Study on Experiment and Molecular Dynamics Simulation of Variation Laws of Crude Oil Distribution States in Nanopores

Abstract

This study is based on an experiment and a molecular dynamics simulation to investigate the distribution states and property variation laws of crude oil in nanopores, aiming to provide theoretical support for efficient unconventional oil and gas development. Focus is placed on the distribution mechanisms of multicomponent crude oil in oil-wet siltstone (SiO₂) and dolomitic rock (dolomite, CaMg₃(CO₃)₄) nanopores, with comprehensive consideration of key factors including pore size, rock type, and CO₂ flooding on crude oil distribution at 353 K and 40 MPa. It is revealed that aromatic hydrocarbons (toluene) in multicomponent crude oil are preferentially adsorbed on pore walls due to π-π interactions, while n-hexane diffuses toward the pore center driven by hydrophobic effects. Pore size significantly affects the distribution states of crude oil: ordered adsorption structures form for n-hexane in 2 nm pores, whereas distributions become dispersed in 9 nm pores, with adsorption energy changing as pore size increases. Dolomite exhibits a significantly higher adsorption energy than SiO₂ due to surface roughness and calcium–magnesium ion crystal fields. CO₂ weakens the interaction between crude oil and pore walls through competitive adsorption and reduces viscosity via dissolution, promoting crude oil mobility. Nuclear magnetic resonance (NMR) experiments further verified the effect of CO₂ on crude oil stripping in pores. This study not only clarifies the collaborative adsorption mechanisms and displacement regulation laws of multi-component crude oil in nanopores but also provides a solid theoretical basis for CO₂ injection strategies in unconventional reservoir development.

1. Introduction

The efficient development of unconventional oil and gas resources is a core challenge in ensuring global energy security and energy structure transformation. According to statistics from the International Energy Agency (IEA, Paris, France), global recoverable shale oil resources exceed 450 billion barrels, accounting for over 30% of undeveloped hydrocarbon resources, yet their recovery efficiency generally remains below 10% [1]. This inefficiency stems from the unique nanoporous structure of unconventional reservoirs—approximately 70% of pore diameters in shale reservoirs are less than 10 nm [2]. The strong confinement effect of nanoscale pores and complex solid–liquid interface interactions invalidate traditional seepage theories, making the distribution states and mobility of crude oil molecules in pores critical scientific issues restricting recovery efficiency [3,4,5]. At the microscale, systematic analysis of adsorption behaviors of crude oil components and the intervening mechanisms of geometric constraints from pore sizes is urgently required to guide innovations in efficient development technologies.

Through experimental approaches, multifactors such as pore size, rock type, and fluid composition interact and couple, making it difficult to independently investigate the influence of a single variable. Real-time observation of the distribution states of crude oil components in pores is not feasible. Molecular simulation, however, enables tracking of the movement trajectories of each molecule and reveals instantaneous changes in interfacial interactions [6]. Current research primarily focuses on using molecular simulation to study the distribution states and movement laws of heavy components. For example, Zhao et al. simulated the gelation process of waxy crude oil by resins and found that the presence of resin components significantly reduces system energy and the interaction ability of paraffin molecules, thereby inhibiting paraffin aggregation [7]. Zhu et al. conducted molecular dynamics research on the influence of water content on asphaltene aggregation behavior and crude oil rheology, concluding that continuous water introduction transforms asphaltene aggregation from bulk-phase to interfacial adsorption, ultimately forming water emulsions in crude oil [8]. However, these studies often overlook the distribution states of light components in nanopores. Additionally, multicomponent collaborative mechanisms are lacking, and competitive adsorption laws of light components remain unclear. Formation pores typically exhibit heterogeneous and irregular structures, necessitating investigation into the influence of components across different pore geometries.

Considerable influence is exerted by different pore walls on crude oil components. In nanopores, smaller pore sizes lead to stronger confinement effects, restricting free molecular diffusion within limited spaces and inducing more ordered molecular arrangements [9,10,11,12,13]. Han et al. investigated the impact of pore size and wettability on CO₂ hydrate stability through molecular simulation, demonstrating that pore size significantly affects hydrate stability while wettability variations alter SiO₂-hydrate interactions and CO₂ nanobubble distributions [14]. Recent studies report that fluid molecules with distinct sizes, shapes, and polarities exhibit differing behaviors in nanopores [15,16,17,18,19,20]. For instance, significant differences exist in the adsorption and diffusion behaviors of polar and non-polar molecules within oil-wet pores. Consequently, systematic comparisons of multicomponent distribution states across varying pore sizes and rock wall surfaces remain lacking. Additionally, the microscale pathways by which CO₂ regulates crude oil distribution through competitive adsorption and dissolution mechanisms are not yet clearly understood.

