Skip to main content
NCBI home page
RSC Advances logoLink to RSC Advances
. 2026 Jul 2;16(34):33135–33145. doi: 10.1039/d6ra02512a

Mechanistic insight into multiscale self-assembly and viscoelastic enhancement in a CO2-responsive amine/zwitterionic surfactant formulation

Mingwei Gao a,b,✉, Zhuoqing Li a,b, Xiaonan Feng c, Xianyun Shi d, Bo Yan a,b, Zixuan Wang a,b, Qing You e,✉, Caili Dai a,b,✉
PMCID: PMC13326686  PMID: 42395735

Abstract

CO2-responsive viscoelastic surfactant (VES) fluids are of growing interest for CO2 flooding and fracturing in oil and gas development, yet the molecular mechanism by which CO2-induced speciation, counterion environment, and hydrotrope-assisted association collectively regulate self-assembly and viscoelasticity in mixed amine/zwitterionic systems remains insufficiently resolved. Here, we use a previously reported amine/zwitterionic formulation composed of EKO, EAHSB and NaPts as a model system to clarify the structure–interaction–rheology relationship responsible for its non-additive viscoelastic enhancement. Rheological measurements show that the mixed EAHSB/EKO/NaPts-CO2 system exhibits composition-dependent enhancement in zero-shear viscosity and elastic modulus, indicating cooperative supramolecular association rather than simple compositional averaging. Cryo-EM reveals the coexistence of vesicles, ring-like features, and thick cylindrical structures, pointing to a more diverse multiscale assembly landscape in the mixed system. 1H NMR and coarse-grained molecular dynamics simulations further suggest that counterion environment and NaPts-involved interfacial reorganization alter the mixed headgroup region and are associated with richer multiscale assembly and enhanced viscoelastic response. Taken together, these results establish a mechanistic framework linking CO2-induced protonation, NaPts-mediated mixed-headgroup organization, multiscale self-assembly, and macroscopic viscoelastic enhancement, thereby providing guidance for the rational design of CO2-switchable amine/zwitterionic surfactant fluids.

NaPts-mediated headgroup organization links CO2-induced protonation to multiscale self-assembly and enhanced viscoelasticity in mixed amine/zwitterionic surfactants.graphic file with name d6ra02512a-ga.jpg

Introduction

Since the industrial revolution, atmospheric CO2 concentrations have risen from ∼290 ppm to ∼420 ppm due to anthropogenic emissions, driving global warming and inducing persistent disturbances to ecosystems and biogeochemical cycles, including ocean acidification. Consequently, reducing CO2 emissions and stabilizing atmospheric CO2 concentrations has become a global priority.1–3 Among the technological pathways proposed to achieve net-zero greenhouse gas emissions by 2050, carbon capture, utilization and storage (CCUS) is widely regarded as indispensable. While carbon capture and storage (CCS) directly targets large-scale emission reduction, its overall economics are often constrained by the energy- and cost-intensive capture step.4–8 CO2 utilization, by assigning industrial value to captured CO2 streams, can alleviate cost pressures and enhance deployment incentives, and has therefore attracted widespread attention from scientists and engineers. Injecting CO2 into oil and gas reservoirs—for example via CO2 flooding, CO2 huff-and-puff, and CO2-based fracturing—represents a promising utilization route that can simultaneously enhance hydrocarbon recovery and provide subsurface CO2 storage.9–13

CO2-responsive viscoelastic surfactant (VES) systems, which offer on-demand and reversible control over viscosity, self-assembly, and microstructure, show great promise for channeling control in CO2 flooding, CO2 foam flooding, CO2 foam fracturing, and CO2-water hybrid fracturing.14–18 In such systems, CO2 responsiveness is directly coupled with surfactant self-assembly, so that CO2-induced changes in molecular state can be translated into changes in viscosity, elasticity, and microstructure. In recent years, single-component CO2-responsive VES systems (formulated from a single CO2-responsive surfactant) have been extensively investigated for such applications.19–21 In most cases, tertiary amine-type surfactants are designed to form micelles or wormlike micelles (WLMs) once protonated in the presence of CO2-generated carbonic acid, enabling reversible thickening and viscoelastic network formation.22–26 For instance, Zhang et al.19 developed a CO2-responsive VES mobility control system based on WLMs using N,N-dimethyloctylamide-propyl tertiary amine, and Wu et al.21 developed a CO2-responsive VES fracturing fluid using erucamidopropyl dimethylamine as the thickener; both studies highlight reversible CO2-triggered gelation-breakage behavior and formation-relevant potential. Nevertheless, most reported CO2-responsive VES systems based on single surfactants still face practical limitations, including relatively high surfactant dosages to reach operational viscosity, limited robustness under deep-reservoir conditions, and an incomplete mechanistic understanding of how CO2-triggered speciation and counterion environment regulate self-assembly and rheology. In VES fluids, macroscopic viscosity and elasticity generally arise from the formation of elongated micelles, wormlike micelles, vesicles, or transiently connected aggregate networks. In single-component tertiary amine VES systems, CO2-induced protonation can increase aqueous compatibility and promote micellar growth or WLM formation. In zwitterionic VES systems, strong hydration and internal charge compensation can improve tolerance to temperature and salinity. In binary or multicomponent VES systems, complementary headgroup interactions, counterion screening, and additive-assisted packing may further enhance aggregate growth and transient network formation.

In recent years, zwitterionic VESs, whose hydrophilic headgroups contain both anionic and cationic groups, have been shown to exhibit superior high-temperature stability and therefore offer greater application potential in high-temperature reservoirs.27–29 However, single-component zwitterionic VES fluids typically require relatively high concentrations to achieve strong viscoelasticity, which increases chemical costs and hinders large-scale field use.30 From a VES-fluid design perspective, an appealing strategy is therefore to construct multicomponent VES systems that couple the CO2-switchability of tertiary-amine surfactants with the robustness of zwitterionic surfactants, while using suitable additives to promote cooperative packing, self-assembly, and viscoelastic network formation.

