Molecular dynamics study on the aggregation behaviour of different positional isomers of sodium dodecyl benzenesulphonate

Abstract

All-atom molecular dynamics (MD) simulations were performed to study the aggregation behaviour of different sodium dodecyl benzenesulphonate positional isomers (xΦ12) where x = 1, 2, 3, 4, 5 and 6. In the simulation, the solvent accessible surface area, carbon and sulphur distribution, angle possibility distribution, chain conformation, hydration numbers, distribution of polar heads on the micelle surface, and the interaction energy among the benzene rings were analyzed. The simulated results showed that these six isomer micelles are more elliptical than spherical and the micelle radius increases with the shifting of the benzenesulphonate group from one side to the middle of the alkyl chain. In the micellar aggregate, the short alkyl chains are located at the polar layer of the micelle while the long alkyl chains assemble in the central region of the micelle. In the six different isomers, 1Φ12 isomer shows some special structural features.

---

1 Introduction

As they have unique amphiphilic properties, surfactant molecules can form a series of aggregates in aqueous solution, such as micelles, vesicles, bilayers and lamellae, and play increasingly important roles in industry and daily life.^1,2 Until now, these aggregates have been widely investigated both for scientific interest and for their important applications using a lot of different experimental techniques including electron paramagnetic resonance (EPR),^3,4 nuclear magnetic resonance (NMR),^4,5 small angle neutron scattering,⁶ light scattering,⁷ and theoretical methods such as Monte Carlo simulations⁸ and molecular dynamics (MD) simulations.^9–12 These techniques have great significance for elucidating the physical properties of the surfactant assemblies, especially for the micelles.

Micelles formed by sodium dodecyl benzenesulphonate (SDBS), one of the important anionic surfactants, play essential roles in cleaning and laundry detergents for industry and household. Especially in the petroleum industry, SDBS is increasingly applied to the enhance oil recovery. In commercial SDBS products,^13,14 different positional isomer of SDBS may exit, and they are usually considered as the impurities of the SDBS products. The physicochemical properties of these isomers are demonstrated to be different in aqueous solution.¹⁵ Since molecular structure decides properties and application, i.e. the linear dodecyl benzenesulphonate has good biodegradation while the branched chain counterparts are with poor biodegradation; further understanding the properties of individual component of the surfactant mixtures is very significant and helpful. The names and structures of different sodium dodecyl benzenesulphonate isomers are defined and shown as Scheme 1.

Scheme 1 Structure of sodium dodecyl benzenesulphonates.

In some experiments, these isomers of SDBS can be successfully separated, thus their different properties can be conveniently investigated. Abdel-Khalek et al.¹⁶ studied surface and thermodynamic properties of SDBS (xΦ12) micelles in aqueous solution, where x = 3, 4, 5 and 6. They found that the values of CMC for SDBS isomers increase gradually as benzenesulphonate group shifts from one side to the centre of the alkyl chain except for 6Φ12, while the micellization free energy for these four isomers is similar at the same temperature. Aoudia et al.¹⁷ used fluorescence spectrum method to investigate the micellar dispersions of several alkyl benzenesulphonates, and they proposed the presence of premicelle in solution. In 1997, Goon et al.⁵ investigated several linear alkylbenzenesulphonate (NaLAS) to discuss about their proton chemical shifts of different isomers using NMR technique. They noted that the chemical environments of alkyl chain were very different in consequence of micellization. According to the experimental results, they proposed that the short chains of SDBS isomers point to the micelle palisade layer while the long chains assemble in the core of micelle. They also found that meta protons to the sulphonate group on the phenyl ring exhibit large upfield shifts after micellization while ortho protons to the sulphonate group keep unshifted. This discrepancy showed that water boundary in micelles is located between meta and ortho protons in the phenyl ring, which is in accordance with the study of Das et al.¹⁸ In addition, Das proposed a schematic of 6Φ12 micelle structure according to the changes of proton chemical shifts before and after micellization.

Although these satisfactory understandings have been obtained by indirect experimental inspection and analyses of macroscopic properties, the acquaintance of physicochemical properties of micelle at the molecular level is not easy to be acquired only via the experiments. With the increase of computer power, computer simulations have been widely performed as one complement of experimental studies. Molecular dynamics (MD) simulation, considered as an effective tool, can provide direct microscopic details of structural and dynamical properties of surfactants aggregations at the molecular level, and have been widely used to investigate the aggregated properties at the interface or in the solution.

