WO2015055651A1

WO2015055651A1 - Method for removing carbon dioxide from a gas stream by means of a solution comprising 2-amino-2-methyl-1 -propanol (amp)

Info

Publication number: WO2015055651A1
Application number: PCT/EP2014/072017
Authority: WO
Inventors: Hans T. Karlsson; Helena SVENSSON; Christian Hulteberg
Original assignee: Individual
Current assignee: Individual
Priority date: 2013-10-14
Filing date: 2014-10-14
Publication date: 2015-04-23
Anticipated expiration: 2016-04-14

Abstract

A method for removing carbon dioxide from a gas stream by contacting the gas stream with a composition comprising an amine and an organic solvent is disclosed.

Description

METHOD FOR REMOVING CARBON DIOXIDE FROM A GAS STREAM BY MEANS OF A SOLUTION COMPRISING 2-AMINO-2-METHYL-1 -PROPANOL (AMP)

Field of the invention

The present invention relates broadly to carbon dioxide capture methods. More 5 particularly, the invention relates a method for removing carbon dioxide from a gas stream using a composition of an amine and an organic solvent.

Background

It is generally well accepted that combustion activities resulting in the emission of carbon dioxide (C0₂) to the environment is undesirable. Efforts to reduce C0₂ emissions include reducing combustion and capturing the emitted C0₂.

Most of the major challenges in post-combustion C0₂ capture revolve around the relatively large parasitic load that the carbon capture and storage (CCS) facility imposes on the power plant. Much of this results from the energy required to regenerate the absorbent. Energy is also required for C0₂ compression, although this is less than that needed for regeneration. It is believed that there is less potential for further significant saving in the area of compression.

Use of chemical solvent (e.g. amine) scrubbing to remove C0₂ from natural gas and hydrogen has been used since 1930 (G. T. Rochelle, 2009). Absorption of C0₂ in a chemical solvent based process can be achieved by using a solvent or a solvent combination (W.I. Echt, 1997). The chemical solvent bonds loosely to C0₂ when exposed to the gas stream e.g. (flue gas) in the scrubber. The loaded solvent is then led through a stripper to regenerate the chemical solvent. Currently there are different solvents being used to capture C0₂ the most common of which are aqueous alkanolamines (K. Thambimuthu, 2002). Primary and secondary alkanolamines are used as solvents in C0₂ capture due to the fact that both primary and secondary amines form stable carbamates with C0₂ in aqueous solutions. Their C0₂ loading is limited to 0,5 mol C0₂/mol amine according to their reaction stoichiometry. Hydrolysis to generate the free amines occurs with all amines and can lead to the C0₂ loading being exceeded, especially at higher absorption pressure (S. Xu et al, 1992). Monoethanoalmine (ME A) is a primary amine that is extensively used as a chemical solvent for absorption processes. However, MEA suffers from solvent degradation and thermal degradation during usage although is a relatively cheap amine and the amine loss is calculated to $5/ton C0₂ (G. T. Rochelle, 2009). Diethanolamine (DEA) is another commonly used amine for removal of C0₂.

In contrast, tertiary amines can reach a higher stoichiometric absorption of 1 mol C0₂/mol amine. However, due to tertiary amines having lower reaction rates than primary and secondary amines, reaction with C0₂ is somewhat harder to achieve. Sterically hindered amines in aqueous solutions also show a high stoichiometric absorption of 1 mol C0₂/mol amine but at absorption rates comparable to those of primary and secondary amines (G. Sarotori, W. Savage, 1983). N- methyldiethanolamine (MDEA) is a tertiary amine that because of its low reaction rate is normally mixed with a promoter such as piperazine (PZ) to improve the reaction rate. PZ is an unpleasant chemical that can cause harm to both environment and human health and has also been shown to react with C0₂ thereby altering the reaction mechanism.

In summary, aqueous solutions of sterically hindered amines have been widely researched as an alternative to commercial systems using monoethanolamine (MEA), and diethanolamine (DEA). The main reason for this is the higher loading capacity of the aqueous hindered-amine solutions which is mainly due to the fact that a hindered amine does not form a stable carbamate. Rather the reactions are forced towards the formation of bicarbonate which doubles the loading capability from 0.5 to 1 mole C0₂/mole amine.