Shale reservoirs typically contain numerous nanoscale pores [21], with natural fractures often insufficient for effective recovery. Thus, hydraulic fracturing is usually applied early to create complex fracture networks for production [22]. Initial extraction relies on fracturing fluids to establish flow paths and release shale oil, while CO₂ flooding or refracturing dominates mid-to-late stages for energy replenishment. During fracturing, shale oil contacts fracturing fluids, altering its in situ distribution. CO₂ dissolution swells oil volume, releasing elastic energy to compensate for reservoir depletion from fracturing [23], while gaseous CO₂ displaces pore water to restore seepage pathways [24]. As shale oil primarily interacts with water and CO₂ in situ, understanding its post-contact state with these components is critical for characterizing post-displacement reservoir oil distribution.

To address the aforementioned issues, this study reveals molecular mechanisms of multicomponent collaborative adsorption and confinement effects in dolomite and siltstone (SiO₂) reservoirs (defined rock types). Crude oil distribution states of simulated components—n-hexane (alkane) and toluene (aromatic hydrocarbon)—are characterized across nanopore sizes of 2–9 nm (critical range in shale and tight carbonate formations). The synergistic effects of CO₂ flooding are analyzed, with a fixed fluid composition of CO₂ and water (1:1 ratio) to simulate in situ displacement conditions. Specifically, competitive adsorption between CO₂ and crude oil, as well as viscosity reduction via CO₂ dissolution, are uncovered. NMR analysis demonstrates that CO₂ effectively displaces crude oil from pores, particularly in small nanopores (2–5 nm). Through integrated analysis of rock type (dolomite/SiO₂), pore size (2–9 nm), and fluid components (n-hexane, toluene, CO₂, water), this study systematically clarifies crude oil distribution states under varying reservoir conditions, providing robust theoretical support for optimizing CO₂-assisted development strategies in unconventional formations.

2. Molecular Dynamics (MD) Simulation Details

In this work, the Materials Studio molecular simulation software (2019) was utilized to compare the adsorption properties of different components on siltstone and dolomitic rock at a temperature of 353 K, so as to determine the distribution state and production law of crude oil in the original formation.

The specific modeling process is as follows: Silica is used to replace shale, with α-silica crystals imported from the database (Figure 1a). Its lattice parameters are a = 4.909996 Å, b = 4.909996 Å, c = 5.402 Å, α = β = 90.000°, γ = 120.000°. The quartz (1,0,0) plane is selected as the cleavage surface (Figure 1b), with the unit cell expanded and the surface methylated (Figure 1c) to construct a lipophilic (methyl) slit-pore model. A dolomite unit cell (CaMg₃(CO₃)₄) is imported (Figure 1d), with lattice parameters a = 4.806398 Å, b = 4.909996 Å, c = 16.006 Å, α = β = 90.000°, γ = 120.000°, to establish a supercell surface.

The use of these two molecules as rock surfaces has been widely validated [25,26,27,28]. N-hexane (C₆) and toluene represent straight-chain alkanes and aromatic hydrocarbons, respectively. With 12 conformational degrees of freedom and no side-chain steric effects, C6 clearly demonstrates van der Waals interactions [29,30,31]. Toluene’s small molecular weight and well-established force field enable accurate reproduction of pore-scale alkane-aromatic distributions [32,33,34].

Densities at 353 K and 40 MPa were obtained from NIST databases [35]. Multicomponent models used an alkane: aromatic ratio of 8:13, gas: liquid and water: oil ratios of 1:1, with molecular counts and densities listed in

Table 1. Molecular weights and density of different models. (A check mark √ indicates that this lithology is used, while a cross mark × indicates that this lithology is not used).

NO.	N-hexane	Toluene	H2O	CO2	Siltstone	Dolomite	353 K and 40 MPa-Density (g/cm3)	Explain
1	72	117	0	0	√	×	0.90658	Competitive adsorption of multicomponents in 5 nm pores
2	197	0	0	0	√	×	0.65231	Distribution state of n-hexane in 5 nm pores
3	0	189	0	0	√	×	0.84646	Distribution state of toluene in 5 nm pores
4	80	0	0	0	√	×	0.65231	Distribution state of n-hexane in 2 nm pores
5	355	0	0	0	√	×	0.65231	Distribution state of n-hexane in 9 nm pores
6	0	189	0	0	×	√	0.84646	Comparison of distribution states between toluene and n-hexane in 5 nm pores
7	197	0	197	0	√	×	0.826155	Distribution state of n-hexane in water-containing pores
8	197	0	0	197	√	×	0.65231	Distribution state of n-hexane under CO2 action
9	0	189	0	189	√	×	0.78323	Distribution state of toluene under CO2 action

. CO₂ was modeled as a supercritical fluid under these conditions. Molecular colors in results were consistent with Figure 2.

