direct comparison among those values and, if possible, suggest potential applications. Figure 22 shows all the results produced in this work, together with ideal equilibrium curves related to both compounds and those already shown in previous diagrams.  Based on the results discussed in this work, the present metallic powder might be used for gas storage. The maximum concentration tested was about 30.66 wt.%, the remaining solid medium consisted of pure quartz sand and was used to guarantee a homogeneous diffusion CuSn12 particles along the whole reactor. With appropriate treatments, a similar powder might be used to directly build a metallic high-porous lattice, thus allowing the use of higher concentrations and optimization of the promoting effect. In that sense, the powder accurately described in those pages may contribute to improving the methane storage capacity of tanks and cylinders, thus increasing the energy density of their content.

Conclusions
The present work allowed us to investigate methane and carbon dioxide hydrate formation in the presence of a porous pure quartz medium and a powder consisting of a CuSn12 metallic alloy. Experiments were carried out with the aim of establishing if this latter material was able to intervene in hydrate formation and dissociation or not and, in positive cases, defining the entity and modalities of its action. A small-scale reactor was used to produce hydrate, firstly in the absence of powder and then by adding it in different concentrations of 4.23, 18.01, and 30.66 wt.%.
The first experimental evidence was provided by the porous medium, which acted as a promoter for methane hydrate, while it inhibited the process in the presence of carbon dioxide. That behavior was related to the geometrical properties of sand grains and their pores.
In both cases, the CuSn12 powder was found to promote hydrate formation. When methane was used as a guest compound, that promoting effect was observed only with a concentration about 30.66 wt.%; however, it was significant and moved hydrate dissociation values to milder thermodynamic conditions. The same occurred in the presence of carbon dioxide with the only difference that also a lower concentration, or 18.01 wt.%, promoted hydrate formation. In this latter case, while sand properties initially fixed conditions of hydrate dissociation above the ideal equilibrium, the involvement of powder brought those values below it.
Based on results reached in this work, it is expected that higher concentrations may further lead to milder dissociation conditions. From here, there is the possibility, which clearly needs to be further researched, to use such powder to build, with the auxilium of 3D print technologies, high-porous lattice in apposite tanks, to improve the gas storage efficiency and the density of energy stored per unit of volume. Further research will be focused on investigating why such metallic powder acted as a promoter for hydrate formation. In particular, analyses on water composition after experiments will be done to detect potential ions release in water. Alongside water, the solid medium will be further investigated as well to determine its variation in thermal conductivity due to the addition of such metallic powder.