Introduction

The elevated concentration of carbon dioxide (CO2) in the atmosphere due to anthropogenic or human activities is perceived to be a major contributor to global warming1. Over 50% of the global energy-related CO2 emissions are from the combustion of oil and gas and fossil fuels2. Under the current energy policies scenario (especially in developing countries) and with increasing global energy demand, the emission of over 25 Gt of CO2 per year from fossil fuels is considered inevitable3. Investments in renewable and alternative energies and CO2 capture and storage (CCS) in subsurface formations have been of interest to mitigate the greenhouse gas effects4. However, social issues surrounding CCS projects in reservoirs close to residential areas imply transportation to remote areas, which is not economically favorable5. Moreover, despite dramatic growth in the renewable energy sector, around 80% of global energy demand is provided from fossil fuel resources, which highlights the importance of positioning the oil and gas industry as a part of the solution towards negative carbon emission6,7. Miscible and immiscible CO2 injection in oil reservoirs has been considered an attractive enhanced oil recovery (EOR) technique for over 50 years due to CO2 physical properties and its multiple interactions with oil over a wide pressure and temperature range, high CO2 dissolution with oil, lowering oil/CO2 interfacial tension (IFT), oil swelling and reducing oil viscosity8,9. The incremental oil produced from CO2-EOR is considered a lower-carbon fuel and the process is identified as an early opportunity for CCS to store CO2 through EOR operation, depending on the amount of CO2 used per barrel of oil recovered3. CO2 utilization factor (mass or volume of CO2 required to produce one incremental barrel of oil) is the most important parameter in the life-cycle of CO2-EOR to enable the CCS and Utilization (CCUS) as viable options for operators7. The CO2 utilization factor depends hugely on the reservoir heterogeneity and rises from 0.3 tCO2/bbl in conventional CO2-EOR operations to 0.9 by applying advanced CO2 conformance and mobility control methods10.

CO2-Foam can alleviate reservoir heterogeneity and mitigate CO2 conformance and mobility contrast issues such as fingering, gravity segregation, and early breakthrough in production well11. However, the need for copious amounts of water for foam to propagate, foam dry-out in the absence of water below critical water saturation, detrimental effect of oil on foam at oil saturations above the critical so-called foaming saturation and surfactant loss due to adsorption onto the rock surface or partitioning at oil/water interface are some limitations of conventional foaming agents12, that make the newly developed CO2-philic foaming surfactants attractive for CO2-EOR and CCUS applications12. High CO2 solubility in CO2-philic surfactant solution also makes it an effective solution for greenhouse mitigation in the CCUS context13, 14.

In our previous studies on newly developed CO2-philic surfactants, we have developed a comprehensive morphology screening on sulfonic, anionic-amphoteric-based surfactant blends with CO2-philic functionalities for CO2 foam-EOR, concluding that the addition of tail and balanced branches in surfactant structure increased CO2-foam performance in terms of stability, mobility reduction factor (MRF), IFT reduction and oil recovery at target high-temperature sandstone reservoir12,15,16. In terms of static adsorption tests on sandstone, the newly developed surfactants performed better than the commercial alpha olefin sulfonate (AOS) surfactant, as reported in our previous publication17. In terms of CO2-philic foam propagation in porous media, surfactant alternating gas (SAG) injection tests in the presence of oil were performed at reservoir conditions (102 ֯C and 125 bar) with the selected single, double and three-tailed surfactants12. The surfactants with higher activity at CO2/aqueous interfaces (D and T) showed better performance in terms of recovery factor (RF%) (95% and 98%, respectively).

The only limiting factor for increasing the number of tails in surfactant structure was observed to be the increased equilibrium adsorption of surfactant onto the sandstone rock surface, upon comparison of single-, double- and triple-tailed surfactants15. Higher adsorption of triple-tailed surfactant onto the rock surface was reported to be due to the longer tail and bulkier structure of this surfactant, which is in line with the previous findings on the increase in the magnitude of adsorption with the increase in length of the hydrocarbon chain18.

