A Review of Caprock Integrity in Underground Hydrogen Storage Sites: Implication of Wettability, Interfacial Tension, and Diffusion

Abstract

As industry moves from fossil fuels to green energy, substituting hydrocarbons with hydrogen as an energy carrier seems promising. Hydrogen can be stored in salt caverns, depleted hydrocarbon fields, and saline aquifers. Among other criteria, these storage solutions must ensure storage safety and prevent leakage. The ability of a caprock to prevent fluid from flowing out of the reservoir is, thus, of utmost importance. In this review, the main factors influencing fluid flow are examined. These are the wettability of the caprock formation, the interfacial tension (IFT) between the rock and the gas or liquid phases, and the ability of gases to diffuse through it. To achieve effective sealing, the caprock formation should possess low porosity, a disconnected or highly complicated pore system, low permeability, and remain strongly water-wet regardless of pressure and temperature conditions. In addition, it must exhibit low rock–liquid IFT, while presenting high rock–gas and liquid–gas IFT. Finally, the effective diffusion coefficient should be the lowest possible. Among all of the currently reviewed formations and minerals, the evaporites, low-organic-content shales, mudstones, muscovite, clays, and anhydrite have been identified as highly effective caprocks, offering excellent sealing capabilities and preventing hydrogen leakages.

1. Introduction

As climate change continues to challenge countries and scientists worldwide, and with CO₂ emissions hitting a record 37.8 Gt in 2024 [1], the European Union is committed to significantly increase hydrogen’s contribution to the energy sector. The target set by the EU is for hydrogen to contribute 10% of its energy needs by 2050 [2]. As a result, safely storing large volumes of hydrogen has become essential, with underground (or subsurface) hydrogen storage (UHS) emerging as a reasonable solution.

Underground gas storage (UGS) for natural gas has already been applied successfully for almost a century now. The first underground gas storage site worldwide was completed in 1915 in Ontario (a gas field in Welland). The first aquifer used for gas storage was in Kentucky in 1946, and the first solution-mined cavern was made in 1961 in Michigan [3,4]. Thus, there is extensive history and experience regarding UGS, which can also be partly applied in UHS. UHS can be implemented in various geological formations similarly to UGS, such as in salt formations (Figure 1), depleted gas reservoirs, and saline aquifers [5,6]. In general, successful site selection requires a thorough evaluation of several criteria, including fluid flow, geochemical reactions, microbial presence, geomechanical stability of both reservoir and caprock formations, and the sealing ability of the surrounding formations [7]. Nevertheless, caution is needed, since the properties of hydrogen differ significantly from those of methane and carbon dioxide. Therefore, the aforementioned criteria should be carefully evaluated to ensure site selection the safest way possible. Among hydrogen’s properties, its much smaller molecule and high diffusivity, raise concerns about the effective sealing ability of caprock formations [8,9].

Effective sealing of the reservoir is of great importance, as leakages can cause economic disaster for the entire project and pose a severe threat to the safety of people and infrastructure, in addition to environmental concerns. Even when the chosen UHS site is a depleted gas field, effective containment is not guaranteed as hydrogen molecules are smaller compared to methane. As a result, specific investigation and evaluation of caprock properties with respect to hydrogen as a fluid should be performed prior to the final decision. From a fluid-flow perspective, caprock integrity is influenced mainly by its physical properties, such as its porosity, permeability, wettability, interfacial tension, and diffusivity [4,8,10]. Recently, many studies examined the effects of these properties on caprock integrity. Focus has been given to hydrogen’s influence on rock wettability and interfacial tension, while fewer studies provide experimental data on diffusion coefficients [10,11,12]. These parameters were found to be mostly influenced by pressure, temperature, the presence of organic material, and microbial action, and to a lesser extent by gas mixture and salinity [11,13,14,15,16]. Experimental data revealed that, depending on the rock type, fluctuations of the aforementioned factors can alter the formation’s wettability, thus presenting a completely different behavior [17,18].

Ongoing research continues to shed light on these aspects, addressing numerous uncertainties and enhancing our understanding of hydrogen behavior under simulated underground conditions. However, substantial opportunities remain for further investigation to establish a more comprehensive and precise understanding of this applied field of science.

The scope of this work is to summarize and review the key parameters (i.e., wettability, interfacial tension, and diffusion) that ensure storage safety, along with the available experimental data so as to demonstrate the impact of pressure, temperature, organic material, and other parameters, such as salinity and microbial action, on the caprock of a potential hydrogen reservoir.

2. Porosity and Permeability

As is well known in the oil and gas industry, porosity and permeability are key factors influencing the fluid flow and sealing efficiency of porous media. The higher their values, the easier the fluid passes through the porous media. The sealing ability of caprock is mainly attributed to the low values these parameters present, otherwise, gas would escape.

Porosity refers to the void space that was created during the sediment’s deposition. It has to be mentioned however, that the pores are not necessarily interconnected with each other, and, as a result, fluid cannot reach and fill the isolated pores. Regarding fluid flow, effective porosity is used for engineering calculations, and its value is obtained from the following equation [19]:

$${\mathrm{\phi }}_{\mathrm{e}}={\displaystyle \frac{\mathrm{interconnected} \mathrm{pore} \mathrm{volume}}{\mathrm{bulk} \mathrm{volume}}},$$

(1)

A formation presenting high porosity values could possibly act as a sealing formation, provided that a very small proportion of these pores are connected to each other. Another parameter that positively influences hydrogen’s entrapment is the tortuosity of the pore network. Tortuosity accounts for the deviation from the straight line due to the complicated pore network. This impedes the fluid flow, providing trapping, and is calculated as follows:

$$\mathrm{\tau }= {\displaystyle \frac{\mathrm{L}}{{\mathrm{L}}_0}},$$

(2)

where L is the length of the path that fluid follows between two points, whereas L₀ is the straight distance between these two points [19]. Tortuosity mainly prevents leakages caused by diffusion, but also increases the travel times of molecules between two points.

Permeability refers to a formation’s ability to conduct fluid through its pores and affects the direction and the rate of fluid flow. This term was introduced by Henry Darcy and can be expressed through Darcy’s law. By rearranging the Darcy flow equation, permeability (mD) can be calculated as follows [19]:

$$\mathrm{k}={\displaystyle \frac{\mathrm{q}\mathrm{\mu }\mathrm{L}}{\mathrm{A}\Delta \mathrm{p}}},$$

(3)

where q is the flow rate through the porous medium (cm³/sec), μ is the viscosity of the flowing fluid (cp), L is the length of the core sample (cm), A is the cross-sectional area across which flow occurs (cm²), and Δp is the pressure difference (atm).

The effective porosity and permeability values of the caprock are, by default, too low, e.g., in shales. Based on a review of relevant publications, the following representative values for these parameters are given in

Table 1. Porosity values range for different rocks/minerals.

Rock/Mineral	Porosity (%)	Reference
Evaporite	0.14 to 7.7	[22,23]
Salt (Halite)	0.04 to 0.9	[24]
Shales	0.11 to 14	[23,25,26]
Mudstone	2 to 8.6	[23,27]
Unfractured metamorphic and igneous rocks	0 to 5	[26]

and Table 2. Typical caprock formations, such as salts, shales, mudstones, and evaporites, present porosity values that are low enough. Granites, basalt, and gneiss formations present low values as well, and could be used as caprocks, although these formations are frequently fractured in the field. Effective seals for hydrogen can also be documented from natural hydrogen accumulations, where rocks such as dolerites and shales have been found to effectively confine the natural hydrogen reservoirs, under the condition that their pore throat radii are less than 100 nm [20]. The well-known Mali dolerite is the best example of the kind of rock that delivers effective containment [21].