Current amine technologies result in a loss of net power output of about 30% and a reduction of more than 10% in efficiency. There is little point of retrofitting power plants of low thermal efficiency since the efficiency losses would then render the plant uneconomical. Furthermore, the capital cost requirements are high and the water requirement is considerable, nearly double the net MWh for water-cooled plants. The solvent volatility and degradation issues have a negative impact on the process economy. The relatively low boiling point of certain amines may result in solvent carryover from the C0₂ regeneration step. The high temperature (i.e. large amount of energy) required for regeneration of the amine and release of captured C0₂ causes degradation of the amine and decrease of the absorber efficiency. Another problem is that steam extraction for solvent regeneration reduces flow to LP turbine, which has an impact on efficiency and turn-down capability. Amine solutions with high reaction rates usually have a high heat of absorption increasing the energy demand for regeneration of the amine solution. Amine solutions with a low heat of absorption generally have a low reaction rate, increasing the amount of amine solution needed in the absorption process.

It would therefore be desirable to provide alternative amine compositions for use in the absorption of C0₂ from gas streams.

Summary

Consequently, the present invention seeks to mitigate, alleviate, eliminate or circumvent one or more of the above-identified deficiencies in the art and disadvantages singly or in any combination by providing, in one aspect, a method for removing carbon dioxide (C0₂) from a gas stream comprising the steps of:

providing a gas stream containing C0₂; and

contacting the gas stream with a composition comprising:

2-amino-2-methyl-l-propanol (AMP); and

a solvent selected from N-methyl-2-pyrrolidone (NMP) or triethylene glycol dimethyl ether (TEGDME);

to remove the C0₂ from the gas stream.

In some embodiments of the method described above, contacting the gas stream with the composition forms a solid that is separated from the composition. The solid may form in crystalline or amorphous form. Alternatively, the solid may form as a slurry. Any suitable separation unit known to the skilled person may be used to separate the solid from the composition. Separation may occur by a mechanical method such as filtration. The solid may then be pumped to a regeneration unit (if present) as described below.

According to an embodiment, the method further comprises the step of recycling the composition separated from the solid back to the contacting step. In an embodiment, the method further comprises the step of heating the solid to release C0₂ and regenerate AMP. A variety of regeneration units typically known to the person skilled in the art may be used for heating the solid.

In yet another embodiment, the method further comprises the step of recycling the AMP separated from the heating step back to the contacting step.

According to some embodiments of the method, the solid is heated at a temperature below 100 °C. In a preferred embodiment of the method, the solid is heated at a temperature between about 70 to 80 °C. Any particular source of heat may be used, including waste heat, to regenerate AMP and release C0₂. As the solid or slurry will usually be highly concentrated, little to no energy will be lost in heating any residual NMP and TEGDME thereby increasing the method's efficiency. The inventors are of the opinion that reducing the energy needed for the regeneration will also reduce the costs of the C0₂ capture operation significantly as it increases the amount of energy that can be sold to market.

In some embodiments of the method, the gas stream is contacted with the composition at a temperature between about 25 to 50 °C. In one preferred embodiment of the method, the gas stream is contacted with the composition at a temperature of about 25 °C. Alternatively, the gas stream is contacted with the composition at a temperature of about 50 °C in the method.

According to a preferred embodiment of the method, the solvent is NMP. In an alternative embodiment, the solvent is TEGDME.

The source of the C0₂-containing gas stream is not intended to be particularly limited and may arise from a variety of applications. In some embodiments of the method, the C0₂-containing gas stream is a flue gas stream.

According to some embodiments of the method, the composition comprises 15 to 25 wt.% AMP.

In some embodiments of the method, the released C0₂ is received at elevated pressure, such as about 6 bars. In a preferred embodiment, a C0₂ gas stream is received at 6 bar when the solid is heated at 90°C.

According to another aspect, there is provided a use of a composition comprising: 2-amino-2-methyl-l-propanol (AMP); and

to remove C0₂ from a gas stream.

In some embodiments of the use described above, AMP and C0₂ react to form a solid that is separated from the composition.