Models of oil components, CO₂, and H₂O were constructed in the Amorphous Cell module at 353 K using Table 1 densities. The Build Layers module combined these with rock surfaces to form complete models.

The COMPASS force field, validated for petroleum systems [36,37,38], described van der Waals and electrostatic interactions. Ewald summation and 9–6 Lennard-Jones potentials were used for electrostatics and dispersion [39,40,41]. An NVT ensemble maintained constant volume; initial density parameters corresponded to 40 MPa. A 12.5 Å cutoff radius and Nose thermostat [42] were applied. The entire simulation was conducted under the canonical (NVT) ensemble with 500 ps annealing, 500 ps dynamics, and 1 ns data collection.

3. Nuclear Magnetic Displacement Experiment

The NMR system employs low-field NMR technology. The low-field NMR analyzer and imaging system (MacroMR12-150I, Niumag Analytical Instrument Co., Ltd., Beijing, China) comprises a permanent NdFeB magnet and NMdigI spectrometer, measuring ¹H signals in pore fluids at a Larmor frequency of 12.66 MHz, 0.3 Tesla magnetic field strength, 20 ppm homogeneity over 60 mm, a 70 mm probe coil diameter, and a 40 μs dead time, Figure 3.

Natural cores were obtained from the Jimusaer shale oil block (3498.9 m depth), Xinjiang, China, with a reported pore volume of 12% and permeability of 90.43 mD. Oil-saturated natural cores were cut and polished into cylindrical samples (2.5 cm diameter, 30 cm length).

(1)

Polished cores were immersed in a mixed solution (deionized water: D₂O, 1:1 v/v) for one week to saturate pore spaces. Immersion rather than vacuum saturation was used to preserve core oil and structure.

(2)

Surface moisture was wiped off, and cores were scanned by NMR to map initial mixed solution distribution; signal-free regions were identified as oil components.

(3)

Scanned cores were placed in a core holder (confining pressure 2 MPa, backpressure 1 MPa) and waterflooded at 82 °C with the mixed solution at 1 mL/min for 2 h (≈1.7 PV).

(4)

After flooding, cores were wiped and re-scanned by NMR to visualize post-waterflood oil distribution.

(5)

Post-scan cores were reinserted for CO₂ (purity > 99.9%) flooding at 1 mL/min for 2 h (immiscible displacement). Post-flood cores were immersed in the mixed solution for one week, wiped, and scanned by NMR to observe post-CO₂ oil distribution in pores.

4. Results and Discussion

4.1. Effects of Multicomponent Crude Oil Composition on Distribution States

Figure 4 depicts the initial model (0 ps) and equilibrium configuration (2000 ps) of multicomponents on the SiO₂ surface. In this model, the slit distance of SiO₂ is set to 5 nm.

Figure 4 illustrates the dynamic evolution and component distribution characteristics of multicomponent crude oil within SiO₂ slit pores. Homogeneous mixing of multicomponent crude oil in the pores is shown by the initial model (a), whereas system equilibrium is indicated by the final configuration (b) after 2000 ps simulation. In Figure 4c, aromatic hydrocarbons exhibit localized aggregation tendencies in the pores but are mostly adsorbed onto the siltstone surface. Conversely, Figure 4d demonstrates that n-hexane distributes relatively dispersedly, with higher concentration in the central pore region. Thus, under multicomponent conditions, stronger adsorption capacity for light hydrocarbon components is exhibited by siltstone, while heavy components form aggregates within the pores.

Based on the above simulation snapshots, the density distribution of each component in the pores was calculated to further validate the distribution characteristics of each component (Figure 5). Aromatic hydrocarbons (toluene) exhibit two symmetric adsorption layers closest to the rock surface, with adsorption peaks located at ±2.61 nm and ±2.21 nm. Saturated hydrocarbons (n-hexane) show an adsorption peak at ±1.89 nm, forming a stable adsorption layer near aromatic hydrocarbons (toluene), which is consistent with the observations from the simulation snapshots.

This phenomenon arises from the synergistic regulation of the physicochemical properties of multicomponent crude oil components, intermolecular interactions, and solid–liquid interfacial interactions [43]. Aromatic hydrocarbons, due to their conjugated π-bond structure, undergo specific adsorption with polar SiO₂ pore walls through strong π-π interactions, leading to their accumulation near the pore walls. In contrast, n-hexane, with weak polarity and low intermolecular van der Waals forces, tends to disperse and occur in the energetically stable central region of the pores [44]. This distribution mechanism is fundamentally driven by the combined effects of component polarity, intermolecular forces (van der Waals forces, hydrophobic interactions), and interfacial interactions.