Apart from porosity, effective sealing is achieved when permeability values are low enough. Generally, to effectively confine hydrocarbons, a reservoir’s caprock permeability values should be in the range of 10⁻³ to 10⁻⁵ mD [28]. As Table 2 illustrates, the permeability values of typical caprock formations lying either between or well below those limits with those of Salt/Evaporites being extremely low.

Table 2. Permeability values for different rocks/minerals.

Rock/Mineral	Permeability (mD)	Reference
Evaporite	10−9 to 10−3	[29]
Salt (Halite)	10−8 to 10−4	[30]
Shales	9 × 10−6 to 6 × 10−3	[25,31]
Mudstone	2 × 10−7 to 2 × 10−1	[32]
Unfractured metamorphic and igneous rocks	10−6 to 10−4	[26]

Table 2. Permeability values for different rocks/minerals.

Rock/Mineral	Permeability (mD)	Reference
Evaporite	10−9 to 10−3	[29]
Salt (Halite)	10−8 to 10−4	[30]
Shales	9 × 10−6 to 6 × 10−3	[25,31]
Mudstone	2 × 10−7 to 2 × 10−1	[32]
Unfractured metamorphic and igneous rocks	10−6 to 10−4	[26]

Furthermore, studies such as [25,33] spotlight the synergistic effect of low porosity and permeability, associating them with elevated breakthrough pressures that hydrogen has to overcome, leading to more effective sealing. Although porosity and permeability values are thought to be constant throughout an underground gas storage project [19], this may not be the case in UHS due to hydrogen’s great reactivity and the potential reactions with rock minerals.

Geochemical reactions in UHS hold a key role in the integrity of caprocks as they can modify their porosity both ways: either create new flow paths for hydrogen to escape (increase in porosity) or induce mineral precipitation leading to reduced porosity [34,35]. In nature, both procedures occur simultaneously. Due to these effects, the overall behavior of the caprock can change over time and should be monitored so as to avoid leakages, which will lead to project failure.

Geochemical modeling and simulations conducted over a seven-year period for hydrogen storage in porous media [36], revealed that acidity of the brine promotes mineral dissolution and precipitation (particularly of calcite), leading to porosity reduction and subsequent permeability decrease, maintaining the containment capacity of the caprock. In contrast, under alkaline conditions, the moderate changes in mineral alteration retain the initial sealing efficiency. Further, the authors [36] note that both high pressure and high salinity promote caprock integrity by impeding mineral dissolution and salt precipitation, respectively, while temperature should be taken into serious consideration, since it appears to be rather equivocal.

Other geochemical modeling work of ref. [37] using shales of Ordovician age from western Australia demonstrated, that geochemical reactions in the framework of an H₂–brine–rock system have a minor effect on the sealing efficiency of the caprocks, highlighting the importance of its thickness. In fact, simulation results indicated that a caprock can contain hydrogen effectively since a very small percentage of the overall volume, less than 1%, intrudes for a few meters (modeling run for a project’s lifespan of three decades).

It seems that differences in the composition and type of the rocks/minerals used for the simulations, as well as changes in the temperature, pressure, and salinity conditions, provide different results; therefore, both modeling data and experimental laboratory tests should be case-tailored.

3. Wettability

Wettability refers to the tendency of a fluid to maintain contact with a solid surface and plays a crucial role in underground hydrogen storage (UHS). Wettability is determined based on the measured value of the contact angle.

When a drop of liquid lies onto a solid surface in a gas environment, the angle that is created between the solid surface and the gas-liquid interface, measured towards the center of the liquid drop, is known as the contact angle (Figure 2). The value of this angle is defined by the magnitude of solid–liquid, liquid–gas, and solid–gas interfacial tensions.

Contact angle (CA) defines the wetting preference of a surface towards a fluid compared to others, varying from 0°(complete wetting) to 180° (complete non-wetting). When a formation presents a CA value (θ) less than 90°, then it is considered as water-wet and intermediate-wet when θ = 90°.Contact angle measurements are performed either by measuring the equilibrium contact angle or the advancing and receding contact angles (e.g., sessile drop and tilting plate method).

Equilibrium contact angle is measured under static conditions, and it is representative of the wetting behavior of the rock in no-flow conditions. On the other hand, dynamic contact angles are representative of the wetting behavior a rock would exhibit when flow occurs [30]. As a rule of thumb, the equilibrium CA value lies in between a maxima and a minima represented by the advancing and receding CA, respectively [30].

Advancing contact angle resembles the imbibition process, where the non-wetting phase is displaced by the wetting phase [19]. For a water-wet formation, this could be the case when hydrogen is withdrawn and brine imbibes the pores, entrapping hydrogen (residual trapping) [18].

Receding contact angle resembles the drainage process, where the wetting phase is displaced from the non-wetting phase. For a water-wet formation, this could be the case of hydrogen injection displacing the formation brine [18]. It must be mentioned that receding contact angle values are lower than the advancing values due to the contact angle hysteresis phenomenon [38].

The contact angle, and thus the wettability of the formation, also affects the maximum column storage height as follows:

$$\mathrm{h}={\displaystyle \frac{2{\mathrm{\gamma }}_{\lg }\cos \mathrm{\theta }}{\Delta \mathrm{\rho }\mathrm{gr}}},$$

(4)

where γ_lg is the liquid–gas interfacial tension, Δρ is the density difference between the two phases, θ is the contact angle, g is the gravitational acceleration, and r is the pore radius. In Equation (4), it is generally suggested to use the receding contact angle to avoid overestimating storage height [39].

Regarding hydrogen storage, water-wet formations are preferred because they ensure storage safety due to residual trapping. In this case, hydrogen will be expelled away from the rock’s surface towards the centers of the pores. Additionally, since a water-wet formation presents a higher capillary entry pressure, hydrogen should exceed this pressure threshold in order to penetrate the caprock, resulting in leakage [35]. Thus, leakage will not occur as easily because hydrogen will be away from the caprock surface. However, wettability can be influenced and physically altered by changes in pressure and temperature, salinity levels, and the composition of gases present in the reservoir.

3.1. Pressure

Table 3. Compilation of available data regarding the pressure variation influence on contact angles.

Rock/Mineral	Temperature (°C)	Pressure (MPa)	Equilibrium CA (°)	Advancing CA (°)	Receding CA (°)	Reference
Basalt	35	5 to 20	-	32.29 to 59.31	~29.4 to 56.9	[39]
	70			47.86 to 68.61	~43.5 to 65.3
S. Arabia Basalt	50	5 to 20		19.3 to 42.1	13.2 to 36.3	[18]
Anhydrate	30	3.44 to 17.23	18 to 20.5	-	-	[11]
	75		19.5 to 17.5
Quartz	40	1 to 10	45 to 51			[17]
Calcite			40 to 59
Basalt			21 to 27
Granite			27 to 48
Shale			39 to 52
Anhydrite			40 to 43
Gypsum			48 to 52
Wolf Camp Shale TOC < 0.3%	50	1.37 to 6.89	65 to 60	-	-	[45]
Eagle Ford Shale TOC = 3.83%			90 to 80
Jordanian Oil Shale (TOC = 13%)	50	0.34 to 11.03	-	43 to 79	43 to 76	[41]
Mica	50	5 to 20	-	21.7 to 42.9	18.3 to 36.6	[40]
Shale TOC = 0.08%	25	5 to 20	-	~26.8 to 47	~21.9 to 41.6	[47]
Shale TOC = 0.1%				~18.3 to 35.1	~15.3 to 30.3
Shale TOC = 0.09%				~14.8 to 21.3	~12.1 to 18.2
Evaporite				~11.7 to 16.3	~9.6 to 13.7
Shale TOC = 0.08%	80		-	~17.3 to 30.4	~14.3 to 27.9
Shale TOC = 0.1%				~10.6 to 20.2	~7.5 to 15.3
Shale TOC = 0.09%				~10 to 14.6	~7.6 to 12.9
Evaporite				~7.9 to 12.7	~5.9 to 9.3
Kaolinite	60	5 to 20	13.4 to 26	-	-	[48]
Illite			16.3 to 31.7
Montmorillonite			19.8 to 38.6