In another embodiment of the use, the composition separated from the solid is recycled to remove C0₂ from the gas stream.

According to another embodiment of the use, the solid is heated to release C0₂ and regenerate AMP.

In some embodiments of the use, the AMP separated from the heating step is recycled back to the composition to remove C0₂ from the gas stream.

According to some embodiments of the use, the solid is heated at a temperature below 100 °C.

In some embodiments of the use, the solid is heated at a temperature between about 70 to 80 °C.

In an embodiment of the use, the gas stream is contacted with the composition at a temperature between about 25 to 50 °C. In one preferred embodiment of the use, the gas stream is contacted with the composition at a temperature of about 25 °C. In an alternative embodiment of the use, the gas stream is contacted with the composition at a temperature of about 50 °C.

In an embodiment of the use, the solvent is NMP. In an alternative embodiment of the use, the solvent is TEGDME.

Similar to the method described above, the source of the C0₂-containing gas stream in the use is not intended to be particularly limited and may arise from a variety of applications. In some embodiments of the use, the C0₂-containing gas stream is a flue gas stream.

In some embodiments of the use, the composition comprises 15 to 25 wt.%

AMP.

According to some embodiments of the use, the released C0₂ has an elevated pressure, such as about 6 bars. The method and use described above allows for waste heat to be utilized in the step of regenerating AMP, lowering of the overall heat demand for regeneration of AMP and release of C0₂, and decrease of the energy demand in the subsequent compression and liquefaction of C0₂. The method and use described above provides an improved C0₂ capture technology with great potential to reduce the energy penalty of the C0₂ capture process compared to current post-combustion capture technologies.

The method and use as described above utilize the specific organic solvents NMP and TEGDME instead of water as is used in the prior art. Without wishing to be bound by theory, the inventors believe that in NMP and TEGDME, with little or no water present, the hindered amine AMP cannot react further to produce bicarbonate. Although this limits the loading to 0.5 mole C0₂/mole AMP due to stoichiometry, the carbamate formed is quite unstable which is advantageous during regeneration of AMP. Preferably, the carbamate formed will not be dissolved in the solution but will form a precipitate that may be separated from the solution. Regeneration of AMP with concomitant release of C0₂ can then be achieved using less solvent or at occasions preferred to the skilled person. The inventors further believe the NMP and TEGDME solvents offer two significant advantages over aqueous amine solutions. Firstly, the absorption rate is at least five times faster in NMP and TEGDME due to the higher solubility of C0₂. Secondly, an insoluble and thermodynamically unstable product, i.e. a carbamate complex, is formed.

Without wishing to be bound by theory again, the inventors are also of the opinion that not only does the sterically hindered amine AMP provides twice the equilibrium C0₂-loading capacity (like the tertiary amine MDEA) compared to primary or secondary amines, but that it also has a much faster reaction rate than MDEA.

Further, advantageous features of various embodiments of the invention are defined in the dependent claims and within the detailed description below.

Brief description of the drawings

Figure 1 shows a schematic of the method of removing C0₂ from a gas stream according to a preferred embodiment of the invention. Figure 2 shows two images of the crystallized C0₂ product obtained by contacting the AMP and NMP containing composition with a C0₂ containing gas stream according to a preferred embodiment of the invention. The picture on the left is a microscopic image at 100^x zoom while the right image is a microscopic image at 400^x zoom.

Figure 3 shows a CC-report in integration mode with custom baseline chosen as used in the C0₂ absorption measurements described further below.

Figure 4 illustrates a graph of solubility of C0₂ in the pure physical solvent NMP according to a preferred embodiment of the invention.

Figure 5 shows a graph of solubility of C0₂ in mixtures of physical and chemical solvents (NMP+AMP). Key: S18 = NMP, R17 = AMP.

Figure 6 shows a graph of solubility of C0₂ in mixtures of physical and chemical solvents (NMP+AMP) with quadratic polynomial regressions. Key: S18 = NMP, R17 = AMP.