4.2. Effects of Single Crude Oil Components on Distribution States

Figure 6 shows the simulation snapshots of the n-hexane/toluene-oil-wet SiO₂ 5 nm slit pore. As seen in Figure 6c, n-hexane maintains a relatively diffuse distribution state in the equilibrium model of oil-wet SiO₂ 5 nm pores. In contrast, Figure 6d shows significant structural rearrangement in the toluene equilibrium model, where molecules aggregated toward the pore walls. The density distribution in Figure 7 further corroborates this characteristic. Figure 7a indicates that although n-hexane exhibits density peaks near the pore walls (±2.61 nm), the curve fluctuates frequently and the density in the central region decays significantly. For toluene Figure 7b, higher-intensity density peaks form at the pore walls, and a “high-low-high” distribution trend is observed from the pore walls to the central pore region, indicating a higher enrichment degree near the pore walls and a more regular distribution pattern. These findings demonstrate distinct differences in the distribution of different crude oil components within oil-wet pores.

The adsorption energy between a liquid and a solid surface can effectively quantify the strength of interactions between them. The greater the value, the stronger the interactions and the higher the adsorption capacity. The definition of adsorption energy is as follows:

$${\mathrm{E}}_{\mathrm{adsorption}}={\mathrm{E}}_{\mathrm{total}}-({\mathrm{E}}_{\mathrm{surface}}+{\mathrm{E}}_{\mathrm{fluid}})$$

(1)

where ${\mathrm{E}}_{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}}$ is the total energy of the SiO₂-crude oil system; ${\mathrm{E}}_{\mathrm{surface}}$ is the energy of a single crude oil; and ${\mathrm{E}}_{\mathrm{fluid}}$ is the energy of a single SiO₂ surface.

Table 2. Adsorption energy of n-hexane/toluene-oil-wet SiO₂.

Components	${\mathbf{E}}_{\mathbf{ele}}$ (kcal/mol)	${\mathbf{E}}_{\mathbf{vdw}}$ (kcal/mol)	${\mathbf{E}}_{\mathbf{adsorption}}$(kcal/mol)
N-hexane-SiO2	1.016	308.65	348.77
Toluene-SiO2	743.679	1535.619	741.607452

presents the adsorption energy of n-hexane and toluene with oil-wet SiO₂. Among them, toluene exhibits a negative ${\mathrm{E}}_{\mathrm{ele}}$ with a large absolute value, indicating stronger electrostatic interactions with SiO₂; both ${\mathrm{E}}_{\mathrm{v}\mathrm{d}\mathrm{w}}$ and ${\mathrm{E}}_{\mathrm{adsorption}}$ are also significantly higher than those of n-hexane.

From the perspective of molecular interaction nature, the benzene ring structure of toluene has delocalized π electron cloud [45,46]. This feature can enhance the van der Waals force via π-interaction and induced dipole interaction with the lipophilic SiO₂ surface, resulting in a high ${\mathrm{E}}_{\mathrm{vdw}}$ (van der Waals energy) value of 1535.619 kcal/mol. Concurrently, the electron cloud distribution of the benzene ring causes ${\mathrm{E}}_{\mathrm{e}\mathrm{l}\mathrm{e}}$ (electrostatic energy) to show a significant negative value, indicating strong electrostatic attraction. Collectively, these interactions cause the ${\mathrm{E}}_{\mathrm{adsorption}}$ energy to rise to 741.607 kcal/mol. In contrast, n-hexane, a non-polar straight-chain alkane, interacts with SiO₂ solely through weak dispersion forces. This results in a much lower ${\mathrm{E}}_{\mathrm{v}\mathrm{d}\mathrm{w}}$ of 308.65 kcal/mol, an ${\mathrm{E}}_{\mathrm{ele}}$ near zero, and, consequently, a lower overall ${\mathrm{E}}_{\mathrm{adsorption}}$. From these results, it is evident that not only heavy components but also light components with strong polarity can affect the pore seepage effect, leading to lower recovery efficiency.

4.3. Effects of Pore Size on Distribution States

Within different-sized pores, molecular translational and rotational degrees of freedom are restricted by spatial confinement. In nanoscale slit pores, crude oil component molecules cannot diffuse freely and are forced to adapt to the pore space morphology, leading to enhanced molecular alignment order and the distribution of layered distribution or oriented adsorption phenomena [47,48,49,50,51]. Taking n-hexane with moderate adsorption properties as an example, the distribution state of crude oil components under confinement effects is investigated.

Figure 8 shows the equilibrium model of n-hexane in oil-wet SiO₂ slit pores of different sizes. Calculating the density distribution of the model (Figure 9) reveals that in 2 nm pores, the density curve fluctuates, showing multiple high-intensity peaks. Five adsorption layers are formed by n-hexane, all in an adsorbed state with high density. As the pore width increases, three stable symmetric adsorption layers appear in the oil-wet SiO₂ pores. The density of each adsorption layer gradually decreases, and the peak density of the first adsorption layer reduces with increasing pore width. The adsorption layer thickness is approximately 2.645 nm, with adsorbed state accounting for 52.3%. Each adsorption layer has a thickness of approximately 0.44 nm, a value consistent with the width of normal alkane molecules. This indicates that n-hexane in the adsorption layer is adsorbed parallel to the shale surface, as shown in Figure 10.