is summarizing all available data regarding the influence of pressure on contact angles for typical caprock formations and minerals. Advancing and receding contact angles were measured by many researchers, such as Ali et al., Hosseini et al., Alanazi et al. and Iglauer et al. [18,39,40,41,42]. On the other hand, Aghaei et al., Higgs et al., Esfandyari et al., Al-Mukainah et al., Hashemi et al. [11,43,44,45,46], and Aftab et al. [13] have provided results for static contact angles.

It can be easily assumed that the contact angle is sensitive to pressure variations and generally increases with increasing pressure. However, this increase remains within the range that preserves the water-wet condition of the formations, even at higher pressures. More specifically, by increasing pressure, Hosseini’s basalt specimen, Saudi Arabian basalt, granite, mica, and Jordanian oil shale presented the greatest increase in advancing contact angles, ranging from 83% (Hosseini’s basalt at 35 °C) up to 118% for Saudi Arabian basalt (Table 3). Receding contact angles also increased with pressure variation and seemed to be more affected, presenting a mean increase of 12%. Percentage values have been calculated in the present work based on published data in Table 3. For most of the samples, higher pressures resulted in a wider range between advancing and receding contact angle values. An exception to this trend was found in the basalt samples and the shale samples of TOC < 0.1%, whose range between advancing and receding contact angle was reduced. It should be noted that for the same rock type, different studies, e.g., [17,41,45], may provide different results, which are attributed to different mineralogical compositions of the samples. For example, by increasing pressure, Hosseini’s basalt became weakly water-wet, whereas Ali’s basalt remained strongly water-wet. This divergent behavior was attributed to differences in plagioclase composition, as mentioned by Ali et al. [18].

Contact angles not only influence the formation of wettability but also the maximum hydrogen column height that can be securely stored, as Equation (3) indicates. Hosseini et al. [39], in their research, calculated the maximum hydrogen storage height by increasing pressure, resulting in a smaller hydrogen column to be stored in elevated pressures and temperatures (from 2141 m column at 5 Mpa and 35 °C to 928 m at 20 Mpa and 70 °C) as the formation turned to weakly water-wet. Also, Al-Mukainah et al. [45] calculated the maximum hydrogen storage height for Eagle Ford (high) and Wolf Camp (low) TOC shale, resulting in a low TOC shale that could safely store up to 325m, whereas the high TOC shale could store up to 100m.

3.2. Temperature

Although the temperature at a specific site of a UHS project is fixed for a certain depth, it varies relevant to the geothermal gradient. As

Table 4. Compilation of available data regarding temperature influence on contact angle.

Rock/Mineral	Temperature (°C)	Pressure (MPa)	Equilibrium CA (°)	Advancing CA (°)	Receding CA (°)	Reference
Basalt	35 to 70	5	-	32.29 to 47.86	~29.4 to 43.5	[39]
		20		59.31 to 68.61	~56.9 to 65.3
S. Arabia Basalt	25 to 50	20	-	38.5 to 42.1	33.2 to 36.3	[18]
Anydrate	30 to 75	3.44	18 to 19.5	-	-	[11]
		10.34	21 to 19.5
		17.23	20.5 to 17.5
Mica	35 to 70	15	-	53.1 to 35.4	47.3 to 29.2	[49]
Calcite	25 to 80	15	-	80.35 to 57.85	76.6 to 57.85	[50]
Calcite	20 to 80	10	40 to 93	-	-	[17]
Quartz			40 to 73
Basalt			17 to 35
Granite			30 to 67
Shale			38 to 81
Anhydrite			38 to 88
Gypsum			45 to 71
Shale TOC = 0.08%	25 to 80	20	-	~47 to 30.4	~41.6 to 27.9	[47]
Shale TOC = 0.1%				~35.1 to 20.2	~30.3 to 15.3
Shale TOC = 0.09%				~21.3 to 14.6	~18.2 to 12.9
Evaporite				~16.3 to 12.7	~13.7 to 9.3

reveals, by increasing temperature (i.e., at different depths of a certain location), some formations/rocks tend to become more hydrophobic while others tend to become more hydrophilic. The following percentage values have been calculated in the present work based on published data of Table 4.

More specifically, calcite and anhydrite studied by Esfandyari et al. [17] presented the maximum increase in their equilibrium contact angles, by 132% and 131%, respectively (with a temperature rise of 60 °C), while calcite even altered its wettability. Similarly, basalt and shale exhibit a significant increase in their equilibrium contact angles that equals 105% and 113%, respectively (as a result of the same temperature rise as above).On the contrary, according to [47], shales with various TOC values ranging between 0.1% and 0.08%, as well as an evaporite sample, all faced a significant decrease in their advancing contact angle values equal to 40%, 42%, and 33%, respectively (under a temperature increase of 55 °C).Generally, receding contact angle values follow the same trend and similar magnitude of change with advancing contact angle values when considering the same rock sample.

3.3. Presence of Organic Material and Bacteria

Organic material can alter the wettability of rock surfaces, as the presence of organics has been shown to reduce the hydrophilicity of rocks, and it is particularly significant in clastic rocks such as sandstone and shale, which can adsorb organic materials by interacting with reservoir fluids [11].To account for organic contamination on rock surfaces, many researchers applied acid aging to their samples with different types of acids and various concentrations, thereby more accurately representing the reservoir conditions. In

Table 5. Compilation of available data regarding the organic content’s influence on contact angle at 50 °C.

Rock/Mineral	Pressure (MPa)	Acid Concentration (Mol/lt)	Acid	Equilibrium CA (°)	Advancing CA (°)	Receding CA (°)	Reference
Basalt	15	10−9 to 10−2	Stearic	-	~73.09 to 92.29	67.09 to 86.29	[39]
S. Arabia Basalt	5	0 to 10−2	Stearic	-	19.3 to 78.4	13.2 to 72.3	[18]
	5 to 25	10−2			78.4 to 100.8	72.3 to 94.2
Mica	15	0 to 10−9	Lignoceric	-	~42.9 to 63.2	36.6 to ~56.2	[49]
		10−9 to 10−2			~63.2 to 91.8	~56.2 to 84
	25	10−2	Hexanoic to lauric		~67.5 to 89.2	~74.5 to 83.8
Mica	15	10−9 to 10−2	Stearic	-	53.2 to 84.6	48.7 to 76.4	[40]
	15 to 25	10−2	Stearic	-	84.6 to 98.8	76.4 to 90.8
Calcite	10	10−9 to 10−2	Stearic	-	75.85 to 115.85	68.7 to 110.85	[50]
Shale(Wolf Camp, and Eagle Ford)	6.89	TOC < 0.3% and 3.83%	Not aged	60 and 90	-	-	[45]
Shale TOC = 0.08%	15	10−9 to 10−2	Stearic	-	~39.8 to 76.2	~34.6 to 69.3	[47]
Shale TOC = 0.1%					~31.5 to 57.9	~27.8 to 51.8
Shale TOC = 0.09%					~26.7 to 55.4	~22.63 to 51.7
Evaporite					~16.4to 42.8	~11.8 to 40.5

, the influence of organic material on the contact angles of different rock samples is illustrated.