In this context, a multicomponent CO2-responsive VES formulation comprising zwitterionic surfactant (erucamidopropyl hydroxypropyl sulfobetaine, EAHSB), CO2-responsive tertiary amine surfactant (erucamidopropyl dimethylamine, EKO), and aromatic hydrotrope sodium p-toluenesulfonate (NaPts) was previously reported by our group31 to exhibit markedly enhanced viscoelasticity relative to the corresponding binary systems at the same overall surfactant concentration. However, that study mainly established the formulation feasibility and macroscopic performance of the system, whereas the molecular origin of the enhancement remains unresolved. In particular, it remains unclear whether the stronger viscoelasticity arises from simple compositional addition or from cooperative mixed-headgroup organization, how CO2-protonated EKO interacts with the zwitterionic EAHSB headgroups, and how Pts− participates in interfacial packing, aggregate evolution, and network formation.

Accordingly, the present work focuses on the unresolved mechanism underlying the viscoelasticity enhancement of the CO2-responsive EAHSB/EKO/NaPts-CO2 formulation. This study examines the relationship between molecular interactions, aggregate structures, and rheological behavior in this mixed amine/zwitterionic/hydrotrope system by correlating composition-dependent rheology, cryo-EM-observed morphologies, 1H NMR chemical-shift changes, and CG-MD-derived counterion/headgroup organization. These results help distinguish cooperative association in the ternary system from a simple combination of the corresponding EAHSB/NaPts and EKO/NaPts-CO2 binary systems, and suggest that CO2-induced protonation, NaPts-involved interfacial organization, and mixed-headgroup association are closely related to the diverse self-assembled structures and enhanced viscoelastic response. By linking molecular-level interactions with observed aggregate morphologies and rheological properties, this work provides mechanistic guidance for designing CO2-switchable amine/zwitterionic surfactant fluids.

Experimental section

Materials

The zwitterionic surfactant EAHSB and the CO2-responsive surfactant EKO were synthesized in our laboratory following procedures reported in the literature.32 The 1H NMR spectra of EAHSB and EKO are shown in the SI. NaPts (99.7% purity), used as an aromatic hydrotrope, was purchased from Shanghai Aladdin Biochemical Technology Co., Ltd (China). Standard solutions of hydrochloric acid (HCl, 1.0 mol L−1) and sodium hydroxide (NaOH, 1.0 mol L−1), utilized for pH adjustment, were purchased from Shenzhen Borinda Technology Co., Ltd High-purity CO2 (99.999%) was supplied by Qingdao Deyi Gas Co., Ltd (Qingdao, China) and used as received. Ultrapure water was used as the solvent for all aqueous solutions.

Sample preparation

Surfactant (EKO and EAHSB) solutions and the EAHSB/NaPts mixtures were prepared by dissolving accurately weighed amounts of each component in ultrapure water, followed by gentle magnetic stirring at ambient temperature (25 °C) until the samples became visually homogeneous. For the CO2-responsive EKO sample (denoted EKO-CO2), high-purity CO2 was bubbled through the EKO solution until the pH reached a stable value (i.e., no further change with continued bubbling). Acidified control samples (denoted EKO-HCl) were prepared by adding calculated volumes of a 1.0 M HCl standard solution to obtain target pH values of 5.38 and 1.40. To prepare the CO2-treated mixed EAHSB/EKO/NaPts-CO2 system, the accurately weighed components were first dispersed in ultrapure water under gentle stirring. High-purity CO2 was then bubbled through the mixture until the pH stabilized and the formulation became visually homogeneous. For the composition-dependent viscosity measurements, the total surfactant concentration (EAHSB + EKO) was fixed at 50 mM, with EAHSB : EKO molar ratios of 0 : 10, 1 : 9, 3 : 7, 5 : 5, 6 : 4, 7 : 3, 9 : 1, and 10 : 0. The NaPts concentration was fixed at 25 mM. For the representative binary and ternary systems compared, the formulations were 50 mM EAHSB/25 mM NaPts, 50 mM EKO/25 mM NaPts-CO2, and 25 mM EAHSB/25 mM EKO/25 mM NaPts-CO2, respectively. It should be noted that EKO-containing formulations without CO2 treatment, including EKO/NaPts and EAHSB/EKO/NaPts, could not form stable homogeneous aqueous solutions at the concentrations used in this study because neutral EKO has limited aqueous solubility. Therefore, these non-CO2-treated EKO-containing samples were not used for quantitative rheological or cryo-EM comparison. These concentrations were selected for mechanistic comparison among the binary and ternary VES systems. Practical field formulations would require further optimization according to reservoir temperature, salinity, cost, and operational requirements.

Methods

1H NMR spectroscopy

Samples were prepared in a mixed solvent of deuterium oxide (D2O) and deuterated methanol (CD3OD). Chemical shifts are reported in parts per million (ppm) and referenced to the residual HDO and methanol-d4 signals. Unless otherwise stated, the pH values discussed for CO2-treated and HCl-acidified samples refer to the corresponding aqueous formulations prior to NMR sample preparation.

Rheological characterization

Rheological properties were investigated using a HAAKE MARS 60 rotational rheometer (Thermo Fisher Scientific, Germany) with a temperature-controlled coaxial cylinder geometry (CC31/Ti rotor, Z43 cup). Steady shear flow curves (viscosity vs. shear rate) and dynamic oscillatory sweeps (storage modulus G′, loss modulus G″, complex viscosity vs. angular frequency) were performed at 25 °C. Prior to frequency sweeps, amplitude sweeps were conducted at 1 Hz to ascertain the linear viscoelastic region. Evaporation was minimized using a solvent trap.

Cryogenic electron microscopy (cryo-EM)

Cryo-EM was used to visualize the native micro/nanostructures with minimal drying or beam-damage artifacts. For cryogenic scanning electron microscopy (cryo-SEM), samples were rapidly vitrified in liquid nitrogen slush, fractured under cryogenic conditions, and etched by controlled sublimation. The fractured surfaces were sputter-coated with Pt and imaged using a Hitachi SU3500 SEM equipped with a cryo-stage. For cryogenic transmission electron microscopy (cryo-TEM), a small aliquot of sample was deposited onto a carbon-coated copper grid, blotted to form a thin film, and plunge-frozen to vitrify the specimen. Cryo-TEM images were acquired on a JEOL JEM-1400 transmission electron microscope operated at 120 kV.