In the last decades, MD simulations have been carried out to study SDBS self-assembly and interface properties.^9–12,19,20 Gao et al.¹¹ studied the structure, shape, and size of SDBS micelle with different concentration in solution. They observed the evolution of micelle shape as surfactant concentration increased. Jang and coworkers¹² investigated the effects of surfactant molecule architecture at the decane/water interface, including series of sodium hexadecane benzene sulphonate isomers (xΦ16, x = 2, 4, 6, 8). They demonstrated that surfactant 4Φ16 packed more compactly at the decane/water interface and led to the lowest interfacial tension in comparison with other isomers. In another MD simulation, Palazzesi et al.⁹ investigated the structure and dynamics of SDBS and SDS micelles including micelle structure, structural and dynamics aspects of surfactants, water penetration, and counterion binding. They noted that SDBS micelle is more spherical and flexible than SDS micelle, and aliphatic chains of SDBS have more compact packing in the micelle. Klein and coworkers¹⁰ carried out the studies of SDBS phase behaviour in the bulk water using a coarse-grained model. The different morphologies of self-assemblies for five SDBS isomers with increasing concentration were systematically studied and compared with experimental explorations.²¹ They thought that the discrepancies of effective alkyl chain length contributed to the partly different phase behaviours in aqueous solution among these isomers. At the water/air interference, He et al.¹⁹ found that SDBS with straight alkyl chains can aggregate while the branched chains prevent the aggregation and make monomers cover the whole interface.

Although these more information about the SDBS isomers at the interface or in the solution have been obtained, the detailed structure properties of SDBS isomers micelles are still worthy to be further and systematically studied at the molecular level via MD simulation. The present work will focus on the surfactant assemblies of different positional isomers via MD method. The SDBS with no branched chain is also examined and discussed for comparison. Interpretation of physicochemical properties of SDBS isomers will be given in the subsequent parts at the molecular level.

2 Simulation details

In this simulation, one SDBS micelle was firstly built consisting of sixty SDBS monomers with all-trans configurations. The choice of one 60 monomer micelle based on Palazzesi's work.⁹ The initial coordinates of SDBS monomer were obtained from the automated force field topology builder (ATB)^22,23 with optimized geometrical configurations. Bonded and non bonded parameters were derived from the GROMOS 53A6 parameter²⁴ set. Initial charges were evaluated through the electrostatic potential (ESP) method introduced by Merz–Kollman.²⁵ Final charges and charge groups were generated through the method described in the ATB paper.^22,23 The detailed parameters of these six systems are shown in the end of the ESI.† Additional bonded parameters were generated using a Hessian matrix. A bond exists if all three eigenvalues of the interatomic force constant matrix between any two atoms are positive. Detailed calculations are in the paper.²⁶ Water molecules were described by the simple point charge (SPC²⁷) model.

In the present work, all-atom MD simulations were used to study the self-assembly behaviour of surfactants. The numbers of water molecules for 1Φ12, 2Φ12, 3Φ12, 4Φ12, 5Φ12 and 6Φ12 systems are 15 976, 15 942, 15 964, 15 989, 15 990 and 15 996, respectively. A 50 ns simulation time was used for these six systems. The simulation started with pre-assembled SDBS micelles. The sphere structure of original micelle containing 60 monomers was constructed by the Packmol program.²⁸ Sodium ions and water molecules were placed randomly in a cubic box with a total volume of 512 nm³. All simulations were carried out using the GROMACS 4.5.4 package.²⁹ The steepest descent method was used to eliminate the possible overlap of the initial configurations. Then, 200 ps MD simulations in NPT ensemble were performed to obtain the appropriate density of the system. The temperature and pressure were controlled using the Berendsen temperature coupling method³⁰ at 298 K (τT) and 1 atm (τP), respectively. The τT taken as 0.1 ps is the temperature time constant for coupling, and the τP taken as 1.0 ps is the pressure time constant for coupling. Additionally, the 200 ps NPT run is adequate to reach the appropriate temperature, pressure and density as shown in Fig. S1 and S2.† Next, based on the NPT equilibrium, a 50 nanoseconds NVT simulation with the important stable density (see Fig. S1 and S2†) was carried out at 298 K with the Berendsen temperature coupling method.³⁰ The last 20 ns of trajectories was used for further analyses. For the whole simulation, periodic boundary conditions were applied in all directions, and bond lengths were restrained via the LINCS³¹ algorithm. The cutoff distance for Lennard–Jones interactions was 1.2 nm and long-range electrostatic interactions were calculated using the Particle Mesh Ewald (PME³²) algorithm. A time step of 2 fs was applied in the simulation.