In general, rocks with a high content of organic material, such as shales, present high contact angles and intermediate wettability [41,45]. Moreover, the contact angle exhibits a positive correlation with total organic content (TOC), as an increase in TOC leads to a more hydrophobic formation, reducing the wettability of the surface. It has been shown that an increase of 5% in TOC for the Jordanian oil shale resulted in a 10° and 15° rise in advancing and receding contact angles, respectively [33]. This also applies to aged samples, as shifting from zero or lower acidic concentrations to higher ones generally increases contact angle values. From the formations presented in Table 5, Saudi Arabian basalt was largely influenced by a rise in organic content, as it resulted in an increment of 59° in the advancing contact angle [18]. Generally, by increasing the acid concentration, shales and evaporites presented significant increases in their contact angles [47]. Despite this increase, shales of low TOC and the studied evaporite remained strongly water-wet, even under higher acidic concentrations, while all the other samples turned out to be intermediate- or hydrogen-wet. Furthermore, it was revealed that mica’s wettability was also sensitive to alkyl chain length [49]. At the same acid concentration, the mica sample treated with lauric acid resulted in higher contact angle values than when treated with hexanoic acid, representing a difference of 24.32%.

Recent work has examined the use of various concentrations of silica nanofluid in order to reverse the wettability alteration due to the presence of high TOC. The method was successfully tested on Jordanian shale samples bearing high TOC to enhance sealing performance and subsequently ensure hydrogen storage safety [51].

Apart from organic content, the presence of bacteria in the reservoir through their metabolism can significantly influence rock properties, such as the wettability of the reservoir and the caprock formations (

Table 6. Compilation of available data regarding microbial action’s influence on contact angles for untreated and treated samples.

Rock/Mineral	Temperature (°C)	Pressure (MPa)	Untreated	Treated	Reference
Quartz	50	0.1	37.8°	54.2°	[13]
		27	5.8°	14.4°
Water-wet quartz	50	13	85°	95°	[53]
Oil-wet quartz			105°	90°
Basalt	50	27	19.5 °	69°	[52]
Calcite	50	8	57 °	40°	[54]

). Aftab et al. [13] found that, bacterial activity caused an increase in the contact angle of quartz by almost 10°, while its effect was more prominent on basalt, which went from being strongly water-wet to weak water-wet [52]. Similar findings occurred in research by Ali et al. [53] regarding the water-wet quartz, which altered its wettability and became hydrogen-wet, whereas the oil-wet quartz presented the opposite tendency. Nevertheless, some minerals, such as calcite and dolomite, enhanced their hydrophilicity, presenting decreased contact angles, following microbial aging [54]. The authors also mentioned that quartz and gypsum followed the same trend but exhibited smaller changes.

3.4. Salinity

By increasing salinity, the contact angle generally increases [55]. Moreover, ref. [56] revealed in their study that the cosine of the contact angle decreases linearly with increasing salinity and, as a result, can be easily calculated. However, this increase is negligible and has a rather slight effect on the advancing contact angle values, as compared to pressure, temperature, and organic content (

Table 7. Compilation of available data regarding salinity’s influence on contact angle (DI stands for de-ionized water).

Rock/Mineral	Temperature and Pressure	Salinity	Advancing CA (°)	Receding CA (°)	Reference
Calcite	50 °C and 15 MPa	0 to 4.9 mol/kg	69.8 to 80.65	63.35 to 73.3	[50]
Calcite	40 °C and 10 MPa	DI to formation brine	51 to 59	-	[17]
Quartz			45 to 51
Basalt			24 to 27
Granite			27 to 48
Shale			46 to 52
Anhydrite			50 to 43
Gypsum			51 to 52

). Out of the rock types mentioned in the literature and illustrated below, only granite presented a remarkable change, altering its contact angle by 77% (CA raised from 27 to 48°).

3.5. Gas Mixture

The type of fluids interacting with the rock surface, such as hydrogen, brine, or other gases (such as CO₂ and CH₄) that are present in the reservoir, may affect formation wettability. As it is documented in the literature (

Table 8. Compilation of available data regarding gas composition’s influence on contact angle.

Rock/Mineral	Temperature and Pressure	Gas Mixture	Equilibrium CA (°)	Advancing CA (°)	Receding CA (°)	Reference
Jordanian oil shale (TOC 13%)	50 °C+ 11.03 MPa	Pure H2	-	79	76	[41]
		Pure CH4		103	88
		H2-CH4		99	88
Anhydrite	75 °C+ 3.44–17.23 MPa	Pure H2	23 (stable)	-	-	[14]
		H2-CH4	23 (stable)
		H2-CO2	24 (stable)
Calcite	75 °C+ 3.44–17.23 MPa	Pure H2	24 to 25			[14]
		H2-CH4	26 (stable)
		H2-CO2	~27 to 26

), in the presence of pure hydrogen, the Jordanian oil shale is weak water-wet, but when methane is added to the mixture, the rock alters its wettability to hydrogen-wet. Contrary to that, other rock types, such as the anhydrite and calcite, under both mixture and pressure changes, remained water-wet, and their contact angles were found to be almost stable as well. It is noteworthy to mention that the Jordanian shale has a high TOC value (equal to 13%) that may act synergistically with the change in the wettability.

4. Interfacial Tension

The interfacial tension (IFT) between hydrogen and various subsurface fluids and rocks plays a crucial role in determining the efficiency and safety of UHS. IFT is closely related to the contact angle and the wettability of rocks, as Young’s equation suggests:

$$\cos \mathrm{\theta }= {\displaystyle \frac{{\mathrm{\gamma }}_{\mathrm{sg}}-{\mathrm{\gamma }}_{\mathrm{sl}}}{{\mathrm{\gamma }}_{\lg }}}$$

(5)

where θ is the equilibrium contact angle, γ_sg is the solid–gas interfacial tension, γ_sl is the solid–liquid interfacial tension, and γ_lg is the liquid–gas interfacial tension.

A strong water-wet formation requires the fraction of Equation (5) to range between 1 (contact angle = 0°) and 0.5 (contact angle = 60°), while values of 0 (contact angle = 90°) as well as negative values stand for neutral or gas-wet formations and should be avoided. As a result, changes in IFTs of the rock–brine–gas system cause changes in formation wettability, which may have a detrimental effect on storage safety.

At the same time liquid–gas IFT plays a crucial role for capillary trapping (Equations (5) and (6)) and, by extension, to the maximum safe storage gas column height as Equation (4) indicates:

$${\mathrm{P}}_{\mathrm{c}}={\displaystyle \frac{2{\mathrm{\gamma }}_{\lg }\cos \mathrm{\theta }}{\mathrm{r}}}$$

(6)

where P_c is the capillary pressure, γ_lg is the liquid–gas IFT, θ is the contact angle, and r is the pore radius. As a result, increased liquid–gas IFT could improve capillary trapping of hydrogen and enhance storage safety as the capillary entry pressure becomes higher, preventing gas leakage through the caprock. Many studies have examined the behavior of IFTs under different pressures, temperatures, salinity, and organic content conditions, and in the presence of both working (H₂) and cushion gas (CO₂ or CH₄).

