CCUS, the abbreviation for Carbon Capture Utilization and Storage technology, has been developed in recent years with the global attention to greenhouse gas emissions and other ecological issues. CCUS technology is not only effective in improving global warming, but also significant for achieving zero carbon emissions by 2050. According to the International Energy Agency, CCUS technology will achieve 38% of the emission reduction in the chemical and engineering industry and is expected to contribute about 14% of the CO₂ emission reduction by 2050 (Mi and Ma, 2019; Jin and Chen, 2020; Qin et al., 2020; Zhang et al., 2022). CCUS technology can be further divided into oil and gas reservoir utilization (CCUS-EOR/EGR), chemical utilization, and biological utilization depending on the utilization method. CCUS-EOR technology is to inject the captured CO₂ into a reservoir in the development stage with complete geological structure and adequate basic information, to enhance the recovery of crude oil through CO₂ flooding and to achieve CO₂ burial (Ku et al., 2023). This technology is not only highly economically viable, but also has a wide range of applications and a very promising future.

Overseas CO₂ flooding originated in the 1950s (Whorton et al., 1952). In the 1980s, it was gradually applied in the oil field and began to be promoted commercially. Since 2000, the globe concern about environmental issues and the continuous development of engineering technology have further promoted the innovation of CO₂ flooding technology (Shen and Liao, 2009). Around the 1960s, China started to pay close attention to CO₂ flooding technology and conducted indoor flooding experiments, and carried out field tests in layer 4-7 of PuⅠin Daqing oilfield and in the eastern transition zone of Sanan; during the 1990s, Jiangsu Fumin oilfield carried out experiments on CO₂ huff and puff (Chen et al., 2011). Before entering the 21st century, due to the lack of understanding and experimental support for this technology, coupled with the lack of CO₂ gas sources in China and serious gas channeling in the oilfield experiments, the development of CO₂ flooding technology in China had been slow (Gao et al., 2014; Zhao et al., 2017; Li, 2018). Over past 10 years, the internal CO₂ flooding and storage field tests had made a remarkable progress which had three phases: pre-experiments, extended tests and industrial application, with increasingly abundant types of reservoirs and scales of tests, achieving certain results and accumulating some practical experience (Luo et al., 2013). Among all enhance oil recovery (EOR) technologies, the most well-known and effective way to enhance the oil production efficiency is to inject CO₂ into the reservoir, furthermore, the injected CO₂ can achieve the goal of CCUS energy saving and emission reduction (Yuan et al., 2020). The reasons in terms of mechanism why CO₂ flooding can enhance oil recovery is that the solution of CO₂ in the crude petroleum causes volume expansion of crude oil and increases oil saturation in the pore; in addition, CO₂ can diffuse easily inside oil and water, giving the opportunity that the CO₂ can redistribute and stabilize the balance of phase system, which acts an effective part in the relative permeability of oil and water (Hu et al., 2017; Guo et al., 2018; Jia et al., 2019; Li et al., 2019; Li, 2020; Jia et al., 2021).

To take into account the limitations of the complex dissolving and diffusing action of CO₂ inside oil and water, and the inability of indoor experiments to simulate the microscopic oil drive efficiency of CO₂, the author chose the molecular dynamics simulation method to carry out the research. Molecular dynamics simulation is a non-quantum mechanical method to work out problems at the molecular level based on the fundamental theory of classical Newtonian mechanics. The calculation process mainly relies on the position or average configuration of the nucleus to establish the required force field function to describe the molecular structure and energy, which is a more widely used calculation method, and it has certain advantages in analyzing the changes of microscopic properties such as intermolecular forces and molecular morphology. A crude oil system model which contains multiple alkane components is constructed and the action of dissolution-diffusion of CO₂ in the multi-component crude oil model is simulated by applying the simulation method of molecular dynamics in this article. The CO₂ density distribution in crude oil as well as the volume change of crude oil are calculated, and the microscopic impact mechanism of CO₂ dissolution-diffusion on the volume expansion of crude oil is further analyzed. The dissolution and diffusion coefficients of CO₂ were predicted based on the variation law of mean square displacement between crude oil molecules, and the impact of CO₂ dissolution diffusion on crude oil mobility was analyzed. The mechanism of CO₂ dissolution-diffusion in multi-component crude oil is revealed at the microscopic level, and provides theoretical guidance for the development of CO₂ flooding.

Primarily based on the results of laboratory determination of the chemical compositions of crude oil, a model of crude oil was constructed within our paper, the crude oil model is mainly composed of octane, including 17 kinds of alkane components such as methane and ethane (Table 1), the density at 20°C of this model is 0.82 g/cm³. The Visualize module in Material Studio software was used to build a model of each alkane molecule in the crude oil component (as shown in Figure 1). In addition, a molecular model of CO₂ (shown in Figure 2) was constructed for subsequent simulation of the CO₂ motion in the crude oil model (Jia et al., 2021).