3 Results and discussion

Few snapshots at different time points of 1Φ12 micelle, as an example, were shown in Fig. S3† to give an intuitional view of the structure for the micelle. Moreover, the gyration radii of these six isomers as a function of simulation time are described in Fig. S4.† According to the time evolution of the micelle structure, we can conclude that all the systems reached equilibration after ca. 20 ns running and the micelle was stable during the MD simulation. The obtained results based on the last 20 ns NVT trajectories were credible.

3.1 Micelle shape and size

Micelle shape. The shape of micelle is one of the principal characteristics. The ratio I_max/I_min^33,34 is often used as a criterion of judging the shape, where I_max is the largest moment of inertia with x, y, and z axis, and I_min is the smallest one. For one perfect sphere, the value of I_max/I_min is equal to 1. As listed in Table 1, the data indicates that 2Φ12 and 3Φ12 micelles have more spherical structure. In comparison with the work of Palazzesi et al.,⁹ another criterion of the micelle shape, the eccentricity, is also calculated. The eccentricity, e, is defined as e = 1 − I_min/I_avg, where I_avg is the average moment of three inertia. The obtained e for 1Φ12 micelle indicates a less spherical shape than the result of Palazzesi et al., where the e is calculate to be 0.117. Considering the standard error, our result is in agreement with this value. For further study, the ratios of square roots of the principal moments of inertia with three axes are calculated and listed in Table S1.† According to the criterion of Tang et al.,³⁵ both these six micelles are oblate ellipsoids.

Table 1 Structural characteristics of micelles

	1Φ12	2Φ12	3Φ12	4Φ12	5Φ12	6Φ12
Imax/Imin	1.36 ± 0.07	1.23 ± 0.06	1.27 ± 0.07	1.44 ± 0.11	1.92 ± 0.18	2.06 ± 0.16
e	0.17 ± 0.04	0.11 ± 0.04	0.13 ± 0.04	0.19 ± 0.05	0.36 ± 0.05	0.38 ± 0.04
Rg (nm)	1.65 ± 0.01	1.67 ± 0.01	1.67 ± 0.01	1.70 ± 0.02	1.77 ± 0.04	1.82 ± 0.03
Rs (nm)	2.13 ± 0.01	2.16 ± 0.01	2.16 ± 0.01	2.19 ± 0.02	2.29 ± 0.05	2.35 ± 0.04
SASA (nm^2)	96.25 ± 2.45	113.53 ± 3.15	111.01 ± 2.93	113.98 ± 3.18	115.00 ± 2.74	119.34 ± 2.95

---

Micelle radius. Another fundamental characteristic of micelle is its size. There are different criterions characterizing the size, and one is about the radius of gyration R_g (Table 1). The values of R_g for six isomers are from 1.65 to 1.82 nm. For the 1Φ12 micelle, our result is similar to the value of 1.55 nm calculated by Palazzesi et al.⁹ The other criterion is about mean micellar radius R_s, defined³⁶ as:

As given in Table 1, the values of six isomers are from 2.13 to 2.35 nm. The values for 1Φ12 isomer and 6Φ12 are in reasonable agreement with the result obtained by Palazzesi et al.⁹ and the result of Triolo et al.³⁷ via small angle scattering method, respectively. From Table 1, the micelle radius increases gradually and the micelle is more ellipsoidal than spherical as the benzenesulphonate group shifts towards the middle of the alkyl chain, except 1Φ12 micelle.