Surfactant adsorption is a transfer of surfactant molecules from the bulk phase to the surface/interface18–20. The adsorption process ultimately lowers the surfactant concentration in the injected material and causes the retardation of the surfactant front, which has adverse technical and financial effects21–24. Therefore, the loss of surfactants by adsorption is considered a major factor in the project economy25,26. Adsorption of surfactant at the liquid-solid interface particularly is a very complex process in nature, which is the combination of many forces such as chemical, electrostatic, hydrogen bonding, lateral chain-chain interactions, and de-solvation of previously adsorbed surfactant molecules27,28.. It can be a function of the type of surfactant and the nature of surfactant polar head groups and tail parts29.

The depletion method is utilized for surfactant adsorption measurement, by which the change in concentration is measured30. The effect of alkali on the surfactant adsorption process is shown to be very beneficial31. The mechanism of lowering adsorption concentration involves the available positive site reduction on the reservoir rock surface that takes place primarily by increasing solution pH. Although alkali such as sodium carbonate can elevate the solution pH and generate in-situ surfactant molecules through reaction with acidic components present in crude oil, it may precipitate in the presence of divalent cations and result in the reduction of rock permeability profile26,31. These issues may be overcome by using sodium tetraborate19. By generating coordination complexes, it reduces divalent cations precipitation and is more stable in a high salinity environment.

The focus of the present paper is understanding and predicting the adsorption mechanism of newly developed anionic CO2-philic surfactants through experimental and modeling approaches. Adsorption measurement of three types of newly-developed CO2-philic surfactants was performed by observing the changes in concentration, static adsorption tests, and point of zero charge measurements. The adsorption behavior of surfactants is also extensively evaluated by performing modeling and kinetics studies along with investigation on the effect of alkali on the adsorption process.

Materials and methods

Three newly developed surfactants were synthesized in our laboratory and used to determine the adsorption isotherms. The properties of the foaming agent surfactants are reported in our previous publication32. The surfactants contain different chain lengths, branches, and a sulfonate head group. The surfactants used in this project are named S, D, and T and their structures are illustrated in Fig. 1.

Fig. 1
figure 1

Surfactant structure (a) Single-tail, (b) Double-tailed, (c) Three-tailed.

Full size image

The Berea sandstone core, purchased from Cleveland Quarries (USA), is utilized as an adsorbent rock. The adsorbent was sieved to get 60–70 mesh-sized particles and several times washed with distilled water. The adsorbent was used for further adsorption experiments after cleaning the residual adsorbent particles and dried at 363 K (90 °C) for 48 h. Surfactant adsorption concentration was observed by using a Spectrophotometer (Shimadzu UV-3150). The pH of the samples was measured by using the Metrohm Titrando. The X-ray Photoelectron Spectroscopy (XPS) equipment was utilized to determine the elemental composition (wt%) of the core samples. The samples were analyzed by using the Leybold spectrometer using Al K α radiation (1486.6 eV). The instrument was run at the fixed analyzer transmission mode at pass energy of 50 eV with a 300 W X-ray power source. The samples were mounted onto the indium holder (by pressing) and then introduced in a pre-chamber under a vacuum of 10 − 5 torr. All binding energies were referenced to C 1s at 285 eV. All of the peaks (C 1s, N 1s, O 1s, and P 2p) were fitted into several components having the same full width at the half maximum (FWHM). The Fourier Transform Infrared (FTIR) spectra of the adsorbent before and after surfactant treatment were measured by using a NICOLE FTIR spectrophotometer in the range of 450–4000 cm−1.

One of the important properties of surfactants is the critical micelle concentration (CMC). The CMC of the synthesized surfactant solutions was determined by measuring the surface tension of different concentrations against air, by using a DATA PHYSICS S20 spinning drop tensiometer instrument under atmospheric pressure. The surfactant solution was placed in a capillary tube, which can rotate inside a measuring cell. An air bubble was introduced inside the capillary by using a fine needle syringe. The rotation speed increased until the bubble changed to a length that is 4 times the width of the drop. The surface tension was calculated through a high-speed camera, focused at the elongated drop. The CMCs were obtained from a plot of surface tension against surfactant concentration. The surface tension of surfactants at CO2/brine (3% NaCl) at a range of pressure (400 to 2500 psi) and temperature of 90 ֯C was measured by using a Vinci-Technologies IFT700 pendant drop tensiometer, equipped with axisymmetric drop shape analysis. The setup diagram and experimental procedure can be found elsewhere33–35.