The two main components of adsorption energy are van der Waals energy (${\mathrm{E}}_{\mathrm{vdw}}$) and electrostatic potential energy (${\mathrm{E}}_{\mathrm{ele}}$). The results of these interaction energies are presented in

Table 3. Adsorption energy between crude oil and rock surfaces in different pore sizes.

Pore Scale	${\mathit{E}}_{\mathit{e}\mathit{l}\mathit{e}}$ (kcal/mol)	${\mathit{E}}_{\mathit{v}\mathit{d}\mathit{w}}$ (kcal/mol)	${\mathit{E}}_{\mathit{a}\mathit{d}\mathit{s}\mathit{o}\mathit{r}\mathit{p}\mathit{t}\mathit{i}\mathit{o}\mathit{n}}$ (kcal/mol)
2 nm	0.56	295.87	317.86
5 nm	1.016	308.65	348.77
9 nm	1.395	303.577	359.16

. As the pore size increases from 2 nm to 9 nm, electrostatic energy (${\mathrm{E}}_{\mathrm{ele}}$) gradually increases from 0.56 kcal/mol to 1.395 kcal/mol. Van der Waals energy (${\mathrm{E}}_{\mathrm{v}\mathrm{d}\mathrm{w}}$) exhibits a trend of first increasing and then decreasing, reaching a peak of 308.65 kcal/mol at 5 nm. Adsorption energy (${\mathrm{E}}_{\mathrm{adsorption}}$) continuously increases from 317.86 kcal/mol to 359.16 kcal/mol. Therefore, pore size exerts a significant regulatory effect on interfacial energy distribution.

The observed differences essentially arise from changes in the intensity of confinement effects induced by pore size. Small 2 nm pores exert strong confinement, restricting the movement freedom of n-hexane molecules and forcing them to form an ordered layered structure near the pore walls. This leads to prominent density peaks and fluctuations. In 5 nm pores, the confinement effect weakens, reducing molecular alignment order and consequently decreasing the amplitude of density fluctuations. For 9 nm pores, the confinement effect is significantly attenuated, reducing the constraints of pore walls on n-hexane molecules. This allows the molecules to disperse in a freer space, leading to a smoother density curve. Furthermore, interactions between oil-wet pore walls and n-hexane are amplified in small pores, enhancing molecular aggregation and ordered arrangement near the pore walls. This further highlights the mechanism by which pore size regulates the microscale distribution state of n-hexane through confinement effects.

In small pores (e.g., 2 nm), strong confinement effects bring molecules extremely close to the rock surface, where van der Waals interactions dominate and concentrate, leading to prominent ${\mathrm{E}}_{\mathrm{vdw}}$ values. As the pore size increases, electrostatic interactions between molecules and pore walls gradually strengthen due to adjusted interaction distances and increased contact opportunities, causing ${\mathrm{E}}_{\mathrm{ele}}$ to rise continuously. Although ${\mathrm{E}}_{\mathrm{binding}}$ increases with pore enlargement, reflecting cumulative effects of overall interaction intensity, the decreasing average adsorption energy indicates reduced uniformity of molecule-wall interactions and fewer localized strong interaction sites as pores expand. Thus, pore size differentially regulates the interfacial energy distribution between crude oil and rock by altering intermolecular interaction distances, contact areas, and confinement effect intensity. Van der Waals forces remain the primary intermolecular interactions between crude oil and rock surfaces.

4.4. Effects of Rock Composition on Distribution States

Figure 11 presents the simulation snapshots of n-hexane on dolomite and oil-wet SiO₂ surfaces. No significant changes were observed after structural equilibration. However, upon magnifying the contact between n-hexane and the wall surfaces, it was found that the dolomite surface is rougher than SiO₂ and exhibits a wavy morphology (Figure 12). This allows the inference that n-hexane has stronger adsorption capacity on the dolomite surface.

Figure 13 shows the density distribution of n-hexane along the Z-axis within dolomite and SiO₂ pores. In both dolomite and SiO₂ pores, n-hexane exhibits high-density peaks near the pore walls (z ≈ ±2.5 nm), with density fluctuations in the central region tending to stabilize. However, it can be clearly observed that the density of n-hexane near the dolomite surface is 1.12 g/cm³, whereas it is 1.31 g/cm³ near the SiO₂ surface. Additionally, the density trough of n-hexane at z ≈ ±2.3 nm on the dolomite surface is significantly lower than that on the SiO₂ surface.