Accessible surface area. The solvent accessible surface area (SASA) is another important property of micelle. Here we used the double cubic lattice method (DCLM)³⁸ to calculate the accessible surface area. In this technique, all counterions and water molecules are firstly removed from the system. One probe molecule is rolled across the surface of the micelle, which is used to mimic water molecule. The radius of this probe molecule is 0.14 nm. In this paper, the contact area is the total accessible surface area. From Fig. S5,† there is a large gap between 1Φ12 isomer (the smallest value) and others from Table 1, the accessible surface area of 1Φ12 isomer (96.25 nm²) corresponds with the result (93.43 nm²) obtained by Palazzesi et al.⁹ It is noted that no large difference exists among the 2Φ12, 3Φ12 and 4Φ12 isomers. However, the values of 5Φ12 and 6Φ12 isomers are increasing gradually. The change of SASA shows similar increasing tendency with the micelle size on the whole.

3.2 Distribution of terminal methyl

In the typical micelle model, hydrophobic chains of amphiphilic molecules tend to gather together in the inner core and polar head groups expose to the external shell. However, Zana et al.³⁹ proposed that parts of hydrophobic terminal carbon atoms should expose to the polar layer, and the distribution of polar head groups was heterogeneous on the micelle surface. This phenomenon was also reported by other simulation results^34,40,41 and our previous study.⁴² In our studied system, each amphiphilic molecule has two terminal methyl groups except 1Φ12. The distribution of the two terminal methyl groups in the micelle will be an interesting phenomenon to be discussed.

In order to study the distribution of long and short chains, the possibility distribution of terminal C and polar head S with respect to the micelle centre of mass (COM) was calculated, as shown in Fig. 1. For 1Φ12 isomer with one long tail chain, the distribution is similar to the result of previous studies.⁹ As benzenesulphonate group shifts from the end to the middle of alkyl chain, the distributions show that both long and short chains can stretch into the region of the sulphur distribution, indicating that the chains expose on the surface of micelle and not all chains have the all-trans structure due to their flexibility. In comparison, the peak position of the terminal C atoms of long chains is the closest to the COM of the micelle. The information above indicates that the short alkyl chains tend to be in the polar layer of the micelle while the long alkyl chains assemble in the central region of the micelle. Goon et al.⁵ have obtained the same conclusion from their experiments.

Fig. 1 Probability distribution of selected atoms C and S with respect to micelle COM. For the clear display, the spacing between each line was revised to 0.5 nm⁻¹.

Other conclusions can be acquired from Fig. 1. (i) There is an increasing portion of terminal carbon atoms exposing to the polar shell as benzenesulphonate group shifts from the end to the middle of alkyl chain. In particular, terminal carbon atoms of both long chains and short chains for 6Φ12 have the largest distribution in the region of the sulphur distribution. (ii) With regard to the terminal carbon atoms, regardless of long chains and short chains, their peak positions are similar, and the width of distribution curve increases with benzenesulphonate group moving to the middle of micelle. This manifests that hydrophobic chains are flexible in the micelle.

3.3 Chain direction and conformation

To characterize the orientation of alkyl chain in the micelle, two angles, α and β, defined as shown in Fig. 2 were calculated, where the angle α is the one between vectors I_C7→COM (C7 represents the atom C directly connected to the benzene ring) and I_C7→Cts (C_ts represents the terminal atom C of the short chain), and the angle β is the one between vectors I_C7→COM and I_C7→Ctl (C_tl represents the terminal atom C of the long chain). The density distribution for the angles α and β of SDBS molecules over the last 20 ns run is shown in Fig. 3. For 1Φ12 micelle, the angle β mainly ranges from 25° to 65° in agreement with the results obtained by Palazzesi et al.⁹ The large proportion of nonzero values of angles indicates that hydrocarbon chains do not orient perfectly to the micelle centre. For xΦ12 (x = 2,3,4,5) micelles, the angles β are centred between 10° and 40° with the peak occurring around 20°, meanwhile, the angles α are centred between 90° and 120° with the peak appearing around 100°. These phenomena are in favour of former conclusion that long chains are more likely to aggregate at the micelle centre than short chains. In 6Φ12 micelle, the wider distribution and similar peak position of these two angles (α and β) are mainly due to the similar length for two chains.

Fig. 2 Different patterns of position of two side chains relative to the connecting line of C7 and micelle center of mass (COM), and the proportion of different patterns. For clear display, the blue ball represents head group, and the 5Φ12 isomer drafts as an example corresponding to different patterns are shown.