Figure 7 shows a graph of solubility of C0₂ in mixtures of physical and chemical solvents (NMP + Specified amine), 25 °C, solid lines. Key: R17 = AMP. The article data (MEA & DEA) from Murrieta-Guevara and Rodriguez 1984 is shown in dashed lines.

Figure 8 shows a graph of solubility of C0₂ in mixtures of physical and chemical solvents (NMP + Specified amine), 50 °C, solid lines. Key: R17 = AMP. The article data (MEA & DEA) from Murrieta-Guevara and Rodriguez 1984 is shown dashed lines.

Figure 9 shows a graph of solubility of C0₂ in mixtures of physical and chemical solvents (NMP + Specified amine), 25 °C. Key: R17 = AMP. The article data (MEA & DEA) is from Murrieta-Guevara, Rebolledo-Libreros, Trejo 1992.

Figure 10 shows graph of solubility of C0₂ in mixtures of physical and chemical solvents (NMP + Specified amine), 50 °C. Key: R17 = AMP. The article data (MEA & DEA) is from Murrieta-Guevara, Rebolledo-Libreros, Trejo 1992.

Figure 11 depicts a partial run of true heat flow measurement during C0₂ absorption, 25 wt% AMP 25 °C. Figure 12 depicts a complete run of true heat flow measurement during C0₂ absorption, 25 wt% AMP 25 °C.

Figure 13 shows a graph of solubility in C0₂ in 15 wt% AMP at 50 °C.

Figure 14 shows a graph of data extracted from the early version of the ProFind script, pure NMP at 50 °C.

Figure 15 shows a graph of data extracted from the latest version of the ProFind script, pure NMP at 50 °C.

Figure 16 shows a graph of data extracted from the ProFind script, pure NMP at 25 °C. Notice the difference between pressure point 1 & 2 (circles) vs. 3 & 4 (stars).

Figure 17 depicts an example of how the three parameters studied (pressure, true heat flow, mass flow meter) looked in the program CCReport for experiment no. 3.3 described below in relation to absorption of C0₂ by a composition comprising AMP and TEGDME.

Figure 18 depicts part of a run in experiment no. 4.1 described below described below in relation to absorption of C0₂ by a composition comprising AMP and TEGDME. Figure 18 shows the boundaries for the integral calculations being made in the program CCReport. The horizontal bold black line is the baseline while the two vertical bold black lines are the boundaries for the integral.

Figure 19 shows a graph of solubility of C0₂ in the pure solvent (100wt%) TEGDME at 25 °C. The dashed bold lines with circles represent the article from Henni (2005) and the article from Sciamanna and Lynn (1988). Bold lines with triangles and squares represent the experimental data according to a preferred embodiment of the invention.

Figure 20 shows a graph of solubility of C0₂ in the pure solvent (100wt%) TEGDME at 50 °C. The dashed bold lines with circles represent the article from Henni (2005) and the article from Sciamanna and Lynn (1988). Bold lines with triangles and squares represent the experimental data according to a preferred embodiment of the invention.

Figure 21 shows a graph of solubility of C0₂ in the pure solvent (100wt%) TEGDME at 25 °C and 50 °C. The dashed thin lines represent the article from Henni (2005) and the article from Sciamanna and Lynn (1988) and the article from Kodama et al. (2011). Bold dashed lines represent the experimental data whilst bold lines represent the experimental data at 50 °C according to a preferred embodiment of the invention.

Figure 22 shows a graph of solubility of C0₂ in 15wt% AMP at both experimental temperatures (i.e. 25 °C and 50 °C). The squares represent the data at 25 °C and the triangles represent the data at 50 °C.

Figure 23 shows a graph of solubility of C0₂ in 15wt% AMP at both experimental temperatures (i.e. 25 °C and 50 °C) with cubic regression. The bold lines with squares represent the data at 25 °C and the dashed bold lines with triangles represent the data at 50 °C.

Figure 24 shows a graph of solubility of C0₂ in 25wt% AMP at both experimental temperatures (i.e. 25 °C and 50 °C). The squares represent the data at 25 °C and the triangles represent the data at 50 °C.

Figure 25 shows a graph of solubility of C0₂ in 25wt% AMP at both experimental temperatures (i.e. 25 °C and 50 °C) with cubic regression. The bold lines with squares represent the data at 25°C and the dashed bold lines with triangles represent the data at 50 °C.