4.1. Pressure

Τhe rock–gas interfacial tension is significantly more affected by pressure variation than the rock–liquid or liquid–gas IFT (

Table 9. Compilation of available data regarding the pressure’s influence on rock–liquid and rock–gas interfacial tension (IFT).

Rock	Temperature (°C)	Pressure (MPa)	Rock–Liquid IFT (mN/m)	Rock–Gas IFT (mN/m)	Reference
Calcite	60	1 to 10	52.66 (stable)	98.84 to 69.52	[44]
Quartz			42.04 (stable)	84.57 to 73.11
Basalt			11.25 (stable)	68.37 to 64.86
Granite			39.55 (stable)	85.15 to 74.85
Shale			51.78 (stable)	98.96 to 72.31
Anhydrite			47.17 (stable)	91.18 to 74.53
Gypsum			39.95 (stable)	81.31 to 71.32
Montmorillonite	60	5 to 20	-	67.26 to 58.15	[58]
Illite				67.89 to 59.78
Kaolinite				68.64 to 61.2
Muscovite	50	5 to 20	~49 (stable)	115 to 95	[57]
	70		~43 (stable)	104 to 90

). As the studies by Esfandyari et al., Ali et al. and Yekeen et al. [44,57,58] have shown, rock–liquid interfacial tension is only slightly or not at all influenced by pressure variation. This happens because brine is generally incompressible and, as a result, pressure variation has no significant effect on intermolecular forces.

The rock–gas interfacial tension decreases with increasing pressure. This reduction varies from 5.13% to 24.66% (percentage values have been calculated in the present work based on the available published data in Table 9) and is more pronounced for calcite and shale, whereas basalt and gypsum rock–gas IFT were only slightly affected. This phenomenon is caused by the increased gas density, which, in turn, leads to more gas molecules adsorbing onto the solid surface, expelling the liquid from it. Thus, higher contact angles are observed, and the formation becomes more hydrophobic.

In addition, the pressure increment affects the liquid–gas IFT, which, in general, reduces significantly (from 2% to 55%) due to the higher gas density (

Table 10. Compilation of available data regarding the pressure’s influence on liquid–gas interfacial tension (IFT).

Temperature (°C)	Pressure (MPa)	Mixture or Rock Presence	Liquid–Gas IFT (mN/m)	Reference
25	2.76 to 34.47	H2 + brine	80.77 to 75	[60]
150			58 to 56
60	5 to 20	Kaolinite + H2 + brine	67.02 to 65.53	[58]
		Montmorillonite + H2 + brine	61.02 to 65.53
		Illite + H2 + brine	67.02 to 65.53
50	1.37 to 11.03	H2 + brine	55 to 53	[61]
50	0.1 to 20		83.48 to 65.15
70			78.93 to 62.59
20	1 to 10	H2 + brine	69.07 to 30.97	[17]
40			77.52 to 36.44
60			78.77 to 39.6
80	1 to 7		82.07 to 58.31
25	2 to 40	H2 + brine	73 to 69.1	[59]
50			69.3 to 65.8
100			59.7 to 57.6
175			44.1 to 43.2

), leading to an increase in contact angle values. Such changes make formations less water-wet and result in a limited storage gas column height. Ref. [59] mentions that liquid–gas IFT values at low pressures and temperatures (up to 5 MPa at 25 °C and 1 MPa at 50 °C) present a positive correlation, with increasing pressure reaching a maximum value, and then exhibit a negative correlation, with decreasing values of liquid–gas IFT (Table 10).

4.2. Temperature

Unlike pressure variations, temperature seems to have a more pronounced effect on solid–liquid interfacial tension, which increases significantly with rising temperature (

Table 11. Compilation of available data regarding the temperature’s influence on rock–liquid and rock–gas interfacial tension (IFT).

Rock/Mineral	Temperature (°C)	Pressure (MPa)	Rock–Liquid IFT (mN/m)	Rock–Gas IFT (mN/m)	Reference
Calcite	20 to 80	4	23.32 to 52.92	73.69 to 84.52	[44]
Quartz			29.64 to 42.84	79.66 to 76.28
Basalt			0.01 to 19.47	60.35 to 71.75
Granite			31.66 to 43.17	88.02 to 78.26
Shale			44.95 to 50.68	99.34 to 86.6
Anhydrite			12.92 to 50.45	63.8 to 75.75
Gypsum			0.76 to 38.5	45.38 to 64.4
Muscovite	35 to 70	5	~58 to 43	~124 to~104	[57]

) [44,57]. This means that as temperature increases, the minerals become less hydrophilic, resulting in higher contact angles. Strongly water-wet minerals, such as basalt and gypsum, presented the most significant increase in their solid–liquid interfacial tension compared to other minerals, implying that although they become less water-wet, they remain hydrophilic. However, there are exceptions, such as the muscovite, which exhibited the opposite tendency-its rock–liquid interfacial tension decreased with a temperature rise, making it more hydrophilic.

The temperature effect on the rock–gas IFT is not straightforward, as minerals do not present a uniform tendency. The increase in IFT was more pronounced for gypsum, anhydrite, and basalt. On the contrary, a significant reduction in rock–gas IFT was presented for shale, granite, quartz, and muscovite with increasing temperature [44]. As noted by the authors in [44], the differing responses of minerals to temperature changes is probably attributed to the different mineralogy that was tested.

Liquid–gas IFT values generally decrease with a temperature rise. If we calculate the percentage reduction from

Table 12. Compilation of available data regarding the liquid–gas mixture influence on liquid–gas interfacial tension (IFT).

Temperature (°C)	Pressure (MPa)	Liquid–Gas IFT (mN/m)	Reference
25 to 150	2.76	80.77 to 58	[60]
	34.47	75 to 56
50 to 70	20	65.15 to 62.59	[61]
	0.1	83.48 to78.93
20 to 40	1	63.6 to 62.6	[17]
60 to 80		60.3 to 58.5
25 to 175	2	73 to 44.1	[59]
	40	69.1 to 43.2

, it is found that it ranges from 2% to 40%. The reduction in IFT values with an isobaric increase in temperature is mainly attributed to the decrease in liquid density which, in turn, reduces the density difference between the two phases [61]. Additionally, results that were produced under pragmatic subsurface conditions point in the same direction [17], i.e., by increasing temperature, the IFT between H₂ and formation brine decreases at constant pressure (Table 12). A comparable effect is observed with pressure changes; for instance, at 20 °C the corresponding IFT is 63.6 and 62.8 at 1MPa and 10MPa, respectively).

4.3. Organic Content

The effect of stearic acid concentration on the interfacial tension relevant to the mineral quartz was studied by [62]. It was shown that under the same thermophysical conditions (at 25 MPa pressure and at 50 °C temperature, given as 323K), a higher concentration of organic acid molecules increases the rock–liquid IFT (γ_SL). As the authors noted, the quartz used was aged in a certain organic acid for one year and then aged in a silica nanofluid.

Ali et al. [57] found that the different concentrations of organic content influence both the mica’s rock–liquid and rock–gas IFT. According to these authors, solid–gas IFT (γ_SG) decreases with higher organic/alkyl chain concentration (by increasing hydrophobicity), while, by increasing the acid concentration, rock–liquid IFT increases. Moreover, inclusion of longer alkyl chains (transitioning from hexanoic to lignoceric acid) in the substrates results in higher rock–liquid IFT values.