Fig. 3 C_ts, C7, COM and C_tl, C7, COM angle probability distribution. The angle α is the one between vectors I_C7→COM (C7 represents the atom C directly connected to the benzene ring) and I_C7→Cts (C_ts represents the terminal atoms C of short chains), and the angle β is the one between vectors I_C7→COM and I_C7→Ctl (C_tl represents the terminal atom C of long chains).

For further details, the positions of two side chains relative to the connecting line of C7 and micelle COM were counted, and 1Φ12 isomer with one long tail chain is not discussed. The third angle θ, shown in Fig. 2, is defined as the one between vectors I_C7→Cts and I_C7→Ctl. The angle size of α and θ was compared and counted. When the angle α is less than θ, two alkyl chains are located on the opposite side of the connecting line. Otherwise, two alkyl chains are on the same side. According to this criterion, near 80% two alkyl chains are located on the opposite side of the connecting line as shown in Fig. 2. Appropriate 20% chains are on the same side probably resulting from the bending of tail chain caused by gauche defect.⁴³

The gauche defect, defined⁴³ as the conformation with the dihedral deviating beyond ±60° from the all-trans conformation of 180°, can be used to characterize the conformation of alkyl chains. The probability of gauche defect of nine dihedrals among hydrophobic chains is presented in Fig. S6.† Large difference of the curves exists between 1Φ12 and other five isomers due to the straight chain in the 1Φ12 monomer. In 1Φ12 micelle, the gauche defects at the beginning and end of the chains show an increasing probability while the defects have the low probability in the centre of the chain. This trend is consistent with the results of Palazzesi et al.⁹ Differently, in other micelles, dihedrals spanning the C atom directly connected to the benzene ring show large gauche defect, especially for the situation where the initial C atom of the dihedral is the short chain C atom nearest to the directly connected C atom. However, the effect of benzene rings to the rest dihedrals in these five micelles is little, and the gauche defects at the end of the chains also show an increasing probability. Our results show that the position of benzene ring mainly affects the ordering of the part adjacent to the benzene ring in the alkyl chain.

3.4 Outer distributions of head groups

In order to describe the uniform of the distribution of polar head groups, 2D number-density maps for S atoms of polar heads in the x–y plane were shown in Fig. 4. They were obtained by the statistics of conformations over 20 ns time. Using the least-squares fitting in this analysis, the holistic translation and the rotation of micelle were removed. In these maps, the colour from blue to red represents the increase of density, and the smaller red area indicates the denser distribution of polar groups in the region.

Fig. 4 2D number-density maps for S atoms of polar head in the x–y plane.

It is observed from Fig. 4 that 2Φ12 and 3Φ12 micelles are more spherical, which is in satisfactory agreement with the mentioned results in Table 1, and the distribution of sulphur atoms in 2Φ12 micelle is more uniform. However, the distribution of head groups in 1Φ12 micelle is the most uneven in these studied systems. Many red regions in the 2D number-density map show that many S atoms aggregate in certain regions.

In order to give an explanation on the particular phenomenon of 1Φ12 isomer, we calculated the radial distribution functions (RDFs) between benzene ring COM, (Fig. 5), and the interaction energies including Coulomb and van der Waals interaction energies between benzene rings of these isomers, (Table 2). As shown in Fig. 5, the curve of 1Φ12 isomer is different from other five isomers whose curves of RDFs are similar. Meanwhile, as listed in Table 2, the value of interaction energy in 1Φ12 micelle is more negative than the one in other isomer micelles. This means stronger interaction between benzene groups in 1Φ12 micelle. The π–π interaction⁴⁴ is common in the aggregates including benzene ring or conjugated structure. From the peak location, ca. 0.5 nm, in the RDF of 1Φ12 micelle, and the interaction energy gap between 1Φ12 and other isomers, we think that the π–π interaction exist in the 1Φ12 micelle. Palazzesi⁹ also proposed that the presence of π-stacking aggregates in 1Φ12 micelle. This π–π interaction can well explain the abnormal properties of 1Φ12 micelle with regard to other isomers.

Fig. 5 Benzene COM to benzene COM radial distribution functions.