These findings indicate distinct adsorption patterns of n-hexane on dolomite and SiO₂ surfaces. On the dolomite surface, n-hexane forms a multi-layer adsorption state with ordered double adsorption layers. Therefore, the adsorption capacity of dolomite for n-hexane is stronger than that of SiO₂.

Table 4. Adsorption energy of n-hexane with different rock surfaces.

Components	${\mathit{E}}_{\mathit{e}\mathit{l}\mathit{e}}$ (kcal/mol)	${\mathit{E}}_{\mathit{v}\mathit{d}\mathit{w}}$ (kcal/mol)	${\mathit{E}}_{\mathit{a}\mathit{d}\mathit{s}\mathit{o}\mathit{r}\mathit{p}\mathit{t}\mathit{i}\mathit{o}\mathit{n}}$ (kcal/mol)
N-hexane-Dolomite (dolomitic rock)	0.95	450.36	482.51
N-hexane-SiO2 (siltstone)	1.016	308.65	348.77

presents the adsorption energy of n-hexane on different rock surfaces. It shows that the adsorption energy of n-hexane to dolomite is greater than that to SiO₂. This is attributed to the formation of a bimolecular adsorption layer of n-hexane on the dolomite surface, which leads to an increase in adsorption energy.

The calcium–magnesium ion crystal structure of the dolomite surface enables stronger van der Waals interactions with n-hexane, particularly significant contributions from dispersion forces, thereby elevating ${\mathrm{E}}_{\mathrm{vdw}}$ and ${\mathrm{E}}_{\mathrm{binding}}$. In contrast, the SiO₂ surface is dominated by silicon-oxygen bonds, with relatively uniform polarity. The weaker van der Waals interactions between SiO₂ and nonpolar n-hexane result in lower energy values. Fundamentally, the complex crystal field environment of the dolomite surface enhances dispersion interactions with n-hexane molecules, while the SiO₂ surface, with fewer interaction sites and limited interaction types, struggles to achieve comparable interfacial interaction intensity.

Therefore, during oilfield exploitation, for dolomite formations, strategies such as adding surfactants can be implemented to weaken the adsorption capacity between crude oil components and dolomite.

4.5. Oil-Water Distribution States in Reservoir Formations

In current petroleum development processes, water flooding is commonly used to displace crude oil in formations. Therefore, it is necessary to investigate the distribution states of crude oil and water in rock pores to understand the effectiveness of water flooding in oil-wet rocks.

Figure 14 shows the simulation snapshots of n-hexane and H₂O in a 5 nm oil-wet SiO₂ slit pore. From Figure 14, it can be observed that after the model reaches equilibrium, most of the H₂O originally adsorbed on the SiO₂ surface is replaced by n-hexane. Figure 15 further demonstrates that the density of n-hexane at the SiO₂ wall is higher than that of H₂O, suggesting that H₂O has little influence on crude oil distribution. The proportion of adsorbed n-hexane in the single-component system is 52.3%, while in the n-hexane + H₂O system, this proportion decreases to 42.69%. H₂O molecules can adsorb on the rock surface, thus reducing the proportion of adsorbed saturated hydrocarbons to a certain extent.

Table 5. Adsorption energy of n-hexane and H₂O with oil-wet SiO₂.

Components	${\mathit{E}}_{\mathit{e}\mathit{l}\mathit{e}}$ (kcal/mol)	${\mathit{E}}_{\mathit{v}\mathit{d}\mathit{w}}$ (kcal/mol)	${\mathit{E}}_{\mathit{a}\mathit{d}\mathit{s}\mathit{o}\mathit{r}\mathit{p}\mathit{t}\mathit{i}\mathit{o}\mathit{n}}$ (kcal/mol)
N-hexane-SiO2	1.02	308.65	348.77
N-hexane-H2O-SiO2	1.31	260.29	304.77

shows that upon the addition of H₂O, the adsorption energy of n-hexane to the rock slightly decreases, but the adsorption trend remains unchanged. Thus, the aqueous phase can reduce the adsorption capacity of n-hexane, yet it cannot effectively desorb a significant amount of n-hexane.

As a strong polar molecule, H₂O enhances electrostatic interactions (${\mathrm{E}}_{\mathrm{ele}}$ increases) in the system due to additional electrostatic interactions between its polar groups and the oil-wet SiO₂ surface. Simultaneously, H₂O preferentially occupies interaction sites on the SiO₂ surface, hindering direct contact between n-hexane and the rock surface. This weakens van der Waals interactions (${\mathrm{E}}_{\mathrm{vdw}}$ decreases) between n-hexane and SiO₂, ultimately leading to a decrease in total adsorption energy (${\mathrm{E}}_{\mathrm{adsorption}}$). Therefore, H₂O regulates the interactions between nonpolar crude oil components and the rock surface by altering interfacial interaction sites and types, yet the aqueous phase cannot effectively desorb adsorbed crude oil.

