(522d) Carbon Dioxide Absorption In Amine-Ionic Liquid Hybrid Solvents

The amount of carbon dioxide (CO₂) emitted into the atmosphere by various industries is of great concern due to the global warming effects of CO₂, and industries worldwide are faced with increasing pressure by environmentalists, and the public at large, to reduce their emissions of the gas. In 2008, nearly 30000 mega tonnes (Mt) of CO₂ was emitted into the atmosphere^[1]. This is nearly a 100% increase since 1971. The increasing worldwide demand for relatively cheap electricity and the advances in coal-to-liquids technology has resulted in 42.9% of CO₂ emissions stemming from coal industries, with oil and gas industries accounting for 36.8% and 19.9% of CO₂ emissions respectively^[1].

South Africa produces 224 million tonnes of coal annually and 25% is exported, making South Africa the third largest coal exporting country. It is estimated that South African coal reserves amount to 53 billion tonnes, enough to meet the country’s electricity needs for nearly 200 years to come^[2]. It is estimated that 53% of coal produced in South Africa is used to generate electricity, while 36% of electricity produced worldwide stems from the burning of coal^[2].

The primary electricity utility in South Africa is Eskom Ltd.. While the company harnesses energy from a diverse range of sources, at least 77% of its electricity production stems from the burning of coal at its 14 coal-fired power plants^[2]. South Africa enjoys cost effective electricity due to its abundant coal reserves, but coal energy also results in the country emitting substantial amounts of CO₂. Eskom’s coal operations have made it the second highest CO₂-emitting company in the world.

The flue gas emitted by a pulverised coal power plant typically contains 13 vol% CO₂, 68 vol% N₂, 16 vol% water, 3 vol% O₂, and lower concentrations of other components, 200ppm SO₂, 60ppm NO_X, 60 ppm hydrocarbons^[3]. The flue gas is typically emitted at 1-1.7 bar^[4].

A conventional technique to remove CO₂ from the flue gas and hence prevent the emission of CO₂, is to pass the flue gas through an absorption column containing a solvent. The solvent selectively absorbs CO₂ while allowing other flue gas components to be emitted. The CO₂-loaded solvent is then heated in a regeneration column to release the CO₂for compression and storage, while the solvent is recycled.

There are many solvents under investigation for the energy efficient absorption of CO₂. Conventional solvents are amine based including, Monoethanolamine, Diethanolamine, Methyl diethanolamine, Diglycol Amine, Triethanol Amine, Methyl Monoethanolamine, Amino Ethyl Ethanolamine, Ethyl Amino-ethanol, and Butyl Aminoethanol^[5]. Amines achieve reactive absorption of CO₂ with a high absorption rate at low pressure, in comparison to physical solvents. A common shortcoming of amines is their high corrosivity, requiring amine solvents to be diluted with water at a composition 50 to 70% H₂O by mass. This not only limits the absorption capacity of the solvent, but also results in very high energy costs when heating the solvent in a regeneration column to isolate CO₂.

A relatively new approach is the study of CO₂ absorption in ionic liquids. Ionic liquids are physical solvents composed entirely of ions. They are chemically and thermally stable, and non-corrosive . In this research we investigated the effect of combining amines with ionic liquids instead of water. The effect was studied in terms of CO₂ solubility in the solvent. The amines investigated were Monoethanolamine (MEA), Diethanolamine (DEA), and Methyl-Diethanol (MDEA) amine. The ionic liquids studied were [1-Butyl-3-Methyl Imidazolium] [Tetrafluoroborate] ([bmim][BF₄­),] and [1-Butyl-3-Methyl Imidazolium] [bis(trifluoromethylsulphonyl) imide] ([bmim][Tf₂N]).

The solvents studied contained 45 to 100% of the above ionic liquids (by mass), with the above amines occurring singularly or in combinations at compositions of 10 to 50% by mass. 

All solvents were contacted with CO₂ gas in order to measure the equilibrium CO₂ loading in each solvent. CO₂ loading measurements were conducted at CO₂partial pressures ranging from 0.5 to 15 bar, and at isotherms of 303.15, 313.15, and 323.15 K (approximately).