According to [63], in the calcite mineral studied, an increase in the stearic acid concentration at constant temperature and pressure conditions, resulted in a 5.1 mN/m increase in the rock–water IFT (γ_SL) (see below

Table 13. Compilation of available data regarding the acidic concentration’s influence on rock–liquid interfacial tension (IFT).

Temperature (°C)	Pressure (MPa)	Concentration (mol/lt)	Acid	Rock–Liquid IFT (mN/m)	Reference
50	15	10−9 to 10−2	Lignoceric	51 to 53	[57]
50	25	10−3	Hexanoic to lignoceric	53 to 58
50	10	10−9 to 10−2	Stearic	51 to 56.1	[63]
50	25	10−9 to 10−2	Stearic	41 to 54	[62]

).

4.4. Salinity

Salinity of the brine and its composition have been found to influence interfacial tension [17,60,61,64]. The salinity effects are more profound at higher temperatures, although at lower salinities, temperature changes are more intense. Salinity rising leads to higher rock–liquid IFT (see below

Table 14. Compilation of available data regarding the salinity’s influence on rock–liquid and rock–gas interfacial tension (IFT) [44]. DI stands for de-ionized water.

Rock/Mineral	Temperature (°C)	Pressure (MPa)	Salinity	Rock–Liquid IFT (mN/m)	Rock–Gas IFT (mN/m)
Calcite	60	7	DI to formation water	48.54 to 52.66	56.57 to 78.69
Quartz				34.54 to 42.04	80.88 to 76.82
Basalt				1.98 to 11.25	63.01 to 66.02
Granite				16.9 to 39.59	70.95 to 78.2
Shale				38.22 to 51.78	84.38 to 80.7
Anhydrite				38.72 to 47.17	76.89 to 79.88
Gypsum				13.68 to 39.95	57.69 to 74.57

). This happens due to the presence of ions in brine, which also interact with minerals and may slightly alter their wettability preferences [44]. If we consider the values of Table 14, it is obvious that salinity does not affect rock–gas IFT uniformly.

The same effect is also observed in the liquid–gas IFT, which increases with increasing salinity (see below

Table 15. Compilation of available data regarding the salinity’s influence on liquid–gas interfacial tension (IFT).

Temperature (°C)	Pressure (MPa)	Brine Composition	Salinity (mol/kg)	Liquid–Gas IFT (mN/m)	Reference
50	10	H2 + brine (NaCl)	0.96 to 2.93	68.56 to 71.43	[61]
		H2 + brine (CaCl2)	0.93 to 2.4	70.57 to 82.02
25	2.76	H2 + brine (NaCl + KCl)	0 to 4.95	~73 to ~81	[60]
	34.47			~69 to ~74
150	2.76			~48 to ~58
	34.47			~47 to ~56
20	10	H2+ brine (NaCl + KCl + CaCl2 + MgCl2)	0 to brine	24 to 39.97	[17]
40				35.81 to36.44
60				37.5 to 39.6

). The composition of brine also affects the magnitude of the increment [61]. In a comprehensive assessment using molecular dynamics and machine learning simulations, [65] noted that IFT increased with higher NaCl salinity. However, the experimental work by [60] has shown that some salts, such as CaCl₂, increased the liquid–gas IFT more than others (e.g., NaCl), due to their higher density and molar mass [61]. This phenomenon was more intense at lower pressures.

4.5. Gas Composition

As mentioned earlier, liquid–gas IFT decreases with increasing pressure, and this phenomenon becomes more pronounced in the presence of CO₂ (see below

Table 16. Compilation of available data regarding the gas composition’s influence on liquid–gas interfacial tension (IFT).

Temperature (°C)	Pressure (MPa)	Gas Mixture	Liquid–Gas IFT (mN/m)	Reference
50	5 to 40	100%H2	58.5 to 56.9	[66]
		50%H2 + 50%CO2	57.3 to 54.5
	3.44 to 20.7	50%H2 + 50%CH4	57.4 to 56	[64]
		70%CO2 + 30%H2	54.7 to 32.1
		50%CO2 + 50%H2	60.6 to 38.2
		30%CO2 + 70%H2	65.4 to 51.7
	1.37 to 11.03	100%H2	55 to 53	[41]
		50%H2 + 50%CH4	54.8 to 50
		100%CH4	54 to 46.5
50 to 80	3.44	70%CO2 + 30%H2	54.7 to 47.7	[64]
		50%CO2 + 50%H2	60.6 to 51.5
		30%CO2 + 70%H2	65.4 to 58.9

). Since CO₂ is denser than H₂ under the same pressure and temperature conditions, a gas mixture presents a higher density than pure H₂. Thus, such a gas mixture becomes more sensitive to pressure changes, contrary to pure H₂ [66]. Moreover, the density difference between the liquid (brine) and the gas mixture is reduced, resulting in lower liquid–gas IFT. The same applies for CH₄, but in this case, the phenomenon is not that prominent. This could probably be attributed to the fact that CH₄ has a lower density than CO₂. This reduction in liquid–gas IFT may enhance H₂ recovery, but it also significantly increases the leakage risk, especially at the transition zone between hydrogen and the respective cushion gas [41]. On the other hand, temperature rise led to a decrease in the liquid–gas IFT regardless of the mixture composition that included H₂. This percentage reduction ranges between 10 and 15% (calculated from the values in Table 16).

Furthermore, salinity plays a synergistic role with gas composition in determining IFT; by increasing salinity, the increase in IFT differs for various gas compositions (see below

Table 17. Compilation of available data regarding changes in the gas composition’s influence on liquid–gas interfacial tension (IFT) at different salinities.

Temperature and Pressure	Salinity (mol/kg)	Gas Mixture	Liquid–Gas IFT (mN/m)	Reference
50 °C + 3.44 MPa	0 to 3.15	70%CO2 + 30%H2	51.5 to 60.4	[64]
		50%CO2 + 50%H2	57.9 to 64.6
		30%CO2 + 70%H2	62.4 to 67.8

). A rise in salinity leads to increased brine density, which, in turn, results in creating a larger differential between gas and liquid phases [64]. It can be observed from the corresponding values (Table 16) that the higher the H₂ fraction is, the smaller the value of the IFT increment. This happens because as the hydrogen fraction in the gas mixture becomes higher, the density of the gas phase decreases, creating a huge differential with the liquid phase.

5. Diffusion

Diffusion refers to the process by which a substance moves from higher to lower concentration areas. According to Fick’s law, the diffusive flux of a single compound through porous media is calculated as follows:

$$\mathrm{J}= -{\mathrm{D}}_{\mathrm{eff}}\nabla \mathrm{C}$$

(7)

J is the diffusive flux (mol/m²s), D_eff is the effective diffusion coefficient (m²/s), and ∇C is the concentration gradient (concentration mol/m³).

Movement in porous media does not follow a straight path between two points but instead follows the pathway of interconnected pores. To account for this, the effective diffusion coefficient is used to describe fluid flow through porous media and can be calculated as follows:

$${\mathrm{D}}_{\mathrm{eff}}={\mathrm{D}}_{\mathrm{H}}{\displaystyle \frac{\mathrm{\phi }}{\mathrm{\tau }}}$$

(8)

where D_H is the bulk diffusion, φ is the porosity, and τ is the tortuosity [67].

Diffusion Coefficient

Regarding hydrogen diffusion coefficients for various rock types, scarce experimental data exist. A compilation of published values measured in water-saturated samples is presented in Table 18. These data indicate that the diffusion coefficient of hydrogen generally decreases with increasing pressure and increases with rising temperature [15,16]. Pressure has a more prominent effect on the effective diffusion coefficient, as changes in pressure affect its value by one order of magnitude [15,16]. As pressure increases, gas density increases as well, resulting in higher concentrations and accelerating diffusion [15], with negative results for the containment.