CG-MD simulation

CG-MD simulations were performed using the GROMACS software package to probe the relative self-assembly tendencies and interfacial organization of the surfactants and additives in the bulk phase.33 Simulations were conducted in the NPT ensemble at 298 K and 0.1 MPa. During equilibration, temperature and pressure were controlled using Berendsen coupling; production runs were then carried out using the velocity-rescale thermostat and Parrinello–Rahman barostat.34–36 A time step of 20 fs was used, and trajectories were visualized with VMD.37 This value was chosen because 20 fs is a commonly used and recommended time step for Martini 3 coarse-grained simulations, and it represents the lower end of the typical 20–30 fs range recommended for Martini models. In the present charged surfactant/counterion system, this relatively conservative time step was adopted to maintain stable integration while still allowing efficient sampling of self-assembly. The Martini 3 coarse-grained force field was adopted,38 and the bead-mapping schemes for EAHSB, EKO, EKOH+, and Pts− are shown in Fig. S1. To accelerate self-assembly and improve sampling within accessible simulation timescales, the total concentrations in the CG models were set higher than those in the experiments while maintaining the same relative composition.39,40 This elevated-concentration strategy is commonly used in surfactant self-assembly simulations because aggregate nucleation and morphological evolution are slow at dilute experimental concentrations and would require much larger simulation boxes and longer trajectories. In the present work, the simulated systems were therefore used to examine relative self-assembly tendencies, headgroup/counterion organization, and possible interfacial association motifs, rather than to reproduce the experimental concentration-dependent morphologies in a one-to-one manner. The molar ratio of EAHSB : EKO : EKOH+ : Pts− was fixed at 10 : 4 : 6 : 10 as a representative model for the partially protonated CO2-treated condition. This ratio is intended as a plausible comparative composition rather than as an exact speciation measurement. The simulated concentrations and their fold differences relative to the experimental formulation are summarized in Table S1. Because of the elevated concentrations and coarse-grained representation, the CG-MD results are interpreted as qualitative and comparative structural tendencies rather than quantitative reproductions of the experimental assemblies. All-atom simulations at experimental concentrations would be valuable in future work to further refine the local interaction picture. Initial configurations were generated using PACKMOL by randomly placing the surfactant/hydrotrope molecules and water beads in the simulation box (Fig. S2).41

Results and discussion

In this section, we focus on how mixed-headgroup interactions and NaPts-associated interfacial reorganization regulate the rheology, multiscale self-assembly, and viscoelasticity of the EAHSB/EKO/NaPts-CO2 system. Supporting reference data for single-component EKO under CO2 and HCl treatment are provided in the SI (Fig. S3 and S4) and are used only as contextual baseline information.

Composition-dependent viscoelastic enhancement and multiscale self-assembly in EAHSB/EKO/NaPts-CO2

Composition-viscosity response

Here, viscoelastic enhancement refers to the simultaneous increase in zero-shear viscosity, plateau storage modulus, and elastic-dominated response, together with a smaller estimated network mesh size. From an application perspective, these properties are relevant to mobility control, fluid diversion, and the stability of CO2-related oilfield working fluids. In this comparison, EAHSB/NaPts was used as a non-CO2-responsive zwitterionic binary reference because EAHSB does not contain the tertiary amine group responsible for CO2-induced protonation, whereas EKO/NaPts-CO2 was used as the CO2-responsive amine binary reference. Fig. 1a shows the zero-shear viscosity (η0) of the EAHSB/EKO/NaPts-CO2 system at a fixed total surfactant concentration while varying the EAHSB : EKO molar ratio. The viscosity increases at intermediate compositions and reaches a maximum at the 5 : 5 ratio, then decreases when either component becomes dominant. Compared with the EAHSB/NaPts and EKO/NaPts-CO2 binary systems, the ternary formulation shows a much higher η0. The steady-shear curves in Fig. 1b show shear-thinning behavior for all samples, but the ternary system maintains a higher low-shear viscosity, indicating stronger association in the mixed formulation.

Fig. 1. Influence of the molar ratio (EAHSB : EKO) of VES on the zero-shear viscosity (a); steady-shear viscosity (b) and dynamic modulus curves (c) of EAHSB/NaPts, EAHSB/EKO/NaPts-CO2, and EKO/NaPts-CO2 aqueous solutions. Filled symbols represent storage modulus G′, and open symbols represent loss modulus G″.

Fig. 1

Open in a new tab

The dynamic rheological properties (Fig. 1c and Table 1) further substantiate the enhanced elasticity in the mixed system. For EAHSB/NaPts and EKO/NaPts-CO2, the storage modulus (G′) remains relatively low, although G′ exceeds G″ over most of the measured frequency range, indicating a weakly elastic-dominated response. The relatively low G″ at low frequencies for the ternary system indicates weaker viscous dissipation and a more elastic-dominated response in this frequency range. This feature is considered together with η0, Inline graphic, and the derived network parameters, rather than being used alone to assign a specific aggregate structure. The difference in G′ between the ternary system and EKO/NaPts-CO2 is moderate, but the ternary system shows a higher estimated plateau modulus and a much higher zero-shear viscosity. Therefore, the viscoelastic enhancement is evaluated from the combined changes in η0, Inline graphic, and the derived network parameters rather than from the frequency–sweep curves alone. Table 1 lists comparative micellar–network parameters for the different systems, where ξM is the mesh size, le is the contour length between entanglements, and Lc is the contour length of the micelles. The standard transient–network relationships used to obtain these comparative micellar–network parameters are provided in SI Section S3. The higher plateau modulus, together with the broader elastic-dominated response of the mixed system, is consistent with slower structural relaxation and a much larger zero-shear viscosity, in agreement with the steady-shear measurements. Compared with the two binary reference systems, the mixed system exhibits a smaller average mesh size and a shorter contour length between entanglements. These parameters indicate that the superior viscoelastic response of the mixed system is more plausibly associated with a more efficiently percolated network, characterized by smaller mesh size, shorter entanglement spacing, and higher effective network density, rather than with contour length alone.