An important parameter for adsorption research, the point of zero charge (PZC) of the adsorbent sample was calculated by using the salt addition process36. A non-destructive analytical technique to determine the elemental composition of materials by measuring the fluorescent X-ray emitted from a sample when it is excited by a primary X-ray source. Static surfactant adsorption was carried out by taking an amount of 1 g of core sample mixed in 20 mL solution of surfactant, which contains alkali and salinity in specific quantities, and agitated for the next 24 h at 90 oC using a horizontal shaker. After that the samples were centrifuged and the supernatant was filtered by a membrane filter of 0.45-micron size. The effect of alkali was also measured using the same method, where the addition of alkali changed the pH of the test from 6.2 to 9.8.

The concentration adsorbed (mg/g) on the rock surface was calculated by measuring the change in the surfactant concentration in the solution by using Eq. 137,38;

$$\:Ads=({C}_{i}-{C}_{e})\times\:\frac{V}{M}$$
(1)

where Ci and Ce are the initial and final surfactant concentration in ppm, V is the volume of solution and M is the mass of core in gram, g. The effects of various parameters such as temperature, pH, and NaCl concentrations on the adsorption were also investigated. The kinetics and thermodynamic studies were also conducted.

Adsorption isotherm models

Typically, adsorption has three types; (a) Chemical (b) Physical (c) Ion Exchange type of Adsorption. Two adsorption isotherm models namely, Freundlich and Langmuir are the most widely used models for the adsorption study22.

Freundlich adsorption isotherm model

Freundlich adsorption isotherm used in literature explaining reversible and non-ideal adsorption22,39,40. The model relates the solute adsorbed amount to the solute amount in liquid. Freundlich Adsorption Isotherm is mathematically expressed as41–43

$$\:\frac{x}{m}=k{c}^{\frac{1}{n}}$$
(2)

It can also be written as:

$$\:{log}\frac{x}{m}={log}k+\frac{1}{n}{log}c$$
(3)

where, x = adsorbate amount (mg), m = adsorbent amount (g), c = Equilibrium adsorbate concentration. K and n are constants of adsorbate and adsorbent.

Freundlich equation is true for real systems at lower concentrations when only electrostatic interactions are active and it fits well to a system where adsorption (mg/g) is assumed proportional to concentration. The adsorbent surface becomes electrically neutralized over time, and electrostatic interactions do not participate in adsorption, so at saturation point, adsorption is independent of initial concentration. The slope of the linear equation of the Freundlich model ranges between 0 and 1, which explains adsorption intensity or surface heterogeneity. A closer value to zero defines a more heterogeneous surface. On the other hand, the below unity value represents a chemisorption system, where 1/n indicates cooperative adsorption. 1/n = 0 explains a completely heterogeneous surface. The equation is modified for surfactant adsorption onto sandstones as22,39–41, 44;

$$\:L{og}({q}_{e})=L{og}({k}_{f})+\left(\frac{1}{n}\right)L{og}({c}_{e})$$
(4)

where ce = Concentration of adsorbate at equilibrium, qe = Adsorption amount of adsorbate (mg/g), kf = Freundlich constant, n = Constant of surface heterogeneity. The extent of adsorption varies directly with concentration until it reaches equilibrium concentration ce. At the equilibrium point, the surfactant solution reaches its critical micelle concentration (CMC) and beyond that point, there is no change in adsorption capacity as the concentration induces micellization. The Freundlich adsorption isotherm therefore fails at higher concentrations.

Langmuir adsorption isotherm model

The Langmuir adsorption isotherm is the most frequently used model for adsorption at the solid-liquid interface. The Langmuir equation relates the amount of solid adsorbate adsorbed to the equilibrium liquid concentration at a fixed temperature. The equation was developed by Irving Langmuir in 1916, and the nonlinear form of it is stated as41:

$$\:{q}_{e}=\frac{{q}_{o}b{c}_{e}}{1+b{c}_{e}}$$
(5)

Langmuir isotherm’s linear form is given below [22, 39–42, 44].

$$\:\frac{1}{{q}_{e}}=\frac{1}{{q}_{max}}+\frac{1}{b{q}_{max}{c}_{e}}$$
(6)

where ce = Concentration of surfactant, qe = Adsorption (mg/g), qmax = Maximum adsorption (mg/g), b = Adsorption constant related to adsorption energy.