Moreover, formation properties, such as porosity and permeability, also influence hydrogen diffusion, which is positively correlated with porosity, permeability, and tortuosity [68,69]. Following that, shales and mudstones exhibit the lowest diffusion coefficient values of all rock types, due to their small pore size, low permeability, and generally unconnected and tortuous pore systems.

Studies of salt caverns as potential storage sites, using experimental and numerical approaches, found that the effective diffusion coefficient for salt rocks varies from 3.5 10⁻¹⁰ to 1.2 10⁻⁹ m²/s over temperature and pressure ranges of 25 to 45 °C and 2.7 to 6.5 MPa, respectively [70]. By contrast, at the same temperature and pressure ranges as the above, the effective diffusion coefficient values for mudstone ranged between 10⁻⁷ and 1.8 10⁻⁷ m²/s [70].

Yuan et al. [71] tested evaporitic cores from the Lotsberg Formation in Alberta, Canada, using helium gas as a hydrogen equivalent due to their similar molecular size. They tested five different samples at 25 °C and an initial pressure of 6.2 MPa, with different mineralogy representing various depths, as the evaporite sequence consisted of salt (pure halite) and marlstone interbedded formations. In their study, diffusion through pure salt rock was found to be negligible, while it was slightly greater for the samples bearing carbonate and clay impurities. In addition, intergranular fractures acted as helium flow-leakage pathways.

Since hydrogen diffusion studies in rocks are rather restricted, we provide a few examples from molecular simulations applied on metals and alloys below; relevant results [72] show that hydrogen induces embrittlement and subsequently impacts structural properties. For example, in the nickel superalloy (IN-718) the diffusion coefficient varies between 10⁻¹² and 10⁻¹⁵ m²/s, depending on temperature (500–1400 K). These findings underscore the temperature dependence, particularly when diffusion occurs at very high temperatures (500 K equals 226.85 °C). However, this is not the case for most underground hydrogen storage conditions, whereas the previous data are primarily useful for designing hydrogen combustion systems. Additional research in fcc metals (Pd, Ni, Al, and Ag) [73] has shown the mechanism by which hydrogen (in its atomic form) diffuses into metals occupying space in the metal lattice. As such, hydrogen acts as an impurity in the metal lattice, thus enhancing embrittlement of metals, which could impact the well integrity, but as before, this phenomenon is not directly relevant to UHS.

Despite previous results regarding the diffusion coefficient, some studies suggest that although hydrogen is by its nature highly diffusive, this property does not threaten storage safety, as other properties of the caprock, such as low porosity, high tortuosity, and low permeability, can effectively limit its impact [37]. In this context, molecular dynamics simulations performed by [74], which tested hydrogen diffusivity in shales, reported a diffusion coefficient of 10⁻⁷ m²/s. According to the authors, this value seems to be slightly higher than the diffusion coefficients of CH₄ and CO₂ through the same medium. Their study, as well as that of [75], revealed the positive role of cushion gases in limiting hydrogen’s diffusivity through the pores. Additionally, ref. [74] used the previous diffusivity coefficient to calculate the distance that hydrogen would travel within a year, and demonstrated that hydrogen would penetrate the caprock by approximately 2.5 m. Based on these results, the authors brought the caprock thickness of a potential UHS site to the spotlight, since they predict a 25 m diffusion front after a century.

Table 18. Compilation of available data regarding hydrogen’s diffusion coefficient: (⊥) perpendicular to plane and (‖) parallel to plane.

Rock/Mineral	Temperature (°C)	Pressure (MPa)	Diffusion Coefficient (m2/s)	Reference
Werra rock salt	25	1	1.40 × 10−9	[76]
Opalinus clay			1.20 10−9
Shale (TOC 3.91%)	30 to 60	4	1.30 × 10−8 to 2.40 × 10−8	[16]
	30	0 to 20	1.80 × 10−8 to 8.00 × 10−9
Boom clay	21	1	(⊥) 2.64 × 10−10	[77]
			(‖) 7.25 × 10−10
			(‖) 5.51× 10−10
Shale (TOC 13%)	30	1.5 to 4.5	7.37 × 10−10 to 3.49 × 10−10	[15]
	60	1.5 to 4.5	1.04 × 10−10 to 1.27 × 10−11
Shale (TOC 18%)	30	1.5 to 4.5	2.45 × 10−9 to 7.49 × 10−9
	60	1.5 to 4.5	4.49 × 10−11 to 9.61 × 10−8
Mudstone	45	1	1.00 × 10−10	[78]
Australian anthracite coal	20 to 60	1.3	0.99 × 10−9 to 6.77 × 10−9	[79]
Albite	25	4.4	1.5 × 10−11	[70]
Quartz			2.5 × 10−11
Illite			10−11
Calcite			2.5 × 10−11
Anhydrite			1.5 × 10−11
Glauberite			1.6 × 10−11
Polyhalite			1.8 × 10−11
Halite			3 × 10−11
Mudstone			2.1 × 10−11
Glauberite-bearing salt			2.5 × 10−11

Table 18. Compilation of available data regarding hydrogen’s diffusion coefficient: (⊥) perpendicular to plane and (‖) parallel to plane.

Rock/Mineral	Temperature (°C)	Pressure (MPa)	Diffusion Coefficient (m2/s)	Reference
Werra rock salt	25	1	1.40 × 10−9	[76]
Opalinus clay			1.20 10−9
Shale (TOC 3.91%)	30 to 60	4	1.30 × 10−8 to 2.40 × 10−8	[16]
	30	0 to 20	1.80 × 10−8 to 8.00 × 10−9
Boom clay	21	1	(⊥) 2.64 × 10−10	[77]
			(‖) 7.25 × 10−10
			(‖) 5.51× 10−10
Shale (TOC 13%)	30	1.5 to 4.5	7.37 × 10−10 to 3.49 × 10−10	[15]
	60	1.5 to 4.5	1.04 × 10−10 to 1.27 × 10−11
Shale (TOC 18%)	30	1.5 to 4.5	2.45 × 10−9 to 7.49 × 10−9
	60	1.5 to 4.5	4.49 × 10−11 to 9.61 × 10−8
Mudstone	45	1	1.00 × 10−10	[78]
Australian anthracite coal	20 to 60	1.3	0.99 × 10−9 to 6.77 × 10−9	[79]
Albite	25	4.4	1.5 × 10−11	[70]
Quartz			2.5 × 10−11
Illite			10−11
Calcite			2.5 × 10−11
Anhydrite			1.5 × 10−11
Glauberite			1.6 × 10−11
Polyhalite			1.8 × 10−11
Halite			3 × 10−11
Mudstone			2.1 × 10−11
Glauberite-bearing salt			2.5 × 10−11

6. Discussion and Conclusions

Subsurface, underground, or geological hydrogen storage has long been considered important for supporting the energy sector in the future [4]. Field project examples, as summarized by [80], are widely distributed and include depleted hydrocarbon fields (2), saline aquifers (3), and salt caverns (5) located in Europe (6), Argentina (1), and the USA (4). As underground hydrogen storage (UHS) is important for a large-scale economy of hydrogen and, subsequently, for the deployment of the energy transition, ensuring the storage safety of a caprock formation capable of preventing leakages is of utmost significance. Caprocks lack cracks and open fractures (i.e., they are not geologically tectonized), which could act as preferential flow paths for hydrogen leakage into the cover rocks of the reservoir and, ultimately, to the surface; in the present work, key physical parameters regarding fluid flow and the caprock were examined. By using examples from the literature and our own results from Greece, parameters, such as the rock properties of wettability, interfacial tension (IFT), and diffusion, were reviewed in addition to their influence on diverse geological formations and rock minerals that, in various proportions, constitute the caprock.