Table 1. Micellar characteristic parameters of EAHSB/NaPts, EAHSB/EKO/NaPts-CO2, and EKO/NaPts-CO2 aqueous solutions.
graphic file with name d6ra02512a-t3.jpg graphic file with name d6ra02512a-t4.jpg ξ M/nm l e/nm L c/nm
EAHSB/NaPts 0.17 ± 0.02 4.91 ± 0.34 94.34 ± 2.15 321.42 ± 12.19 9343.82 ± 1023.31
EAHSB/EKO/NaPts-CO2 0.35 ± 0.02 12.39 ± 0.70 69.26 ± 1.31 192.03 ± 6.03 6864.54 ± 139.68
EKO/NaPts-CO2 1.21 ± 0.12 7.23 ± 0.40 82.89 ± 1.56 259.06 ± 8.14 1552.85 ± 168.30
Open in a new tab

These parameters were estimated using standard transient–network relationships and are used mainly for comparative evaluation of the relative network characteristics of the three systems, rather than as absolute structural dimensions of uniquely defined micelles.

Overall, Fig. 1 shows that the ternary EAHSB/EKO/NaPts-CO2 formulation gives higher η0 and Inline graphic than either binary reference system, especially at intermediate EAHSB : EKO ratios. This non-additive response suggests cooperative association in the mixed formulation rather than simple compositional averaging. We next examine whether this rheological behavior is accompanied by changes in aggregate morphology.

Aggregate morphologies

Cryo-SEM was used mainly to compare the overall frozen-fracture morphology at the micrometre scale, rather than to directly assign nanoscale aggregate types. The nanoscale aggregate morphology was therefore interpreted primarily from the cryo-TEM images. To compare the aggregate morphologies associated with these rheological differences, the EAHSB/NaPts, EKO/NaPts-CO2 and EAHSB/EKO/NaPts-CO2 samples were examined by cryo-EM. Cryo-SEM images (Fig. 2a, b, and c) show that the EAHSB/NaPts system forms a relatively loose network of thin lamellae, with visible voids and a comparatively open texture. The EKO/NaPts-CO2 system also exhibits elongated aggregates and sheet-like domains, but the overall network appears less continuous, with larger pores and fewer obvious junctions between structures. In striking contrast, the EAHSB/EKO/NaPts-CO2 mixed system develops a more compact microstructure. Cryo-SEM reveals thickened strands and irregular but relatively smaller voids that are more evenly dispersed across the observed fractured surface than in the two binary counterparts, indicating a comparatively finer and more compact network texture. Here, the improved void uniformity refers to the reduced occurrence of large isolated pores and the more even dispersion of voids within the network, rather than a regular or periodic pore arrangement.

Fig. 2. Cryo-SEM (a, b and c) and cryo-TEM (d, e and f) images of EAHSB/NaPts, EKO/NaPts-CO2, and EAHSB/EKO/NaPts-CO2 aqueous solutions.

Fig. 2

Open in a new tab

Cryo-TEM images (Fig. 2d, e, and f) provide more detailed information at the nanoscale. The EAHSB/NaPts system shows elongated micellar features that are broadly consistent with WLM-like aggregates, exhibiting high structural homogeneity. The EKO/NaPts-CO2 system predominantly self-assembled into giant multilamellar vesicles hundreds of nanometers in size, displaying a complex “multi-chamber nested” internal morphology. The vesicles appeared significantly deformed due to dense packing, suggesting strong interactions at the soft matter interfaces. Remarkably, the EAHSB/EKO/NaPts-CO2 mixed system exhibited exceptional structural diversity: it not only retained some WLMs and vesicles but also featured prominent ring-like features (approximately tens of nm in diameter) and larger, thicker elongated cylindrical structures. Cryo-TEM images provide clearer nanoscale differences among the three systems. EAHSB/NaPts mainly forms long WLMs with relatively uniform morphology, whereas EKO/NaPts-CO2 forms large multilamellar vesicles. In contrast, the ternary EAHSB/EKO/NaPts-CO2 sample contains coexisting WLMs, vesicles, ring-like features and thicker cylindrical structures. Imaging alone cannot assign the mechanical role of each morphology, but the broader distribution of aggregate structures suggests that the ternary system is not a simple superposition of the two binary assemblies. This morphological diversity is consistent with the higher η0 and Inline graphic observed for the ternary formulation.

To probe the local interactions behind these morphological differences, we next examine the headgroup environment using 1H NMR and CG-MD analysis.

Molecular-level headgroup organization and NaPts-assisted association

After establishing the differences in rheological behavior and aggregate morphology, we next used 1H NMR spectroscopy and CG-MD analysis to examine whether these changes are accompanied by differences in the local headgroup environment. To gain deeper insight into the molecular basis of the observed synergy, 1H NMR spectra (Fig. 3a) and the schematic model in Fig. 3b are used to interpret how NaPts-assisted association reorganizes the mixed headgroup region at the molecular level. Supporting binary-system 1H NMR reference data and plausible schematic packing models for the EAHSB/NaPts and EKO/NaPts-CO2 systems are provided in the SI (Fig. S5–S8). Compared with the two binary reference systems, the mixed EAHSB/EKO/NaPts-CO2 solution shows noticeable chemical-shift changes for protons located near the headgroups and amide linkages of both surfactants. For EAHSB, the methylene protons adjacent to the sulfobetaine moiety experience downfield shifts, indicating a more constrained and polar environment around the zwitterionic headgroup. For EKO, protons near the tertiary amine and amide segments are also shifted, reflecting altered local polarity and hydrogen-bonding patterns in the presence of EAHSB and NaPts.

Fig. 3. 1H NMR spectra of EAHSB/NaPts, EAHSB/EKO/NaPts-CO2, and EKO/NaPts-CO2 solutions (a); plausible schematic packing model used to illustrate NaPts-involved headgroup organization in the EAHSB/EKO/NaPts-CO2 system (b). The schematic should not be regarded as a uniquely determined three-dimensional structure.

Fig. 3

Open in a new tab

Because 1H NMR alone cannot uniquely define the three-dimensional assembly structure, Fig. 3b is presented only as a plausible schematic model to aid discussion rather than as direct structural proof. The aromatic protons of NaPts remain clearly visible in all formulations, but their chemical environment is perturbed in the mixed system, suggesting that the local environment of NaPts changes upon mixing and is more closely associated with the interfacial headgroup region.42–44 The aromatic protons of NaPts also show a changed local environment in the ternary formulation, suggesting closer involvement of Pts− in the interfacial headgroup region. Although the chemical-shift changes cannot define a unique binding geometry, together with the CG-MD coordination analysis they support a NaPts-involved interfacial organization in which Pts− preferentially associates with cationic regions.