Kinetics and thermodynamic studies

A 20 g of adsorbent was transferred into 100 ml of surfactant solution portions at different concentrations of 400 ppm, 800 ppm, and 1000 ppm, respectively. The adsorption measurements were performed after constant time interval (10 min) until the total adsorption equilibrium was achieved. Thermodynamic studies were performed at three different temperatures of 30, 40, and 50 °C. The samples were centrifuged after 6 h and the concentrations of surfactants were measured.

Prediction of surfactant adsorption is very important for the implementation of adsorption design45. Chemical kinetics is an efficient approach to study reaction rates and the factors affecting the rate of a reaction. The rate constant is a parameter to evaluate some basic quantities for chemical flooding. Kinetic data is concerned with the rates of change involved during a process while equilibrium data only provides information about the system’s final state30. Pseudo-first-order and pseudo-second-order models were used for adsorption kinetics measurements of anionic surfactants40, 45–47.

Pseudo first-order kinetic model

The pseudo-first-order kinetic model can be written as42, 44,48,49

$$\:\frac{d{q}_{t}}{dt}={k}_{f}({q}_{e}-{q}_{t})$$
(7)

where qt is the amount of surfactant adsorbed (mg/g) at time t, qe is the amount of surfactant adsorbed (mg/g) at equilibrium, kf is the rate constant for pseudo-first-order model (1/min), and t is time (min). After integration and applying initial conditions i.e. qt = 0 at t = 0 and qt = qt at t = t, the equation becomes48:

$$\:L{og}({q}_{e}-{q}_{t})=L{{og}q}_{e}-{k}_{f}\frac{t}{2.303}$$
(8)

The rate constant for the adsorption of surfactants on Berea sandstone was evaluated from the slope and intercept of the graph between log (qt - qe) vs. t.

Pseudo second-order kinetic model

The mathematical form of Pseudo-second-order can be presented in the following form42,44,48,49;

$$\:\frac{d{q}_{e}}{dt}={k}_{s}({q}_{e}-{q}_{t}{)}^{2}$$
(9)

where ks= pseudo-second-order rate constant (1/min). Applying definite integrals and boundary conditions qt = 0 at t = 0 and qt = qt at t = t, this equation is reduced to the following form48;

$$\:\frac{\varvec{t}}{{\varvec{q}}_{\varvec{t}}}=\frac{1}{{\varvec{q}}_{\varvec{e}}}\varvec{t}+\frac{1}{{\varvec{k}}_{\varvec{s}}{\varvec{q}}_{\varvec{e}}^{2}}$$
(10)

The initial adsorption rate constant “h” (mg/g. min) at t = 0 can be defined as;

$$\:h={k}_{s}{q}_{e}^{2}$$
(11)

Pseudo-second order rate constant ks (min−1) can be determined by a slope and intercept of a plot of t/qt vs. t for surfactant adsorption on Berea sandstone.

Results and discussions

The composition of Berea sandstone was determined by XRF. The Berea core contains silica as its major constituent (91.02%) and alumina (2.97%). Other components include P2O5 (0.015%), K2O (0.871%), MgO (0.6%) and F2O3 (1.33%). The surface chemistry of the core sample has been defined by these two species i.e., Silica and Alumina. The FTIR peaks of adsorbent (before and after treatment) with surfactants at 521.27 cm−1, and 693.91 cm−1 are bending vibration of Si-O group adsorption peaks of untreated adsorbent shows at 776.33 cm−1 and 1080.17 cm−1. These peaks were symmetric and the asymmetric bond vibration stretching of Si- O. The peaks at 2851.85 cm−1 and 2924.64 cm−1 indicate the–CH2 presence in the structure.

Surface tension values of three surfactants at different concentrations were measured at different concentrations, and the critical micelle concentration was identified at the inflexion point of the curve50. The CMCs of the surfactants with different tail groups are shown in Fig. 2 (a), (b) and (c). As shown in Fig. 2 (a), (b), and (c), the CMC values were found to be 750, 480, and 40 ppm for S, D, and T, respectively. The measured surface tensions between CO2-philic surfactant solutions and CO2 (IFTc/w) is presented in Fig. 3 (a), (b), and (c), which highlights the positive effect of addition of tail groups in surfactant structure on lowering the IFTs and higher surfactant activity at c/w interface.