In accordance with the established molecular models of alkanes and CO₂ above, the geometry and energy optimization of each molecular model were carried out and the optimized the molecular model structure was obtained by using the Forcite component, as shown in Figure 3. Figure 4 indicated that we used the Construction tool which is belong to the Amorphous Cell Tools component to build the multi-component crude oil molecular system model in accordance with the experimental testing consequences of crude oil components with the dimensions of 42.67 × 42.67 × 42.67 Å³ and the density of it is set to 0.82 g/cm³ in accordance with the real crude oil density. To ensure that the properties of the system remain unchanged during the simulation process, the crude oil system model is therefore subjected to three-dimensional periodic boundary conditions. CO₂ was added to the multi-component crude oil molecular system model, and the temperature were set to 313 K, 353 K, 393 K and the pressure were set to 10–50 MPa, in units of 10 MPa respectively in order to be similar to the reservoir properties. This part is a preparation for the research of the action of the CO₂ dissolution and diffusion in crude oil under different reservoir conditions. It should be noted that the idea of establishing the model is also applicable to other multi-component fluids including two phases of oil and gas or three phases of oil, gas and water, but it needs to be further determined according to experimental results and research objectives. Multi-component fluids are more reasonable in real formation fluid characterization and simulation.

Author used the BIOVIA Material Studio 2019 software package for molecular dynamics simulations. First, the geometric and energy optimization of the multicomponent crude oil model was performed using steepest descent method to obtain the optimized structures for different temperature and pressure conditions. After that, the kinetic equilibrium was performed, and the force field parameters for both CO₂ and n-alkanes were assigned using the COMPASS force field developed by Sun et al. which is suitable for organic molecules and inorganic covalent bond molecular systems, and its potential energy function form is expressed as: E t o t a l = E b o n d + E a n g l e + E t o r s i o n + E c r o s s + E v d w + E c o u l o m b i c (1) where the first previous four terms show the bonding energy and the last two terms denote the non-bonding interaction energy. The non-bonding interaction energy can be represented by the Lennard-Jones 9-6 potential and the Coulomb electrostatic potential. The Lennard-Jones 9-6 potential describes the potential energy due to the weak repulsive and gravitational forces of the van der Waals force. Its specific form is: E = ∑ i > j q i q j r i j + ∑ i > j E i j 2 r i j 0 r i j 9 − 3 r i j 0 r i j 6 (2) r i j 0 = r i 0 6 + r j 0 6 2 1 6 (3) where ij denotes an atomic pair, E _ij is the potential well depth, r _ij ⁰ is the zero potential distance of the atomic pair, and r is the distance between two atoms.

The other kinetic simulation parameters are: the canonical ensemble (NVT), temperature control using the Nosé-Hoover method, with the temperature set to 313, 353, and 393 K respectively; after that, the isothermal isobaric system synthesis (NPT) is selected, with the pressure set to 10–50 MPa, in units of 10 MPa, respectively, and the Berendsen method for pressure control. All other parameters were consistent in both equilibria, including: Ewald and Atom Based ways were adopted for electrostatic interaction as well as van der Waals interaction, separately. The dynamic simulation of 1 ns (1,000,000 steps) was performed with a step size of 1 fs at a distance of 1.25 nm. The consequences of the simulation procedure were exported every 1,000 steps which divided into two parts, the first part including 400 ps, they were applied to maintain the balance of the system, and the other part containing 600 ps was utilized for collecting the density data in whole system.

The way in which we calculate the crude oil molecular system density at a temperature of 20°C is to apply the simulation of molecular dynamics, afterwards, we compared it with the US NIST database, as shown in Figure 5. The difference between the value of crude oil system density as determined via molecular dynamics simulation that used in our article and the density value of reference given by NIST database is slight, which indicating that the established multi-component crude oil system model and procedure of simulation are reasonable.

Zhao et al. (2016a) used a PVT instrument to measure the solubility change of CO₂ in crude oil and formation water under reservoir conditions respectively and express those changes through curve representation. It was found that with the enhancement of reservoir pressure, the ability of CO₂ solution in both oil and water enhanced, and the ability of CO₂ solution within oil phase was about 7 times that in water.

Ao et al. (2019) studied the dissolution and diffusion law in crude oil and brine layer by using the pressure depletion method, and found that the reservoir pressure, temperature and brine concentration all directly affect the dissolution process of CO₂ in brine layer. The higher the salt concentration in the brine layer, the lower the CO₂ dissolution and the slower the diffusion rate.