To examine this interpretation further, CG-MD simulations were carried out for representative CO2-treated formulations.

Coarse-grained simulations: multiscale evolution, interfacial properties, and ordered organization

CG-MD simulations were used to follow aggregate evolution and headgroup/counterion organization in representative models of the CO2-treated system. The relative proportions of EAHSB, EKO, EKOH+ and NaPts were kept constant, while the total concentration was varied. The simulations enable us to follow the time-resolved evolution of surfactant aggregates, the interfacial exposure of hydrophobic and hydrophilic moieties, and the spatial organization of polar headgroups within the final micellar structures.

Assembly kinetics

As suggested by the cryo-TEM image of the EAHSB/EKO/NaPts-CO2 solution (Fig. 2f), the mixed system exhibits the coexistence of multiple micellar morphologies with planar self-assembly characteristics, including vesicles, ring-like features, and cylindrical aggregates. The end-of-simulation snapshots show vesicle-, lamella- and cylinder-like features, which are qualitatively consistent with the cryo-TEM observations.

The time evolution of aggregate formation is illustrated in Fig. 4b and c. At early times, small spherical or short cylindrical micelles nucleate from randomly dispersed surfactant molecules. These aggregates then merge into larger bilayer-like patches, which can curve into vesicles or grow into thicker cylindrical structures. Because the CG models use elevated concentrations, these trajectories should be viewed as qualitative assembly pathways rather than direct kinetic reproductions of the experiments. CO2-induced protonation-regulated self-assembly.

Fig. 4. Snapshots of equilibrium configurations for different systems (a); snapshots of micelle formation processes for System 1 (b) and System 2 (c) at different time stages.

Fig. 4

Open in a new tab

Micelle interfaces

To further characterize the self-assembled architectures obtained from CG-MD, we selected a representative aggregate formed in System 1 and calculated the solvent-accessible surface area (SASA) as a function of simulation time, which was further decomposed into contributions from hydrophilic and hydrophobic beads (Fig. 5a and b). The total SASA gradually decreases as individual surfactants pack into more compact aggregates, reflecting a reduction in exposed surface area. Decomposition of SASA into hydrophobic and hydrophilic contributions reveals that, for the mixed system, approximately 24.03% of the accessible surface is associated with hydrophobic tail beads, while 75.97% arises from hydrophilic headgroups and counterions. The total SASA decreases as dispersed surfactants pack into aggregates, indicating progressive compaction. In the representative mixed aggregate, hydrophilic headgroups and counterions account for most of the accessible surface area, while hydrophobic tail beads still contribute about 24.03%. This mixed surface character suggests a solvated but heterogeneous interface, which may facilitate contacts between neighboring aggregates.

Fig. 5. Change in SASA over time for hydrophilic and hydrophobic groups during the simulation (a); distribution of hydrophilic (red) and hydrophobic (blue) zones on the surface of a representative self-assembled aggregate (b).

Fig. 5

Open in a new tab

Organization of polar headgroups

In the present system, CO2 mainly functions through pH-mediated protonation of the tertiary amine group in EKO. Therefore, the observed assembly behavior is more appropriately described as CO2-induced protonation-regulated self-assembly rather than CO2-driven self-assembly. During the CO2-induced protonation-regulated self-assembly process, the polar headgroups of EAHSB, EKOH+ and EKO, together with Pts−, redistribute within the interfacial region. Radial distribution functions (RDFs) between representative functional groups (Fig. 6) provide a quantitative view of these rearrangements. Pronounced first peaks are observed for unlike-charged pairs—most notably between sulfonate groups and cationic sites such as the protonated amine in EKOH+—indicating strong short-range electrostatic association. In addition, Pts− displays prominent near-contact correlations with the cationic regions, consistent with its role as a preferential counterion that accumulates in the interfacial headgroup layer.

Fig. 6. RDFs between different functional groups: negative and negative groups (a); positive and positive groups (b); negative and positive groups (c); tertiary amine of EKO and charged groups (d).

Fig. 6

Open in a new tab

To quantify the strength of these short-range associations, coordination numbers (CNs) were obtained by integrating the RDF, g(r), from r = 0 to the first minimum beyond the first RDF peak (Table 2), which corresponds to the first coordination shell. The resulting CN matrix highlights several dominant motifs. EAHSB exhibits strong local ion pairing between its sulfonate and quaternary ammonium sites (CN = 1.90), consistent with efficient internal/near-neighbour charge compensation within the zwitterionic headgroup layer. Pts− coordinates strongly around EKOH+ (CN = 1.70) and also around the cationic moiety of EAHSB (CN = 1.49), supporting preferential association of Pts− with cationic centres. A direct hetero-association between EAHSB sulfonate and EKOH+ is also evident (CN = 0.60), which is consistent with cooperative headgroup packing via complementary charges. By contrast, like-charged contacts remain weak (e.g., EAHSB+–EKOH+ CN = 0.25; EKOH+–EKOH+ CN = 0.24), indicating that counterion-mediated screening suppresses short-range cation–cation crowding.

Table 2. CN between different functional groups.
CN Negative groups Cationic groups Neutral groups
–SO3− (Pts−) –SO3− (EAHSB) –N(CH3)2+–(EAHSB) –NH(CH3)2+ (EKOH+) –N(CH3)2 (EKO)
Negative groups –SO3− (Pts−) 0.44 0.31 1.49 1.70 1.16
–SO3− (EAHSB) 0.51 1.90 0.60 0.33
Cationic groups –N(CH3)2+–(EAHSB) 0.45 0.25 0.43
–NH(CH3)2+ (EKOH+) 0.24 0.44
Neutral groups –N(CH3)2 (EKO) 0.59
Open in a new tab

Importantly, the simultaneously high coordination of Pts− around both EAHSB+ and EKOH+ suggests that Pts− preferentially localizes within mixed cation-rich interfacial regions. This behavior is consistent with transient bridge-like association motifs rather than with a uniquely defined bridging structure. Such ion-mediated short-range associations may help screen residual headgroup repulsion and favor closer local packing in the interfacial region. When considered together with the 1H NMR perturbations and the cryo-EM/rheological results, these simulations provide a plausible molecular interpretation for the richer multiscale assembly and stronger viscoelastic response observed in the mixed system. We next examine the relative position of hydrophobic tail segments near the headgroup region, which may further affect interfacial packing.