Fig. 2
figure 2

CMC graph surfactant (a). S, (b). D, (c). T, at 23 °C and atmospheric pressure.

Full size image
Fig. 3
figure 3

IFT c/w graphs of surfactant (a). S, (b). D, (c). T, as a function of pressure at 90 °C.

Full size image

The PZC was calculated from the plot illustrated in Fig. 4, by identifying the point where.

∆pH approaches zero, an indication of pH of zero surface charge. It was found that PZC value is 7.8 as indicated in Fig. 4. Any pH value higher than the PZC pH value would make the surface charge predominantly negative.

Fig. 4
figure 4

PZC determination by salt addition method.

Full size image

The adsorption values of surfactants used in the current study are found to be 0.65, 0.94, and 1.65 mg/g for S, D, and T surfactants, respectively. The same trend for all surfactants was observed showing a sudden increase in concentration, which may indicate the formation of “hemi micelles”, where the surface aggregates on the adsorbent surface are formed by lateral chain-chain interaction. These lateral interactions produce a driving force additionally, that superimposes established electrostatic attraction and ultimately causes a very sharp increasing trend in adsorption, as observed in the literature24. Normally, adsorption isotherms show a four-region trend as shown in Fig. 5. In these four-region isotherms, the first region shows a linear increase in adsorption as the concentration of surfactant increases with a slope of almost unity. In the second region, the adsorption is higher compared to the first region having an increased slope, this trend is normally due to the interactions of the rock surface and the surfactant molecules. Usually, no substantial distinction is present between the second and third regions, as the lateral interaction of chains of the surfactants and the surface slightly reduces the adsorption in the third region. The adsorption approaches its maxima in the fourth region of the isotherm. Any further surfactant concentration increase leads to micellization of molecules in the solution and this micellization prevents surfactants from adhering to the sand particles24. In all cases of the surfactants, the adsorption increased very steadily as the concentration increased up to some extent and showed no increase further. General adsorption isotherm of surfactants is illustrated in Fig. 5.

Fig. 5
figure 5

Adsorption isotherm.

Full size image

The surfactant structure and head group play a pivotal role in the adsorption of surfactants on the adsorbent particles. The adsorption at the interface of solid-liquid media is strongly influenced by adsorbent compositions, which makes the surface of the adsorbent negatively charged and shows weak interaction with the head part of anionic surfactant, a negatively charged portion. As shown in Fig. 6 (a), while adsorption of all surfactants was lower than conventional Alpha Olefin Sulfonate (AOS) foaming agent with reported adsorption of 2.5 mg/g on Berea33, surfactant T shows higher adsorption compared to D and S in the absence and presence of alkali, which can be attributed to the extra group attached in T structure that produces bulkiness.

Fig. 6
figure 6

Adsorption isotherm of the surfactants (a) S, (b) D, and (c) T, at 90 °C.

Full size image

Effect of pH on adsorption of anionic surfactant onto adsorbent

The effect of pH on the adsorption of anionic surfactants was performed by using sodium metaborate. The effect of pH on S, D, and T surfactants is shown in Fig. 6. From experimental results it was found that the anionic surfactant adsorption was decreased with an increase in solution pH. This happens due to two key reasons, first the molecular size and second the effect of surface charge. At the higher pH values the bigger particle generates a loss in conformational entropy of the chain of the surfactant on adsorption, which results in a decrease in adsorption. In the second case, the pH affects the charge on the surface of the adsorbent. Silica molecules can acquire a different charge which depends upon the solution pH. At pH lower than PZC, the silica particles acquire a positive charge and the reverse is true at higher pH, the mechanism is shown in Eq. 12 and Eq. 13:

$$\:SiOH+{H}^{+}\to\:SiO{H}_{2}^{+}$$
(12)
$$\:SiOH+O{H}^{-}\to\:Si{O}^{-}+{H}_{2}O$$
(13)

Concluding the above discussion, the charge present on the sandstone surface shifts by changing pH of the material and shows negative values at high pH resulting in a repulsion between the surfactant and sandstone surface and decreased adsorption.

At pH higher than 8.0, the adsorbent carries a negative charge as the PZC was found to be 8.0 for the sandstone sample used for the present study. Therefore, the anionic surfactants will have a lower surfactant adsorption. This is also experimentally confirmed by its low adsorption values by using different alkalis.