4.6. Effects of CO₂ on Crude Oil Distribution States

CO₂ flooding is a crucial enhanced oil recovery (EOR) technology. Upon injection into reservoirs, CO₂ dissolves in crude oil, reducing its viscosity, causing volume expansion, altering the distribution of crude oil in pores, and promoting release and migration. Therefore, it is essential to understand the migration of crude oil components under the influence of CO₂.

CO₂ exists as a supercritical fluid that fully dissolves in crude oil upon contact at 353 K and 40 MPa. This property was incorporated into modeling by forming miscible interfaces between CO₂ and oil molecules to simulate oil component dynamics during miscibility.

Figure 16 shows that after reaching equilibrium configuration, CO₂ is thoroughly dispersed within crude oil, with a certain amount of CO₂ molecules accumulating on the SiO₂ surface. Figure 17 presents the density distribution characteristics of multi components (n-hexane, toluene) and CO₂ in a 5 nm oil-wet SiO₂ pore. CO₂ forms distinct high-density peaks near the pore walls (on both sides of the z-axis) with a broad distribution range. The density fluctuations of n-hexane are attenuated, and the density in the central pore region is significantly lower than that near the pore walls, indicating that CO₂ has a certain stripping effect on n-hexane. The density peaks of toluene at the pore walls are significantly reduced, and the central pore region exhibits complex fluctuations, leading to decreased overall distribution uniformity. Thus, the effect of CO₂ on toluene is dissolution, expansion, and viscosity reduction.

Table 6. Adsorption energy of n-hexane and CO₂ with oil-wet SiO₂.

Components	${\mathit{E}}_{\mathit{e}\mathit{l}\mathit{e}}$ (kcal/mol)	${\mathit{E}}_{\mathit{v}\mathit{d}\mathit{w}}$ (kcal/mol)	${\mathit{E}}_{\mathit{a}\mathit{d}\mathit{s}\mathit{o}\mathit{r}\mathit{p}\mathit{t}\mathit{i}\mathit{o}\mathit{n}}$ (kcal/mol)
N-hexane-SiO2	0.894	234.646	264.869
CO2-SiO2	0.975	51.515	60.0911
N-hexane-CO2	2.757	515.128	521.808

and

Table 7. Adsorption energy of toluene and CO₂ with oil-wet SiO₂.

Components	${\mathit{E}}_{\mathit{e}\mathit{l}\mathit{e}}$ (kcal/mol)	${\mathit{E}}_{\mathit{v}\mathit{d}\mathit{w}}$ (kcal/mol)	${\mathit{E}}_{\mathit{a}\mathit{d}\mathit{s}\mathit{o}\mathit{r}\mathit{p}\mathit{t}\mathit{i}\mathit{o}\mathit{n}}$ (kcal/mol)
Toluene-SiO2	4.84	247.48	285.92
CO2-SiO2	1.19	59.59	69.43
Toluene-CO2	5.959	806.834	806.45

present the adsorption energy of n-hexane and toluene with CO₂ and oil-wet SiO₂, respectively. The adsorption energy of the n-hexane-CO₂ system is significantly higher than that of n-hexane-SiO₂, and the toluene-CO₂ adsorption energy is also far higher than that of toluene-SiO₂. Additionally, all interaction energies of CO₂-SiO₂ are lower than those of crude oil component-SiO₂ systems. This demonstrates distinct effects of CO₂ on different crude oil components. The strong van der Waals energy between CO₂ and n-hexane/toluene dominates the high adsorption energy, originating from dispersion forces between CO₂ molecules and hydrocarbon components, particularly the complex induced dipole effects formed between the benzene ring of toluene and CO₂.

With its small molecular size and strong diffusion capacity, CO₂ preferentially occupies oil-wet SiO₂ pore wall sites through competitive adsorption, forming an adsorption layer via van der Waals interactions. This weakens direct interactions between n-hexane/toluene and the pore walls. For n-hexane, CO₂ intervention dilutes its interaction intensity with the pore walls, forcing a more dispersed distribution. For toluene, CO₂ disrupts π-interactions between the benzene ring and pore walls, triggering redistributions within the pores.

CO₂ alters the distribution state of crude oil components in reservoir pores through competitive adsorption, reducing the adsorption strength between crude oil and rock surfaces and promoting crude oil desorption and flow. During development, adjusting CO₂ injection volume and timing can enhance its “displacement-desorption” effect on crude oil, effectively improving oil recovery.