Packing state of hydrophobic tail chains

The spatial arrangement of hydrophobic tails relative to the interfacial headgroup layer in the EAHSB/EKO/NaPts-CO2 system was analysed by calculating RDFs between the coarse-grained headgroup bead N and two representative tail beads C4 and C6 along the hydrocarbon chains of EAHSB, EKO, and EKOH+ (Fig. 7). Here, C4 represents a tail segment closer to the headgroup, whereas C6 is located further down the hydrophobic chain; therefore, comparing the N–C4 and N–C6 correlations provides insight into how different parts of the tail populate the headgroup–tail interfacial region.

Fig. 7. RDFs between the surfactant headgroup bead (N) and tail chain beads (C4 and C6).

Fig. 7

Open in a new tab

For all three surfactant species, the N–C4 RDF exhibits a pronounced first peak at short distances, indicating that the near-headgroup tail segments are frequently located in close proximity to the headgroup region. This is consistent with the expected anchoring of the hydrophobic chain near the interfacial shell in mixed micelles/bilayer fragments. The N–C6 RDF also shows short-to-intermediate distance correlations, indicating that some distal tail segments remain close to the headgroup layer instead of being fully buried in a uniform hydrophobic core. Such behavior is consistent with partial tail backfolding and/or interdigitation near the interface in the mixed assembly, leading to a more heterogeneous interfacial zone in which hydrophobic segments partially penetrate toward the headgroup-rich region.45

This arrangement may create additional hydrophobic contact points near the interface and facilitate association between neighboring aggregates. In conjunction with the NaPts-associated reorganization of the mixed headgroup region discussed above, this coupled headgroup–tail organization offers a plausible explanation for why thicker cylindrical and ring-like architectures are more readily formed in the mixed system.

Conclusion

In this work, rheology, cryo-EM, 1H NMR and CG-MD simulations were used to examine the origin of the enhanced viscoelasticity of the CO2-treated EAHSB/EKO/NaPts formulation. Compared with the EAHSB/NaPts and EKO/NaPts-CO2 binary systems, the ternary formulation shows higher viscosity and elasticity, together with more varied aggregate morphologies including WLMs, vesicles, ring-like features and thick cylindrical structures. 1H NMR and CG-MD results suggest that CO2-induced protonation, mixed-headgroup interactions and Pts− localization near cationic regions contribute to NaPts-involved interfacial organization. These findings explain the cooperative viscoelastic enhancement of the ternary system and provide guidance for designing CO2-switchable amine/zwitterionic surfactant fluids.

Conflicts of interest

There are no conflicts to declare.

Supplementary Material

RA-016-D6RA02512A-s001

Acknowledgments

We gratefully acknowledge the financial support from the Major Scientific and Technological Innovation Projects in Shandong Province (No. 2025ZLGX07), the National Natural Science Foundation of China (No. 52304061), the Postdoctoral Fellowship Program of CPSF (No. GZB20240851), the Natural Science Foundation of Shandong Province of China (No. ZR2024QE088), the Qingdao Postdoctoral Funding Project (No. QDBSH20240201005), and the Fundamental Research Funds for the Central Universities (No. 24CX06040A).

Data availability

All data supporting this study are included in the article and its supplementary information (SI). Supplementary information: coarse-grained mapping schemes and simulation details, additional 1H NMR and rheological reference data, binary-system schematic packing models, and equations used for extracting comparative micellar network parameters. See DOI: https://doi.org/10.1039/d6ra02512a.