Porosity and permeability are the first properties to consider in evaluating a formation’s ability to effectively seal a reservoir [19]. This enhances the safety of hydrogen storage when compared to other gases. Experiments have been conducted on various rocks/minerals and systems, as well as on a variety of total organic carbon concentrations (such as basalts, anhydrites, evaporites, granite, and shales, including their various clay minerals, calcite, quartz, and gypsum) [11,17,18,39,40,47]. A general remark for all rocks is that, under increasing pressure, wettability reduces as the contact angle increases (Table 3). In contrast, the effects of temperature changes were not straightforward (Table 4); a number of samples became more hydrophilic and wettability increased [11,39], while the rest became more hydrophobic and their wettability dropped [17]. Concerning the organic acid concentration (Table 5), it was revealed that, as concentration rises, the hydrophobicity of rocks increases, turning most of them to intermediate- or hydrogen-wet [18,39,48,49]. Apart from the rising concentration values, longer alkyl chains affected wettability in the same manner [48]. Among the other parameters reviewed, salinity and gas-mixture composition proved to be either insignificant or of no effect on wettability (Table 7 and Table 8).

Along with wettability, the interfacial tension (IFT) of the rock–liquid–gas system provides an additional sealing mechanism, as it determines the capillary entry pressure that hydrogen must surpass to intrude the caprock and escape. Among the parameters that significantly influence [81] interfacial tension, the pressure and temperature conditions play a key role [82]. The desirable liquid–gas IFT values should be high enough to ensure safe hydrogen containment. The literature review (Table 9 and Table 10) revealed that rock–liquid IFT has no dependence on pressure, while it is slightly affected by temperature changes [44,48,56]. Contrary to that, rock–gas and liquid–gas interfacial tensions are highly dependent on pressure changes [56,59], while, temperature changes can reduce IFT by 2 to 40% (Table 12). More specifically, as pressure increases (from 5 to 20 MPa), the rock–gas IFT decreases almost linearly for most of the samples. Similarly, as temperature rises (Table 11), the rock–gas IFT values decrease ([60,82] is a compilation of the data review in [82]). Even for the same mineral (i.e., quartz), the decreasing values of IFT at constant pressure show a direct dependence on temperature [63]. The variation in the liquid–gas IFT for the same pressure range does not present similar linearity, as previously mentioned [82]. Decreasing IFT values with increasing pressure were documented by [45] for a hydrogen–brine (liquid–gas) system. Moreover, studies on the influence of organic content on IFT have shown that increasing its concentration raises the rock–liquid IFT (Table 13), resulting in higher hydrophobicity [57,63]. The same stands with increasing alkyl chain length. The effect of salinity was more pronounced at higher temperatures, significantly raising the rock–liquid IFT values, as shown in Table 15 [17,59].

As hydrogen is a highly diffusive substance, researchers [15,64] have attempted to measure its diffusion coefficient in various rocks/minerals (Table 18). By using this coefficient along with porosity and tortuosity values, fluxes through a caprock can be calculated and potential losses can be quantified. Research on hydrogen diffusivity has mainly focused on metallurgy and materials science, evaluating its effects in metals and alloys. Molecular simulations applied in metals and alloys highlight the intrusion of hydrogen in the metal lattice, causing embrittlement under high temperatures [72,73]. Nevertheless, the limited literature on hydrogen diffusion in underground storage has revealed [37,74,75] that hydrogen fluxes can be minimized by a caprock’s low porosity and high tortuosity, as well as by the existence of a cushion gas. The latter gases (CH₄, CO₂) have higher affinities for rock surfaces, thereby expelling hydrogen to the pore centers [74,75].

Considering the relations of CCS and UHS, a direct comparison between CCS, carbon capture, and storage (or CGS, carbon geo-storage), and underground hydrogen storage (UHS) is rather uncertain [81]. CGS and UHS might not be correlated, since significant deviations occur. An important difference between underground gas/hydrogen storage and CO₂ subsurface storage is the consideration of the cyclic nature of operations (during injection/withdrawal) in hydrogen storage, which leads to wettability hysteresis. This mostly affects the reservoir rather than the caprock, by changing relative permeability, which, in turn, results in reduced withdrawal efficiency (e.g., in [81]). Further, we should not forget that CO₂ must be sequestered for long periods, if not forever, whereas stored gas and hydrogen are meant for recovery and energy production. When converting underground gas storage sites (typically depleted gas reservoirs)to hydrogen storage, recent findings have revealed that a specific storage site can contain approximately 25% less volume of hydrogen than the original natural gas volume [83].

Accordingly, this review concludes that for safety and leakage prevention of subsurface/geological hydrogen storage, a potential caprock should meet the following criteria:

• Poses low porosity, preferably unconnected or with a highly complicated pore system, and low permeability.
• The sealing geological formation should be strongly water-wet, regardless of pressure and temperature conditions.
• In order for the previous criterion to be met, the caprock should present low rock IFT, while exhibiting high rock–gas and liquid–gas IFT.
• Its effective diffusion coefficient should be the smallest possible, which is highly dependent on caprock properties such as porosity and tortuosity.
• The organic content of the caprock formation should be minimal, as the literature revealed that the greater the TOC values, the less the wettability (less water-wet).
• Concerning microbial action, it may have a double role in wettability, as it could reduce the formation’s wettability or increase it by developing a biofilm.

With respect to these criteria, and among the reviewed geological formations and minerals, evaporites, low-organic-content shales, mudstones, muscovite, and anhydrite have been identified as highly effective sealants, offering excellent sealing capabilities. Special caution should be given to anhydrite, which is better avoided due to its reactivity with hydrogen.

Further research is needed regarding underground hydrogen storage (UHS) and, more specifically, on the hydrogen interaction with the caprock formation. This relationship is of critical importance, since the interaction between hydrogen and caprock during subsurface storage poses significant risks to the viability of a project, including potential failure, primarily through geochemical and physical processes that compromise caprock integrity. Additionally, there are gaps in current research that are relevant to the present rock types, where there is a lack of experimental tests for specific temperature and pressure ranges based on representative, realistic conditions. Contradictions include the following: (a) literature reporting contrasting results for systems of rock–hydrogen–brine on a positive or no-correlation between contact angle and pressure [82]; and (b) those mixed results relevant to salinity produced by molecular simulations, whether NaCl or other salts have a more intensive effect on the IFT [65,79].

Among future perspectives, the following items are believed to be highly influential additions to safe hydrogen subsurface storage:

• The wettability of other potential caprock formations under various pressures, temperatures, microbial conditions, and gas mixture compositions should be investigated.
• More experimental data should be generated on the wettability behavior of different rock types under long-term injection/production cycles of large- and medium-scale projects.
• Further TOC and mineralogy, as well as comparative experiments on various rock types, should be performed in order to assess the influence of organic and mineral content on wettability.
• Since, IFT reduction varies in rate and extent with increasing pressure across various rocks and temperatures, caution should be given, and more targeted studies should be performed.
• The effective diffusion coefficient should be further examined for a wider variety of geological formations under various pressure and temperature conditions.
• Finally, the use of surfactants, which enhance rock hydrophilicity, should be examined in depth along with their economic and other possible side effects (e.g., environmental impact).