4.7. Distribution State of Crude Oil Under Different Displacement Methods Distribution

Nuclear magnetic resonance (NMR) spectroscopy was used to analyze the distribution state of crude oil in oil-saturated core pores. Figure 18 shows T₁–T₂ NMR plots of natural cores at different stages. The horizontal axis (T₂) reflects pore size: longer axes indicate larger pores, shorter axes smaller pores. The vertical axis (T₁) represents fluid-rock surface interactions: longer axes signify weaker interactions, shorter axes stronger interactions. Color intensity denotes fluid signal strength, i.e., crude oil component content in pores. D₂O was added to deionized water to enhance displacing fluid signals, leveraging the ¹H nuclear properties to distinguish hydrogen in crude oil from that in displacing fluids via differential signal amplitudes. Figure 18 displays hydrogen signals from crude oil, enabling quantification of residual oil content across pore sizes.

In Figure 18a, initial crude oil content in pores is high. After water flooding, the red region weakens, indicating reduced oil content (Figure 18b). Post-CO₂ flooding, residual oil is largely displaced (Figure 18c). Minimal changes in the vertical axis suggest unchanged oil-rock interactions, demonstrating high efficiency of CO₂ in mobilizing residual oil.

In the original core, remaining oil exists in multiple states: free oil in large pores corresponds to strong signals in the high-T₂ interval of the T₂ spectrum; oil in small pores or throats is constrained by capillary forces, manifesting as low-T₂ interval signals. After water flooding, free oil in large pores is preferentially displaced. Remaining oil mostly remains as thin films or clusters in small pores or complex pore structures. The high-T₂ signal in the T₂ spectrum significantly attenuates, but residual oil signals persist in the low-T₂ interval, highlighting the limited mobilization of water flooding for oil trapped in small pores.

After compiling the T₂ signal intensities at different stages, Figure 19 is obtained. The curve exhibits the highest peak concentrated in the high-T₂ interval (10–100 ms), indicating that oil in the original core is primarily in a free state within large pores. The high hydrogen proton content and relaxation characteristics correspond to the strongest signal amplitude. Following water flooding, the peak amplitude decreases, and the high-T₂ interval signal attenuates significantly. As water displaces free oil in large pores, remaining oil mainly exists as capillary-bound oil in small pores or residual oil, leading to an overall reduction in signal intensity. This reflects the effective displacement of large-pore oil by water flooding but highlights its limited ability to mobilize remaining oil in small pores.

After CO₂ flooding, the peak further adjusts, with continued weakening of the high-T₂ interval signal and altered distribution of low-T₂ interval signals. Through mechanisms such as dissolution, expansion, and extraction, CO₂ disrupts the distribution state of remaining oil. It not only mobilizes residual oil in large pores but also influences capillary-bound oil in small pores.

Therefore, CO₂ exerts a certain stripping effect on crude oil, disrupting the adsorption equilibrium between oil and rock surfaces and prompting residual oil to detach from the surfaces of rock pores. The further adjustment of signal distribution in the NMR spectrum directly reflects that the stripped remaining oil has detached from its original distribution state, enabling efficient exploitation of “difficult-to-mobilize remaining oil.”

5. Conclusions

(1)

The distribution state of multicomponent crude oil in nanopores is synergistically regulated by a component’s physicochemical properties, intermolecular interactions, and solid–liquid interfacial interactions. Aromatics (toluene), due to their conjugated π-bond structure, exhibit specific adsorption with polar SiO₂ pore walls through strong π-π interactions, accumulating near the pore walls. N-hexane, with weak polarity and low intermolecular van der Waals forces, tends to disperse in the energy-stable central region of pores.

(2)

Pore size significantly influences the distribution state of crude oil. Small-sized pores (2 nm) impose strong confinement effects, forcing n-hexane molecules to form ordered layered structures near pore walls, resulting in intensely fluctuating density curves with multiple high-intensity peaks. As pore size increases, confinement effects weaken, molecular alignment order decreases, density fluctuation amplitude reduces, and n-hexane molecules gradually disperse more diffusely within pores.

(3)

Rock type affects the adsorption energy between crude oil and rock surfaces. The calcium–magnesium ion crystal structure of dolomite surfaces enables stronger van der Waals interactions with n-hexane, as well as particularly significant contributions from dispersion forces, elevating the adsorption energy with n-hexane. Thus, dolomite exhibits stronger adsorption capacity for n-hexane than SiO₂.

(4)

CO₂ flooding exerts a critical impact on crude oil distribution states. CO₂ preferentially occupies oil-wet SiO₂ pore wall sites through competitive adsorption, weakening direct interactions between n-hexane/toluene and pore walls. It demonstrates a certain stripping effect on n-hexane and triggers the redistribution of toluene within pores. Collectively, this reduces the adsorption strength between crude oil and rock surfaces, promoting crude oil desorption and flow, thereby effectively enhancing oil recovery.

(5)

NMR results show that water flooding cannot effectively strip crude oil from pores, while CO₂ can form a competitive adsorption state with crude oil on pore walls, thereby enabling the stripping of crude oil from pores.