References

  1. Zhao J. Zhang Y. Bie S. Bilsback K. R. Pierce J. R. Chen Y. Changes in global DMS production driven by increased CO2 levels and its impact on radiative forcing. npj Clim. Atmos. Sci. 2024;7:18. doi: 10.1038/s41612-024-00563-y. [DOI] [Google Scholar]
  2. Namikawa Y. Suzuki M. Atmospheric CO2 Sequestration in Seawater Enhanced by Molluscan Shell Powders. Environ. Sci. Technol. 2024;58:2404–2412. doi: 10.1021/acs.est.3c09273. [DOI] [PubMed] [Google Scholar]
  3. Peter R. Kuttippurath J. Sunanda N. Chakraborty K. Effect of thermal and non-thermal processes on the variability of ocean surface pCO2 and buffering capacity in the north Indian Ocean. Prog. Oceanogr. 2025;233:103442. doi: 10.1016/j.pocean.2025.103442. [DOI] [Google Scholar]
  4. Bourtsalas A. C. Goldberg D. Zorn T. Techno-economic and environmental assessment of carbon capture and storage in northeastern U.S. industrial clusters. Int. J. Greenhouse Gas Control. 2025;148:104499. doi: 10.1016/j.ijggc.2025.104499. [DOI] [Google Scholar]
  5. Agbejule A. Sempron-Namuag P. A bibliometric analysis of carbon capture, utilization and storage (CCUS): Identifying barriers and drivers. Appl. Energy. 2025;400:126604. doi: 10.1016/j.apenergy.2025.126604. [DOI] [Google Scholar]
  6. Ren Q. Wei S. Du J. Wu P. Research progress and perspectives on carbon capture, utilization, and storage (CCUS) technologies in China and the USA: a bibliometric analysis. Environ. Sci. Pollut. Res. 2023;30:76437–76454. doi: 10.1007/s11356-023-27749-w. [DOI] [PubMed] [Google Scholar]
  7. Wang W. Wei N. Wei F. Ali M. Lin K. Lei S. Research on the progress of Carbon Capture, Utilization, and Storage (CCUS) technologies based on bibliometric analysis. J. Cleaner Prod. 2025;529:146826. doi: 10.1016/j.jclepro.2025.146826. [DOI] [Google Scholar]
  8. Yao J. Han H. Yang Y. Song Y. Li G. A Review of Recent Progress of Carbon Capture, Utilization, and Storage (CCUS) in China. Appl. Sci. 2023;13:1169. doi: 10.3390/app13021169. [DOI] [Google Scholar]
  9. Liu B. Yao J. Sun H. Zhang L. Revealing the effects of thermal properties of supercritical CO2 on proppant migration in supercritical CO2 fracturing. Gas Sci. Eng. 2024;121:205172. doi: 10.1016/j.jgsce.2023.205172. [DOI] [Google Scholar]
  10. Cai C. Kang Y. Wang X. Hu Y. Huang M. Liu Y. Liu J. Chen H. Li X. Experimental study on shale fracturing enhancement by using multi-times pulse supercritical carbon dioxide (SC-CO2) jet. J. Pet. Sci. Eng. 2019;178:948–963. doi: 10.1016/j.petrol.2019.04.009. [DOI] [Google Scholar]
  11. Oh J. Kim K.-Y. Han W. S. Kim T. Kim J.-C. Park E. Experimental and numerical study on supercritical CO2/brine transport in a fractured rock: Implications of mass transfer, capillary pressure and storage capacity. Adv. Water Resour. 2013;62:442–453. doi: 10.1016/j.advwatres.2013.03.007. [DOI] [Google Scholar]
  12. Chen H. Hu Y. Kang Y. Wang X. Liu F. Liu Y. Advantages of supercritical CO2 compound fracturing in shale on fracture geometry, complexity and width. J. Nat. Gas Sci. Eng. 2021;93:104033. doi: 10.1016/j.jngse.2021.104033. [DOI] [Google Scholar]
  13. Liu X. Jia X. Niu Y. Nie B. Zhang C. Song D. Alterations in coal mechanical properties and permeability influenced by liquid CO2 phase change fracturing. Fuel. 2023;354:129254. doi: 10.1016/j.fuel.2023.129254. [DOI] [Google Scholar]
  14. Lei X. Liu B. Hou Q. Wang Y. Ahmadi M. Liu Z. Chen Z. Switchability and synergistic effect of a CO2-responsive surfactant with co-surfactants at an O/W interface: A molecular insight. J. Mol. Liq. 2024;405:125051. doi: 10.1016/j.molliq.2024.125051. [DOI] [Google Scholar]
  15. Zhang H. Gong H. He X. Wu J. Sun H. Xu L. Dong M. Simultaneous adjustment on the CO2/oil and CO2/water interface tensions using hydrophilic, lipophilic and CO2-philic surfactants during CO2 flooding to enhance shale oil recovery after hydraulic fracturing. Chem. Eng. J. 2024;486:150244. doi: 10.1016/j.cej.2024.150244. [DOI] [Google Scholar]
  16. Xu T. Mao J. Yang X. Zhang Y. Sun Y. Lin C. Zhang Q. Lu Q. Effect of the number of hydroxyl groups of CO2-triggered surfactants on capability and performance in CO2-stimulated response. Colloids Surf., A. 2023;658:130761. doi: 10.1016/j.colsurfa.2022.130761. [DOI] [Google Scholar]
  17. Dowlati S. Mokhtari R. Hohl L. Miller R. Kraume M. Advances in CO2-switchable surfactants towards the fabrication and application of responsive colloids. Adv. Colloid Interface Sci. 2023;315:102907. doi: 10.1016/j.cis.2023.102907. [DOI] [PubMed] [Google Scholar]
  18. Fu L. Zhang Z. Qiu X. Zheng Y. Shao M. Bai Y. Ren Z. Si W. Jin J. Xu Y. Liao K. A rapidly switchable CO2-responsive Pickering emulsions based on supramolecular amphiphilic surfactants and nano-SiO2. Colloids Surf., A. 2023;676:132275. doi: 10.1016/j.colsurfa.2023.132275. [DOI] [Google Scholar]
  19. Zhang Y. Gao M. You Q. Fan H. Li W. Liu Y. Fang J. Zhao G. Jin Z. Dai C. Smart mobility control agent for enhanced oil recovery during CO2 flooding in ultra-low permeability reservoirs. Fuel. 2019;241:442–450. doi: 10.1016/j.fuel.2018.12.069. [DOI] [Google Scholar]
  20. Shen H. Yang Z. Li X. Peng Y. Lin M. Zhang J. Dong Z. CO2-responsive agent for restraining gas channeling during CO2 flooding in low permeability reservoirs. Fuel. 2021;292:120306. doi: 10.1016/j.fuel.2021.120306. [DOI] [Google Scholar]
  21. Wu X. Dai C. Huang Y. Gu C. He H. Zhao M. Developing New Recyclable and CO2 Sensitive Amphiphile for Fracturing Fluid. IOP Conf. Ser. Earth Environ. Sci. 2018;153:022037. doi: 10.1088/1755-1315/153/2/022037. [DOI] [Google Scholar]
  22. Wu X. Huang Y. Fang S. Dai C. Li H. Xu Z. Zhao M. CO2-responsive smart wormlike micelles based on monomer and “pseudo” gemini surfactant. J. Ind. Eng. Chem. 2018;60:348–354. doi: 10.1016/j.jiec.2017.11.021. [DOI] [Google Scholar]
  23. Xiong C. Wei F. Zhou Q. Peng K. Ye Z. Yang H. A CO2-responsive smart fluid based on supramolecular assembly structures varying reversibly from vesicles to wormlike micelles. RSC Adv. 2020;10:25311–25318. doi: 10.1039/D0RA03854G. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Wang R. Qi C. Zhao H. Wang L. Li Q. Wang L. Loosely bonded dual-functionalized ionic liquid-based phase change solvent for energy-saving CO2 capture. Fuel. 2023;350:128748. doi: 10.1016/j.fuel.2023.128748. [DOI] [Google Scholar]