At a certain temperature, the adsorption isotherm is plotted by calculating the number of small molecules adsorbed in the periodic box at different pressure points, which represents the relationship between the concentration of adsorbed gas and its partial pressure p at a certain temperature. Where the slope of the adsorption isotherm for a pressure value of 0 is the dissolution coefficient S (Jin et al., 2020; Jia et al., 2023): S = lim p → 0 c p (4) where S is the dissolution coefficient, dimensionless. This factor represents the maximum number of grams of solute dissolved in 100 g of solution at a given temperature and pressure.

The dissolution coefficient decreases with the increasing temperature. While determining the dissolution rate and final CO₂ dissolution in crude oil is the diffusion coefficient, the larger the diffusion coefficient, the larger the diffusion flux, which determines the gas distribution in the reservoir at different times and affects and improves the physical characteristics of crude oil (Chen et al., 2010). Grogan and Pinczewski (1984) established a mathematical model for the diffusion coefficient under atmospheric pressure conditions by directly observing the interfacial movement of the oil or water phase during CO₂ diffusion. Renner (1988) used artificial cores to test the coefficients of CO₂ and rich gas diffusing motion in crude oil at high pressure. Riazi (1996) used the pressure drop method to study the diffusion coefficients between gas and crude oil (Zhang et al., 2000). Wang (1996) systematically described the CO₂ mixed-phase drive diffusion percolation equation, molecular diffusion coefficient and physical dispersion coefficient models and related experimental determination methods (Li et al., 2001; Fu et al., 2003). Zhao et al. (2016b) proposed an approximate calculation model of the variable diffusion coefficient applicable to the pressure drop method, but the effect of crude oil expansion on diffusion was not considered. In response to the inability of previous authors to comprehensively consider the impacts of varying temperatures and pressures on the diffusion of dissolved CO₂ in crude oil in indoor experiments. This paper takes advantage of molecular dynamics simulations at the microscopic level to carry out simulation studies.

The diffusion coefficient in this paper is calculated from the mean square displacement and is mainly expressed as (Jin and Chen, 2019; Jin et al., 2020): D = 1 6 N α lim t → ∞ d d t ∑ i = 1 N α r i t − r i 0 2 (5) where Nα denotes the diffusion of atoms in the system and ri(t) represents the displacement vector of molecule i from 0 to time t. The diffusion coefficient is obtained from the best trend line of the MSD curve y = ax + b. It is specifically obtained from the following equation: D = a 6 (6)

According to the Strokes-Einstein formula, different diffusion coefficients of CO₂ lead to different viscosities of crude oil and ultimately different crude oil fluidity. The two show an inverse relationship, mainly calculated by the following equation: μ = k T 6 π α D (7) where T represents the simulated temperature, α represents the molecular radius of CO₂, which is approximately 1.65 × 10⁻⁸ cm, k denotes the Boltzmann constant, which is 1.38 × 10⁻²³ J/K, D denotes the diffusion coefficient, as well as μ represents the crude oil viscosity.

The dissolution and diffusion coefficients of CO₂ were calculated mainly by Material Studio software. In the simulation process, the software will count the specific dissolution amount of CO₂ at different times, and according to these dissolution amounts, the dissolution coefficient of CO₂ at different times can be calculated. The diffusion coefficient was calculated by calculating the change of Mean Square Displacement of CO₂ over a period of time, and then solving the diffusion coefficient according to linear regression fitting.

In this paper, we established a multi-component model close to the real formation crude oil and employed molecular dynamics simulations to investigate the dissolution and diffusion process of CO₂ in the crude oil systems, and its influence on the mobility of crude oil was examined. By establishing a multi-component crude oil model consistent with the results of experiments, the dissolution-diffusion procedure of CO₂ in multi-component crude oil at 313 K, 353 K, 393 K and 10–50 MPa, in units of 10 MPa were simulated. Our aims were to quantitatively characterize the density distribution of CO₂ dissolved in crude oil, reveal the mechanism of crude oil expansion and improvement of the crude oil mobility due to CO₂ dissolution and diffusion, and systematically analyze the effects of temperature and pressure on the dissolution and diffusion of CO₂. In addition, the research ideas in this manuscript are also applicable to the molecular dynamics simulations of other fluid models. We expect that our study will provide some insights into the mechanism of CO₂-EOR and provide some ideas for molecular dynamics simulations. The following conclusions were obtained:

1) By simulating the reservoir status including temperature and pressure, the dissolving process of CO₂ in multi-component crude oil was observed as well as the density distribution of CO₂ was analyzed; with the change of temperature or pressure, the movement of CO₂ was intensified and driven into the interstices of alkane molecules.