Figure 26 shows a graph of solubility of C0₂ in different amounts of the amine, AMP, at 25 °C. The squares represent the data with 15wt% AMP and the rhombs represent the data with 25wt% AMP.

Figure 27 shows a graph of solubility of C0₂ in different amounts of the amine, AMP, at 50 °C with cubic regression. The squares with bold lines represent the data with 15wt% AMP and the rhombs with bold dashed lines represent the data with 25wt% AMP.

Figure 28 shows a graph of solubility of C0₂ in different amounts of the amine,

AMP, at 50 °C. The triangles represent the data with 15wt% AMP and the circles represent the data with 25wt% AMP.

Figure 29 shows a graph of solubility of C0₂ in different amounts of the amine,

AMP, at 50 °C with cubic regression for 15wt% AMP and for 25wt% AMP. The circles with bold lines represent the data with 15wt% AMP and the triangles with bold dashed lines represent the data with 25wt% AMP. Figure 30 shows a graph of solubility of C0₂ in different amounts of the amine, AMP, at 25 °C and 50 °C with cubic regression. The circles with bold lines represent the data with 15wt% AMP at 50 °C, the triangles with bold dashed lines represent the data with 25wt% AMP at 50 °C, the squares with bold lines represent the data with 15wt% AMP at 25 °C, the rhombs with bold dashed lines represent the data with 25wt% AMP at 25 °C. Key: R17 = AMP, S10 = TEGDME.

Figure 31 shows a graph of solubility comparison of C0₂ in mixtures of the amine AMP (average for each wt% of AMP) with TEGDME, with MEA and DEA at 25 °C. The squares with bold lines represent the data with 15wt% AMP, the lines with rhombs represent the data with 25wt% AMP. Article data from Murrieta-Guevara and Rodriguez (1984) for MEA are the dotted lines and the data for DEA are the dashed lines. All lines are cubic regressions. Key: R17 = AMP.

Figure 32 shows a graph of solubility comparison of C0₂ in mixtures of the amine, AMP (average for each wt% of AMP) and TEGDME, at 50 °C. The triangles with bold lines represent the data with 15wt% AMP, the lines with circles represent the data with 25wt% AMP. Article data from Murrieta-Guevara and Rodriguez (1984) for MEA are the dotted lines and the data for DEA are the dashed lines. All lines are cubic regressions. Key: R17 = AMP.

Figure 33 is a graph showing the differential heat of absorption as a function of C0₂ loading for 15wt% AMP at different temperatures. Experimental results at 25 °C are represented by the squares with bold lines, the experimental results at 50 °C are represented by the triangles with dashed bold lines. The line at 0,5 mol C0₂/mol amine represents the maximum C0₂ loading according to the reaction stoichiometry. All the lines follow the experimental data.

Figure 34 is a graph showing the integral heat of absorption as a function of

C0₂ loading for 15wt% AMP at different temperatures. Experimental results at 25 °C are represented by the squares with bold lines, the experimental results at 50 °C are represented by the triangles with dashed bold lines. The line at 0,5 mol C0₂/mol amine represents the maximum C0₂ loading according to the reaction stoichiometry. All the lines are cubic regressions of the experimental data. Figure 35 is a graph showing the differential heat of absorption as a function of C0₂ loading for 25wt% AMP at different temperatures. Experimental results at 25 °C are represented by the rhombs with bold lines, the experimental results at 50 °C are represented by the circles with dashed bold lines. The line at 0,5 mol C0₂/mol amine represents the maximum C0₂ loading according to the reaction stoichiometry. All the lines follow the experimental data.

Figure 36 is a graph showing the integral heat of absorption as a function of C0₂ loading for 25wt% AMP at different temperatures. Experimental results at 25 °C are represented by the rhombs with bold lines, the experimental results at 50 °C are represented by the circles with dashed bold lines. The line at 0,5 mol C0₂/mol amine represents the maximum C0₂ loading according to the reaction stoichiometry. All the lines follow the experimental data.

