Carbon capture is increasingly garnering attention as an effective method to reduce carbon emissions released into the atmosphere. Industrial projects testing and implementing carbon capture have been launched worldwide over the past decades, with ongoing research to develop even more effective methods.  

This use case showcases how molecular dynamics simulations in QuantistryLab can provide valuable insights into the viscosity of solutions used in carbon capture applications, a key property for enhancing their efficiency and speed.  

Application of interest: Carbon capture

One of the most popular carbon capture methods is chemical absorption, also known as carbon scrubbing, where chemical solutions remove CO₂ from the air near sources such as thermal power plants.  

Carbon scrubbers commonly use solutions containing various alkanolamines, which bind carbon dioxide at lower temperatures and release it at higher temperatures. This allows captured carbon to be stored by heating the solution, regenerating its absorbent properties.  

A major challenge with this approach is that most conventional amine absorbents are water-based, which can cause issues such as corroding steel containers and consuming high amounts of energy when heated up and cooled down.

In recent years, research efforts have explored non-aqueous solutions as an alternative to water-based amine absorbents. Replacing water with an organic solvent could significantly reduce total energy consumption, minimize thermal degradation, and prevent damage to equipment. Studies have shown that non-aqueous solutions can absorb faster and more efficiently, while lowering the amount of energy required for regeneration.  

However, developing effective non-aqueous solutions for carbon capture applications requires identifying and selecting compounds with low viscosity to speed up carbon absorption, but also low volatility, as highly volatile alcohols often used as organic solvents can be quickly lost to evaporation.    

Since viscosity and volatility tend to be inversely correlated, research efforts focus on finding solvent mixtures that optimize both properties. Recently, polyethylene glycol ethers, such as diglyme, triglyme and tetraglyme, have been proposed as promising candidates to create solvents with optimal properties for carbon capture applications.  

However, R&D efforts to identify the optimal mixture of solvents for fast and efficient carbon capture can be time-consuming and resource-intensive, as multiple iterations and experimental measurements are required to optimize key properties of the final formulation, such as viscosity. Chemical simulations offer a time- and cost-effective alternative to predict and optimize key properties of chemical solutions, such as viscosity, when compared to traditional experimental approaches.  

Use case: Viscosity simulations of polyethylene glycol ethers

With just a web browser, QuantistryLab allows the user to run molecular dynamics simulations of chemical systems, including plasticizers, lubricants, polymers, and electrolytes. This enables the prediction of many of their macroscopic properties, like viscosity.  

To estimate the viscosity of a chemical system, QuantistryLab deploys molecular dynamics simulations to measure fluctuations in the stress tensor over time. Using the Green-Kubo approach, a statistical mechanics method that employs autocorrelation functions of the stress tensor, a prediction of the viscosity at equilibrium is obtained.    

The first step in running these simulations is to generate a model of the chemical systems of interest. This can be easily done with just a few clicks In QuantistryLab by selecting the desired chemicals from the compound library create a liquid preparation.  

For this use case, separate models of an alkanolamine and three polyethylene glycol ethers solvents were first created:

• 2-ethylhexan-1-amine (EHA);

• Bis(2-methoxyethyl) ether (diglyme);

• Triethylene glycol dimethyl ether (triglyme);

• Tetraethylene glycol dimethyl ether (tetraglyme).

Once the models are generated, viscosity can be calculated by initiating a workflow with QuantistryLab’s viscosity feature and selecting the desired temperature and pressure for the simulation, in this case 323.2 °K (~50 °C) and 101 kPa. The simulation yields a 3D atomistic model for each chemical system, representing the dynamical behavior of the atoms as a function of time.

The results obtained with QuantistryLab were compared to experimental data reported in peer-reviewed literature, and summarized in the figure above. The simulations demonstrate exceptional accuracy, with deviations as low as 0.02 mPa·s for diglyme and 0.15 mPa·s for tetraglyme. The overall average difference is just 0.1775 mPa·s, underscoring the model’s precision in closely matching experimental viscosity values and supporting its reliability for predicting viscosity in carbon capture applications.

Using the same workflow described above, models of multiple mixtures of these three polyethylene glycol solvents, together with EHA, can be created to recreate multiple possible formulations of carbon capture solutions and determine their viscosity. These simulations can be implemented in R&D workflows to accelerate the optimization of formulations for carbon capture applications, significantly reducing the need for experiments to a few selected systems that have been pre-selected using computational methods.  

This same approach can be extended to any number of applications across multiple areas of interest, including energy storage, metals and alloys, lubricants, semiconductors and specialty chemicals. For all these applications and more, QuantistryLab offers a full range of computational solutions to study the properties of a wide range of chemical and material systems.

‍

‍