Table 2 The energy (kJ mol⁻¹) of Coulomb interactions and van der Waals interactions among benzene rings

	1Φ12	2Φ12	3Φ12	4Φ12	5Φ12	6Φ12
Energy (vdW)	−261.93	−165.10	−156.50	−153.56	−153.80	−142.59
Energy (Coul)	−37.85	−21.54	−21.17	−18.64	−20.48	−21.91
Total	−299.78	−186.64	−177.67	−172.20	−174.28	−164.50

---

3.5 Interactions between micelle and water

The interaction between micelle and water can be characterized by hydration numbers of atoms or groups. In this paper, the integration of the RDFs of water oxygens around different groups within the distance of 3.5 Å (ref. 34 and 41) was calculated to obtain the hydration numbers. The hydration numbers of sulphonate group, benzene ring, short chain and long chain are listed in Table S2.† We note that discrepancies among different isomers for the same group are not apparent. The hydration numbers of sulphonate group and benzene ring are appropriately 2 and 1, respectively. This is reasonable for the stronger hydrogen bonds between sulphonate group and water molecule. For hydrophobic chains, the hydration numbers are much less than those of the above groups, but the short chains have much larger hydration number than the long chains. We think that it may be attributed to more opportunities orienting to the polar shell for the short chains. The hydration numbers of alkyl chains decrease from 2Φ12 to 6Φ12, except for 1Φ12 molecule.

To study more detailed differences of interaction with water, the hydration numbers of the S and Cn groups on the benzene ring are shown in Fig. 6. The values of sulphur atoms for six isomers are similar. We deduce that the effect of different hydrophobic chains on the interaction between head group and water molecule is weak. From Fig. 6, the hydration numbers of ortho-C atoms (C3 and C5) to sulphonate group are larger than those of meta-C atoms (C2 and C6), which is consistent with the experimental result of Goon et al.⁵ This difference is reasonable due to the outward position of ortho-C atom. For the further observation, the decrease of hydration number from ortho-C atom to meta-C atom for 1Φ12 isomer is slower in comparison with other five isomers due to the π–π interaction. From Fig. 6, in 1Φ12 and 2Φ12 micelles, the hydration numbers of C1, C2 and C6 are large while the numbers of C3 and C5 are similar to those in other four micelles. Hence, we think that the effect of steric hindrance of the short tail is not significant when the number of carbon atoms of short chain goes beyond 2. Particularly, the hydration number of C4 atom for 1Φ12 isomer is much smaller than the one for other isomers, mainly resulting from the π–π interaction that has been discussed in the preceding parts.

Fig. 6 Hydration numbers of the S and Cn groups on the benzene ring, which are counted for all water molecules within 3.5 Å of a group.

The hydration numbers from terminal C atom of short chain to terminal C atom of long chain are plotted in Fig. 7. It is observed that the hydration numbers decrease rapidly from the terminal C atom of short chain to the C atom connecting directly to the benzene, and the hydration numbers rise to a platform until the numbers of terminal C atom of long chain increase again. However, this law is not suitable for 1Φ12 isomer with one alkyl chain. For these six isomers, the hydration numbers of C atoms in short chains are larger than those in long chains on the whole, which is consistent with the results in Table S2.† This indicates that the short chains are more likely to direct to the polar shell while the long chains are more likely to aggregate in the inner core. Similar conclusion can be acquired from experiment.⁵

Fig. 7 Hydration numbers of Cn groups on the hydrocarbon tails, which are counted for all water molecules within 3.5 Å of a group.

The water penetration to the micelle has also attracted much attention for its direct correlation to the micelle application such as reactor and drug delivery. The density distributions of ortho carbon atoms, meta carbon atoms and water molecules for these six systems are shown in Fig. S7.† From the figure, we found that the density of water molecules in all studied systems only occurs between ortho carbon atoms and meta carbon atoms. Thus, the boundary of polar region and hydrophobic region is between ortho position and meta position on the benzene ring, and this division was proposed by Das¹⁸ and Goon⁵ through NMR experiments.