Figure 37 is a graph showing the partial pressures as a function of C0₂ loading for 15wt% AMP at different temperature. Experimental results at 25 °C are represented by the squares with bold lines, the experimental results at 50 °C are represented by the triangles with dashed bold lines. The line at 0,5 mol C0₂/mol amine represents the maximum C0₂ loading according to the reaction stoichiometry. All the lines are cubic regressions of the experimental data.

Figure 38 is a graph showing the partial pressures as a function of C0₂ loading for 25wt% AMP at different temperature. Experimental results at 25 °C are represented by the rhombs with bold lines, the experimental results at 50 °C are represented by the circles with dashed bold lines. The line at 0,5 mol C0₂/mol amine represents the maximum C0₂ loading according to the reaction stoichiometry. All the lines are cubic regressions of the experimental data.

Detailed description of preferred embodiments

General remarks

Although the present invention has been described above with reference to specific illustrative embodiments, it is not intended to be limited to the specific form set forth herein. Any combination of the above mentioned embodiments should be appreciated as being within the scope of the invention. Rather, the invention is limited only by the accompanying claims and other embodiments than the specific above are equally possible within the scope of these appended claims.

In the claims, the term "comprises/comprising" does not exclude the presence of other species or steps. Additionally, although individual features may be included in different claims, these may possibly advantageously be combined, and the inclusion in different claims does not imply that a combination of features is not feasible and/or advantageous. In addition, singular references do not exclude a plurality. The terms "a", "an", "first", "second" etc. do not preclude a plurality.

In the claims, the term "about" is to be interpreted as "approximately" and implies that not only is the exact value or range (e.g. temperature) recited after "about" included but also values slightly lower or higher than the recited value or range.

Method and Use

A schematic of the method according to one embodiment of the invention will now be described with reference to Figure 1. In the method a C0₂ containing gas stream 100 is fed into an absorbing unit 102 wherein the gas stream is contacted with a composition 104 that is also fed into the absorber 102. The composition 104 contains AMP and NMP or the composition 104 contains AMP and TEGDME. Carbon dioxide in the gas stream 100 is absorbed by the composition 104 to obtain a spent slurry of carbamate. A gas stream 101 with reduced C0₂ content passes out from the absorber 102. In fact, gas stream 101 may have up to 80% reduced C0₂ content. The slurry 106 is transferred to a mechanical separator 108 to separate most of the slurry 106 from the unreacted composition 104 which is then recycled back to the absorber 102. The concentrated slurry 110 is sent to a heater 112 to decompose the unstable concentrated slurry 110 to C0₂ 114 and AMP 116. Carbon dioxide 114 is obtained at elevated pressure from the heater 112 and AMP 116 is recycled back for use in the absorber. Only modest temperature levels are required to reversibly decompose the solids in heater 112 to lean carbon dioxide 114 and AMP 116. For instance, a lean CO₂ stream 114 can be received at about 7 bars if heating up to 90 °C. However, sufficient regeneration of CO₂ 114 and AMP 116 can be done at, for instance, 70 °C. Figure 2 illustrates an example of the crystallized C0₂ product obtained by contacting the AMP and NMP containing composition with a C0₂ containing gas stream. The product in Figure 2 has a clear crystal structure in the form of a cuboid with pointy and/or scalpel edges. Without wishing to be bound by theory, the inventors believe C0₂ in the gas stream reacts with the composition by the following reaction mechanism. One molecule of AMP reacts with a single C0₂ molecule to form a carbamate which then reacts with a second molecule of AMP to form a carbamate anion and a protonated cation:

RNH₂ + C0₂→ RNHCOOH

RNHCOOH + RNH₂ RNHCOO^' + RNH

The mechanism is sensitive to water. In the presence of water the carbamate reacts with water to form bicarbonate. The reaction between a carbamate ion and water may occur as follows:

RNHCOO- + H₂0 ≠ RNH₂ + HCO^

It is important that little to no water is present in the described method and use to ensure no bicarbonate is generated.

Experimental

The following examples are mere examples and should by no mean be interpreted to limit the scope of the invention. Rather, the invention is limited only by the accompanying claims.