  25. Lin X. Lu H. Zhang R. Chen Y. Huang B. Dual-functionality of CO2-responsive ionic liquids for CO2 storage: Fatty acid particles re-precipitation and clay swelling inhibition. Chem. Eng. Res. Des. 2025;217:64–71. doi: 10.1016/j.cherd.2025.03.024. [DOI] [Google Scholar]
  26. Wei X. He X. Zhang D. Su X. CO2-Responsive Wormlike Micelles Based on Pseudo-Tetrameric Surfactant. Molecules. 2022;27:7922. doi: 10.3390/molecules27227922. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Zhang P. Zhao L. Lai L. Study on the oil displacement mechanism of anionic/zwitterionic surfactant mixtures in low-permeability reservoirs. Colloids Surf., A. 2026;728:138614. doi: 10.1016/j.colsurfa.2025.138614. [DOI] [Google Scholar]
  28. Baruah A. Pathak A. K. Ojha K. Study on rheology and thermal stability of mixed (nonionic–anionic) surfactant based fracturing fluids. AIChE J. 2016;62:2177–2187. doi: 10.1002/aic.15175. [DOI] [Google Scholar]
  29. Li B. Zhang Q. Xia Y. Gao Z. Surface properties and aggregation behavior of cationic gemini surfactants with dipropylammonium head-groups. Colloids Surf., A. 2015;470:211–217. doi: 10.1016/j.colsurfa.2015.02.005. [DOI] [Google Scholar]
  30. Zhang W. Mao J. Yang X. Zhang H. Zhao J. Tian J. Lin C. Mao J. Development of a sulfonic gemini zwitterionic viscoelastic surfactant with high salt tolerance for seawater-based clean fracturing fluid. Chem. Eng. Sci. 2019;207:688–701. doi: 10.1016/j.ces.2019.06.061. [DOI] [Google Scholar]
  31. Sun N. Gao M. Liu J. Zhao G. Ding F. You Q. Dai C. A novel temperature-resistant fracturing fluid for tight oil reservoirs: CO2-responsive clean fracturing fluid. Colloids Surf., A. 2023;665:131247. doi: 10.1016/j.colsurfa.2023.131247. [DOI] [Google Scholar]
  32. Gao M.-W. Zhang M.-S. Du H.-Y. Zhao M.-W. Dai C.-L. You Q. Liu S. Jin Z.-H. A novel triple responsive smart fluid for tight oil fracturing-oil expulsion integration. Pet. Sci. 2023;20:982–992. doi: 10.1016/j.petsci.2023.01.008. [DOI] [Google Scholar]
  33. Abraham M. J. Murtola T. Schulz R. Páll S. Smith J. C. Hess B. Lindahl E. GROMACS: High performance molecular simulations through multi-level parallelism from laptops to supercomputers. SoftwareX. 2015;1–2:19–25. doi: 10.1016/j.softx.2015.06.001. [DOI] [Google Scholar]
  34. Berendsen H. J. C. Postma J. P. M. van Gunsteren W. F. DiNola A. Haak J. R. Molecular dynamics with coupling to an external bath. J. Chem. Phys. 1984;81:3684–3690. doi: 10.1063/1.448118. [DOI] [Google Scholar]
  35. Bussi G. Donadio D. Parrinello M. Canonical sampling through velocity rescaling. J. Chem. Phys. 2007;126:014101. doi: 10.1063/1.2408420. [DOI] [PubMed] [Google Scholar]
  36. Parrinello M. Rahman A. Polymorphic transitions in single crystals: A new molecular dynamics method. J. Appl. Phys. 1981;52:7182–7190. doi: 10.1063/1.328693. [DOI] [Google Scholar]
  37. Humphrey W. Dalke A. Schulten K. VMD: visual molecular dynamics. J. Mol. Graph. 1996;14(33–38):27–38. doi: 10.1016/0263-7855(96)00018-5. [DOI] [PubMed] [Google Scholar]
  38. Souza P. C. T. Alessandri R. Barnoud J. Thallmair S. Faustino I. Grünewald F. Patmanidis I. Abdizadeh H. Bruininks B. M. H. Wassenaar T. A. Kroon P. C. Melcr J. Nieto V. Corradi V. Khan H. M. Domański J. Javanainen M. Martinez-Seara H. Reuter N. Best R. B. Vattulainen I. Monticelli L. Periole X. Tieleman D. P. de Vries A. H. Marrink S. J. Martini 3: a general purpose force field for coarse-grained molecular dynamics. Nat. Methods. 2021;18:382–388. doi: 10.1038/s41592-021-01098-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Wallace E. J. Sansom M. S. P. Carbon Nanotube/Detergent Interactions via Coarse-Grained Molecular Dynamics. Nano Lett. 2007;7:1923–1928. doi: 10.1021/nl070602h. [DOI] [PubMed] [Google Scholar]
  40. Zhao G. Yan Z. Qian F. Sun H. Lu X. Fan H. Molecular simulation study on the rheological properties of a pH-responsive clean fracturing fluid system. Fuel. 2019;253:677–684. doi: 10.1016/j.fuel.2019.05.027. [DOI] [Google Scholar]
  41. Martínez L. Andrade R. Birgin E. G. Martínez J. M. PACKMOL: A package for building initial configurations for molecular dynamics simulations. J. Comput. Chem. 2009;30:2157–2164. doi: 10.1002/jcc.21224. [DOI] [PubMed] [Google Scholar]
  42. de Souza R. N. Jora M. Z. Duarte L. G. T. A. Clinckspoor K. J. Atvars T. D. Z. Sabadini E. A new interpretation of the mechanism of wormlike micelle formation involving a cationic surfactant and salicylate. J. Colloid Interface Sci. 2019;552:794–800. doi: 10.1016/j.jcis.2019.05.025. [DOI] [PubMed] [Google Scholar]
  43. Bi Y. Wang T. Xiao J. Yu L. Wormlike Micelles with Photo and pH Dual-Stimuli-Responsive Behaviors Formed by Aqueous Mixture of Cationic and Anionic Surfactants. Colloids Surf., A. 2023;668:131441. doi: 10.1016/j.colsurfa.2023.131441. [DOI] [Google Scholar]
  44. Liu P. Pei X. Li C. Li R. Chen Z. Song B. Cui Z. Xie D. pH-switchable wormlike micelles with high viscoelasticity formed by pseudo-oligomeric surfactants. J. Mol. Liq. 2021;334:116499. doi: 10.1016/j.molliq.2021.116499. [DOI] [Google Scholar]
  45. Gokhale D. Chen I. Doyle P. S. Coarse-grained molecular dynamics simulations of immobilized micelle systems and their interactions with hydrophobic molecules. Soft Mater. 2022;18:4625–4637. doi: 10.1039/D2SM00280A. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

RA-016-D6RA02512A-s001
RA-016-D6RA02512A-s001.pdf (1.1MB, pdf)

Data Availability Statement

All data supporting this study are included in the article and its supplementary information (SI). Supplementary information: coarse-grained mapping schemes and simulation details, additional 1H NMR and rheological reference data, binary-system schematic packing models, and equations used for extracting comparative micellar network parameters. See DOI: https://doi.org/10.1039/d6ra02512a.

Articles from RSC Advances are provided here courtesy of Royal Society of Chemistry

RESOURCES