Measurements were conducted using an Intelligent Gravimetric Analyser (IGA-01) constructed by Hiden Analytical Ltd., which utilises a sensitive microbalance to track the absorption of CO₂gas into the solvent by noting an increase in the mass of the solvent. The microbalance has a resolution of ±0.0001 mg or ±0.1 µg. The apparatus achieves a vacuum of 1e-7mbar and operates at pressure up to 20 bar. Temperature of 273.15K to 773.15 K is achievable using a refrigerated waterbath or cryofurnace. 80 to 100 mg solvent was required for loading into the stainless steel reactor.

The low quantity of sample required and the programmable nature of the apparatus, made absorption measurements efficient and less time consuming. The weight change of the sample as gas absorption occurs was calculated by taking into account all contributing forces including the mass of the counterweight and its associated attachments (hook and chain), the sample container and its associated attachments (hook and chains), the sample itself, and the buoyancy force exerted by the gas in the reactor.

Due to the sample being of liquid form, there is a significant change in the sample density upon absorption of the gas, thus making the sample significantly more buoyant and altering the actual weight reading of the sample. This buoyancy force was taken into account by first noting the weight reading of the sample at different gas densities using a non-absorbent gas^[6]

CO₂ solubility measurements were presented in terms of CO₂ mole fraction versus CO₂partial pressure for each sample and at each isotherm.

The data for the above systems are to be presented and discussed, along with the modelling of the data using the Soave-Redlich-Kwong (SRK) equation of state for CO₂ absorption into the ionic liquid component, and Deshmukh-Mather model for reactive absorption of CO₂into primary, secondary and tertiary amine components. 

The measurements found that the combining of amines with ionic liquids was significantly beneficial to the absorption of CO₂ in the solvent, in comparison to the use of pure ionic liquids. All hybrid samples containing amines and ionic liquids achieved higher CO₂ absorption at low pressures of 0.5 to 4 bar, than conventional amine solvents diluted with water^[7], as well as pure ionic liquid solvents measured.

However, at pressure greater than 4 bar, certain hybrid solvents, particularly those containing [bmim][BF₄] ionic liquid and higher compositions of secondary and tertiary amine achieved lower CO₂ absorption than pure ionic liquid solvents of [bmim][BF₄] and [bmim][Tf₂N], indicating superior absorption capacity of pure ionic liquids over hybrid solvents. Hybrid solvents containing the [bmim][Tf₂N] ionic liquid however, achieved superior CO₂absorption compared to pure ionic liquids even at high pressure.

Overall, the combining of amines with ionic liquids resulted in significantly higher CO₂absorption, indicating that this direction of research is certainly worth pursuing further.

References: 

1. International Energy Agency (IEA), 2010, “Key World Energy Statistics – 2010”, International Energy Agency, Paris Cedex, France.
2. Eskom Power Generation, 2011, “Eskom – Coal Power”, Eskom Holdings Ltd., South Africa. Accessed 28/02/2011.  http://www.eskom.co.za/live/content.php?Item_ID=279
3. Brennecke J.F. and Gurkan B.E., 2010, J. Phys. Chem. Lett., 1, 3459–3464.
4. National Energy Technology Laboratory (NETL), 2010, “Doe/Netl Advanced Carbon Dioxide Capture R&D Program: Technology Update”, NETL, U.S.A. Accessed 25/5/2011. http://www.netl.doe.gov/technologies/coalpower/ewr/pubs/CO2%20Capture%20Tech%20Update%20Final.pdf
5. Ma’mun S., Nilsen R. and Svendsen H.F., 2005, J. Chem. Eng. Data, 50, 630-634
6. Macedonia M.D., Moore D.D., and Maginn E.J., Langmuir 2000, 16, 3823-3834.
7. Park S.H., Lee K.B. , Hyun J.C. , Kim S.H., 2002, Ind. Eng. Chem. Res., 41, 1658-1665