Integral design of non-toxic surfactants for industrial applications

Abstract

Surfactants are chemicals that increase solubility and dispersion of hydrocarbons and oils, and they are widely used in detergents and other cleaning products. After use they are discharged into sewage systems or directly into surface waters, ending up dispersed in the environment polluting the water, the soil and other sediments. The toxic effects of surfactants on various aquatic organisms are known. The vital role surfactants play in modern society means there is no option to stop using them. Thus, non-toxic and biodegradable surfactants are required to decrease their final impact on current water resources.

In this thesis, the approach chosen to discover novel surfactants is to create a computational method, which could act as a screening stage to ensure only the most promising non-toxic and/or biodegradable product formulations make it to experimental testing. This would have the effect of reducing the number of necessary experiments, leading to savings in associated time and costs. Molecular Dynamics (MD) simulations underline the computational method, which represent the system as a collection of particles interacting by mathematical relations termed force fields. This permits the calculation of structural, thermodynamic, and dynamical properties. Therefore, it is an ideal tool for the study of surfactant-water-oil ternary mixtures because the molecular structure is known to strongly influence the surfactant behaviour. Whilst this concept is not new, there has been little progress in applying it to the discovery of non-toxic and biodegradable surfactants.

To simulate surfactants at the molecular scale requires large system sizes that must be run for long durations of time. This demand will be difficult to meet by accounting for all atoms in a molecule. It is for this reason it is chosen to work with models of lower resolution, termed ‘coarse-grained’ (CG) models. Here multiple atoms are grouped into one bead which interacts via an effective force field. The parameterisation of a CG model is an important step, impacting the model accuracy. In this work recent incarnations of the Statistical Associating Fluid Theory (SAFT) were used to develop the CG force fields used in MD simulations. Here the versatile Mie force field is used to represent the intermolecular interactions with a Coulombic relation used to account for electrostatic interactions in explicitly charged beads. The SAFT approach is a top-down CG method where the force field parameters are determined via a molecular based equation-of-state. In this thesis a corresponding states version of the equation-of-state was used to find force field parameters for the like-like interactions of the uncharged beads. The SAFT approach allows for faster model parameterisation compared to solely optimising parameters via iterative simulation. There has also been extensive progress in extending SAFT force fields to surfactant-water systems. Despite this, CG MD simulations of surfactant ternary mixtures have not yet been performed using force fields obtained by the SAFT route. Therefore, another key challenge of this thesis is to move the research line forward.

In this thesis SAFT-derived force fields were used to study two different ternary ionic surfactant mixtures: sodium bis (2-ethylhexyl) sulphosuccinate (AOT) in water and cyclohexane, and sodium bis (3-(trimethylsilyl)-1-propanol) sulphosuccinate (AOTSiC) in water and supercritical (sc) CO2. Derivation of the surfactant models required the study of a variety of fluid systems, including pure esters and binary mixtures comprising methyl acetate and water, AOT and water, and scCO2 and water. The described systems contain sufficient experimental data to test the SAFT approach. This is a necessary preliminary step which would give more confidence in using SAFT to study properties of promising novel environmentally friendly surfactants for which data may be limited. This could shed light on the impact of structure on product/process performance whilst reducing the number of necessary experiments.

The development of a SAFT model for AOT and its aqueous mixtures was the first objective of this thesis, since a previous one did not exist. The surfactant model is represented in a group contribution manner, where each chemical moiety is represented by a unique bead. By optimising the surfactant-surfactant and surfactant-water intermolecular interactions this allowed the study of phase behaviour and structural properties. MD simulations using these force field parameters showed the formation of a lamellar phase at ambient conditions. At high temperature a transition to an isotropic phase occurs. The MD simulations do not indicate a structural transition in the middle of the lamellar phase region, instead there is a transition from flexible to rigid bilayers. These observations, as well as the calculated bilayer thickness at room temperature, are in agreement with experimental data. This is encouraging, since only thermodynamic data was included in the force field parameterisation.

Studying the AOT-water-cyclohexane system allowed for the assessment of the SAFT force field to model ternary surfactant mixtures. Due to the group contribution nature of SAFT-𝛾 Mie, many of the intermolecular interactions were simply transferred from the AOT-water model. This model is able to capture both phase morphologies and structural properties. The MD simulations reveal a transition from phase-separated systems to isotropic reversed micelle (RM) phases at low water content. The calculated average cluster size is in good agreement with experimental findings. The relationship between water content and RM morphology was studied. The reduction of water content results in a sharp reduction in average cluster size, highlighting its importance when considering RM stability. When considering the relationship between water content and RM shape, there appears to be little impact since a predominant spherical shape exists for all systems investigated. This has implications for RM experimental investigations, where there are still many differing views regarding the shape of RMs, and where incorrect shape assumptions can affect the accuracy of the results. The advantage of this MD methodology is that no such assumptions about the RM shape must be made for analysis.

Preliminary steps have been completed to study the AOTSiC-water-scCO2 system using the SAFT-based MD methodology. Due to the similarity in structure to AOT, many of the force field parameters were carried over from the previous model. A model for the water – scCO2 binary mixture has been created. A single binary interaction parameter was used to optimise the fit to the liquid-liquid equilibrium. These interactions can be transferred to create a preliminary AOTSiC-water-scCO2 model, which can be tested by comparing to experimental data. Once validated, this would allow the study of the effect of surfactant tail structure and functional groups on the resultant properties of the water-in-scCO2 RMs. The results of this investigation could then guide the discovery of alternative surfactants to replace expensive and environmentally unfriendly fluorinated surfactants.

Despite their simplicity, the CG models created in this work possess a level of transferability, robustness, and representability. They can be used to predict properties not included in the original parameterisation strategy, with good levels of accuracy. The SAFT theory is hence a suitable approach to parameterise the force fields that could be used in a computational screening method. As has been shown in this work, the methodology requires the prior knowledge of experimental data in order to assess the surfactant structure-performance relationship. This could be obtained from literature or via collaborations with experimentalists both in industry and academia. This approach can be used for the discovery of non-toxic surfactants for a wide range of applications, including drug delivery, food, cleaning products and enhanced oil recovery.

Date of Award	14 Sept 2022
Original language	English
Awarding Institution	University of Bath
Supervisor	Carmelo Herdes Moreno (Supervisor) & Tina Düren (Supervisor)