4 Conclusions

Molecular dynamics simulations were carried out to investigate the effects of different positions of benzenesulphonate group attaching to the dodecyl chain of SDBS on the aggregation behaviours. These six micelles are more ellipsoidal than spherical. The micelle radius decreases in the sequence of 6Φ12 > 5Φ12 > 4Φ12 > 3Φ12 ≈ 2Φ12 > 1Φ12. In the micelles, alkyl chains are not perfectly aligned to the micelle COM, in which short chains prefer to be in the polar shell and long chains gather in the micelle interior. Conformational analyses show that alkyl chains are the most ordered in 1Φ12 micelle. The benzene ring position on the chain largely makes the neighbouring section of the chain less ordered in other five micelles. The side chain can affect the hydration of meta-C atom and para-C atom to the sulphonate group on the benzene ring, hardly influencing that of the connected C atom and ortho-C atom. This effect turns to be weak when the number of C atoms of the short chain is over two. On the basis of MD results, two typical micelle structures are proposed, Fig. 8. In the models, the boundary for polar shell and nonpolar region and the orientation of different tail chains are described. Two conformational changes of tail chains with different proportion were also shown in the figure. In particular, for 1Φ12 isomer, the SASA evaluation shows that it packs more closely than other five isomers in the micelle. Head groups of 1Φ12 isomer distribute differently on the micelle surface compared with others. The above two phenomena can be explained with the π–π interaction exiting in the 1Φ12 micelle. Most of these behaviours are in favour of comprehending the essential micelle properties in order to make better applications of the relevant systems.

Fig. 8 Schematic diagram of anionic structures (1Φ12 and 6Φ12). The red in picture represents oxygen atoms in water molecules, and the blue represents hydrogen atoms in water molecules.

Acknowledgements

This work is supported by the Natural Science Foundation of China (21573130). We thank Dr Pamela Holt, Shandong University, for editing some parts of the manuscript.