Effect of salt concentration on adsorption

The adsorbed quantity of surfactant increases by increasing the salinity. Adsorption of surfactants S, D, and T at different salinities. It was found for all sets of surfactants the adsorption increases as the salinity increases. This trend was due to an unequal electrical charge distribution at the interface between adsorbent particles and surfactants. As a result, an electrical double layer formed at the interface. Electrostatic repulsion between the adsorbed surfactant species decreases as the electrical double layer is compressed by increasing salinity and ultimately increases the adsorption capacity. This increase in adsorption with an increase in salinity implies that the adsorption of synthesized surfactant on adsorbent particles is seen to be a chemical process, which increases with salinity.

Modeling adsorption isotherms of anionic surfactants

The adsorption data of synthesized anionic surfactants on adsorbent particles at equilibrium state were fitted over Langmuir and Freundlich isotherm models. Table 1 summarizes the corresponding fitting parameters of experimental adsorption data to Langmuir and Freundlich equations. As seen in Table 1, R2 values are relatively improved in the Langmuir model compared to Freundlich model values, which indicates that data is best fitted in the Langmuir model. Figure 7 presents the Freunlich adsorption fitting isotherm, which shows a straight line between Log qe vs. Log ce, where the line slope is 1/n and the intercept is Log kf. Langmuir isotherm model, as shown in Fig. 8, suggests a plot of 1 / ce vs. 1 / qe, which is a straight line where 1 / (qo.b) is the slope and the intercept is 1 / qo.

Table 1 Modeling parameters.
Full size table
Fig. 7
figure 7

Freundlich model graph.

Full size image
Fig. 8
figure 8

Langmuir model fitting.

Full size image

Adsorption kinetics

The adsorption of S, D, and T onto the adsorbent reached equilibrium in 120, 150, and 200 min respectively. It was observed that the kinetic of surfactants on adsorbent particles follows a bi-phase pattern i.e. the rate was very fast at the start and increased linearly with time until reached to second phase, where increased slowly with a lower slope. This shows that maximum adsorption occurs in the first 90 min and finally, after 90 min, the kinetic isotherm reaches to equilibrium.

To determine rate constants, the data were fitted to pseudo-first and pseudo-second-order models and validated the adsorption kinetic process. Linear fits of both models are shown in Figs. 9 and 10. Linear regression method was used to determine the rate constants kf and ks while the unknown parameters were evaluated from the slope and intercept of the plots as shown in Tables 2 and 3. The best kinetics model was identified by comparing the correlation coefficients R2 and was found that synthesized surfactant adsorption kinetics follows pseudo-second-order kinetic model.

Table 2 Pseudo-first-order parameters.
Full size table
Table 3 Pseudo-second-order parameters.
Full size table
Fig. 9
figure 9

Pseudo-first-order kinetic model.

Full size image
Fig. 10
figure 10

Pseudo-second-order kinetic model.

Full size image

Conclusions

The surfactants S, D, and T adsorption onto crushed core particles were systematically studied. Experimental investigations for CMC, surface tensions, adsorption equilibrium, isotherm, and kinetic behaviors were performed. CMC values was reduced from 750 ppm for S to 480 for D and 40 for T surfactant, and by increasing surfactant activity at c/w interface, IFT values at high pressure reduced from 6 mN/m for S to lower than 2 mN/m for D and T surfactants. Silica active sites for surfactant adsorption were confirmed by XRF and FT-IR studies. Spectral changes following the treatment of adsorbent with surfactant were investigated, confirming the adsorption process. The effect of alkali concentration and salinity were also thoroughly examined, revealing that adsorption increases with higher salinity, whereas the reverse is true for increased alkali concentration. It was found that the surfactant concentration adsorbed on the surface of the adsorbent increases until it reaches the saturation point. Adsorption of S surfactant was reduced from 1.3 mg/g to 0.36 mg, and adsorption of D and T surfactants reduced from 0.96 to 0.34 and from 1.6 to 0.56, respectively. The Langmuir and Freundlich isotherms for adsorption were determined and the experimental data was best fitted in the Langmuir model as compared to the Freundlich model. It was also found that the adsorption isotherms follow pseudo-second-order kinetics models.