Notes and references

1. D. H. Everett, Basic principles of colloid science, Royal Society of Chemistry, 1988.
2. C. Tanford, The Hydrophobic Effect: Formation of Micelles and Biological Membranes, 2nd edn, 1980.
3. B. L. Bales, L. Messina, A. Vidal, M. Peric and O. R. Nascimento, J. Phys. Chem. B, 1998, 102, 10347–10358.
4. G. L. Closs, M. D. E. Forbes and J. R. Norris, J. Phys. Chem., 1987, 91, 3592–3599.
5. P. Goon, S. Das, C. J. Clemett, G. J. T. Tiddy and V. V. Kumar, Langmuir, 1997, 13, 5577–5582.
6. S. Kumar, D. Sharma, D. Sharma and D. Kabir ud, J. Surfactants Deterg., 2006, 9, 77–82.
7. G. A. Sorci and T. D. Walker, Langmuir, 2005, 21, 803–806.
8. M. A. Floriano, E. Caponetti and A. Z. Panagiotopoulos, Langmuir, 1999, 15, 3143–3151.
9. F. Palazzesi, M. Calvaresi and F. Zerbetto, Soft Matter, 2011, 7, 9148.
10. X. He, W. Shinoda, R. DeVane, K. L. Anderson and M. L. Klein, Chem. Phys. Lett., 2010, 487, 71–76.
11. J. Gao, W. Ge and J. Li, Sci. China, Ser. B: Chem., 2005, 48, 470–475.
12. S. S. Jang, L. Shiang-Tai, P. K. Maiti, M. Blanco, W. A. Goddard Iii, P. Shuler and Y. Tang, J. Phys. Chem. B, 2004, 108, 12130–12140.
13. S. K. Hait, P. R. Majhi, A. Blume and S. P. Moulik, J. Phys. Chem. B, 2003, 107, 3650–3658.
14. S. Chakraborty, A. Chakraborty, M. Ali and S. K. Saha, J. Dispersion Sci. Technol., 2010, 31, 209–215.
15. J.-G. Ma, B. J. Boyd and C. J. Drummond, Langmuir, 2006, 22, 8646–8654.
16. N. A. Abdel-Khalek and A. M. A. Omar, J. Chem. Eng. Data, 1999, 44, 138–141.
17. M. Aoudia, W. H. Wade and M. A. J. Rodgers, J. Colloid Interface Sci., 1991, 145, 493–501.
18. S. Das, R. G. Bhirud, N. Nayyar, K. S. Narayan and V. V. Kumar, J. Phys. Chem., 1992, 96, 7454–7457.
19. X. He, O. Guvench, A. D. MacKerell Jr and M. L. Klein, J. Phys. Chem. B, 2010, 114, 9787–9794.
20. Z.-Y. Liu, N. Wei, C. Wang, H. Zhou, L. Zhang, Q. Liao and L. Zhang, AIP Adv., 2015, 5, 117203.
21. J. G. Ma, B. J. Boyd and C. J. Drummond, Langmuir, 2006, 22, 8646–8654.
22. K. B. Koziara, M. Stroet, A. K. Malde and A. E. Mark, J. Comput.-Aided Mol. Des., 2014, 28, 221–233.
23. A. K. Malde, L. Zuo, M. Breeze, M. Stroet, D. Poger, P. C. Nair, C. Oostenbrink and A. E. Mark, J. Chem. Theory Comput., 2011, 7, 4026–4037.
24. C. Oostenbrink, A. Villa, A. E. Mark and W. F. van Gunsteren, J. Comput. Chem., 2004, 25, 1656–1676.
25. U. C. Singh and P. A. Kollman, J. Comput. Chem., 1984, 5, 129–145.
26. J. M. Seminario, Int. J. Quantum Chem., 1996, 60, 1271–1277.
27. H. J. C. Berendsen, J. P. M. Postma, W. F. Gunsteren and J. Hermans, in Intermolecular Forces, ed. B. Pullman, Springer, Netherlands, 1981, vol. 14, ch. 21, pp. 331–342.
28. L. Martinez, R. Andrade, E. G. Birgin and J. M. Martinez, J. Comput. Chem., 2009, 30, 2157–2164.
29. B. Hess, C. Kutzner, D. van der Spoel and E. Lindahl, J. Chem. Theory Comput., 2008, 4, 435–447.
30. H. J. C. Berendsen, J. P. M. Postma, W. F. van Gunsteren, A. DiNola and J. R. Haak, J. Chem. Phys., 1984, 81, 3684.
31. B. Hess, H. Bekker, H. J. C. Berendsen and J. G. E. M. Fraaije, J. Comput. Chem., 1997, 18, 1463–1472.
32. U. Essmann, L. Perera, M. L. Berkowitz, T. Darden, H. Lee and L. G. Pedersen, J. Chem. Phys., 1995, 103, 8577.
33. A. R. Rakitin and G. R. Pack, J. Phys. Chem. B, 2004, 108, 2712–2716.
34. A. D. MacKerell, J. Phys. Chem., 1995, 99, 1846–1855.
35. X. Tang, P. H. Koenig and R. G. Larson, J. Phys. Chem. B, 2014, 118, 3864–3880.
36. S. Bogusz, R. M. Venable and R. W. Pastor, J. Phys. Chem. B, 2000, 104, 5462–5470.
37. L. J. Magid, R. Triolo, J. S. Johnson and W. C. Koehler, J. Phys. Chem., 1982, 86, 164–167.
38. F. Eisenhaber, P. Lijnzaad, P. Argos, C. Sander and M. Scharf, J. Comput. Chem., 1995, 16, 273–284.
39. B. L. Bales, M. Benrraou, K. Tiguida and R. Zana, J. Phys. Chem. B, 2005, 109, 7987–7997.
40. J. Shelley, K. Watanabe and M. L. Klein, Int. J. Quantum Chem., 1990, 38, 103–117.
41. J. Gao, W. Ge, G. Hu and J. Li, Langmuir, 2005, 21, 5223–5229.
42. G. Liu, H. Zhang, G. Liu, S. Yuan and C. Liu, Phys. Chem. Chem. Phys., 2015, 18, 878–885.
43. H. Domínguez, J. Phys. Chem. B, 2007, 111, 4054–4059.
44. C. A. Hunter and J. K. M. Sanders, J. Am. Chem. Soc., 1990, 112, 5525–5534.

---

Footnote

† Electronic supplementary information (ESI) available: Hydration numbers of different sections for SDBS molecules, the fluctuation of solvent accessible surface area with time evaluation, and the density distribution of ortho and meta carbon atoms on the benzene ring and water molecules with respect to the micelle center of mass. See DOI: 10.1039/c6ra05188j

---