General Procedures

All commercial reagents were used as received. All experiments were conducted with a Chemical Process Analyzer (CPA 202) from ChemiSens AB, Sweden. CPA202 is a factory calibrated reaction calorimeter measuring true heat flow with a stable baseline (no reaction zero baseline). The CPA202 reactor has an effective volume of 250 cm³ and is constructed out of stainless steel with glass side windows. A CPA202 reactor of glass and stainless steel was filled with the specific reactants and used in the experiments. During the experiment the glass windows allow the user to see changes to the solutions physical properties during an experiment e. g. color change, precipitation or foaming. The CPA202 is connected to the universal dosing controller (VCR) and a controlling unit that together regulate the experiments. An automated script written in ProFind Automation is used in the experiments. ChemiCall v2 is a software developed by ChemiSens AB used to monitor and control the experimental equipment. Data and graphs are extracted from the program and integrated in CC-report to be able to conduct all calculations, mainly integrations.

In general, two different ProFind Automation scripts were created with some help from ChemiSens in order to automate the process as much as possible to reduce human error. The main experiments used script number 1 which seeks stability in the True Heat Flow (THF) between each step and script number 2 seeks stability in Pressure and THF between each step to better study the absorption abilities when the system moves to equilibrium.

Abbreviations

The following abbreviations are used

AMP 2-Amino-2-methyl-l-propanol

C0₂ Carbon dioxide

DEA Diethanolamine

MDEA N-Methyldiethanolamine

MEA Monoethanolamine

NMP N-methyl-2-pyrrolidone

PZ Piperazine

TEGDME Triethylene glycol dimethyl ether

Absorption of CO? by a composition comprising AMP and NMP

Each experiment was conducted with 100 grams of total sample with different amounts of the included solvents:

1. 100 wt% NMP

2. 85 wt% NMP and 15 wt% AMP

3. 75 wt% NMP and 25 wt% AMP Samples were weighted on a Sartorius scale, type 1574 with a maximum weight of 4200.0 g. The accuracy of measurement according to the scale was 0.05 for weight of 2000 g and 0.1 for weight of 4200 g. All samples were measured and filled in the reactor on the scale first with NMP to 85.00 g or 75.00 g then added up with AMP to lOOg.

For an overview of the process see Table 1 below. Table 2 and Table 3 clarify the variables and constants included in the ProFind automation. A summarized and detailed description of the ProFind automation scripts is now presented.

The operator alters three factors before the start of the run which are:

1. Final_pressure_level - At what pressure is the experiment conducted? Be sure to add 1 because the measurements start from vacuum at -1 bar (0 bars absolute), (e.g. 3.5 bar)

2. Stepping - How large shall the pressure steps be? (e.g. 0.5 bar)

3. Temperature - At what temperature is the experiment conducted? (e.g. 25 °C)

First off the logging of all available parameters is started. Then a variable scope is created which holds all the different variables used in the ProFind Automation. A variable scope is a facility used to calculate new variables and they are only available inside the scope. This automation has two variable scopes (User set parameters, and Background calculations). Inside the background calculation scope there are two repeaters, notified in the table with dashed lines. Reaching dosing level is the deepest repeater which is responsible for controlling and dosing the pressure in small steps on a total of 0.5 bars. This is then repeated in the outer repeater Stepping from vacuum up to 2.5 bars over pressure (3.5 bars absolute).

Step by step explanation to the ProFind automation script:

1. Start log - the logging of the data is started

2. Experiment temperature - sets the temperature in the experiment

3. Stirrer speed - sets the stirrer speed in the experiment 4. Condition Set [Wait for signal stability in THF]- in this case a Signal Stability test of THF which means that the system awaits a stable signal of THF according to different parameters

5. Set initial dosing level [0] - zeroing the mass flow input signal to be sure that no mass flow is started

6. Signal Control [Pump_MS3+(Stepping/100)] - Adds a specified number to the open position Pump_MS3 which is used by the program to know where in the experiment it is.

7. Valve Control [Open] - opens the valve

8. Dosing Control [CTRL] - pressure stepping in small intervals to a total of 0.5 bar

9. Valve Control [Close] - close the valve

10. Condition Set [Wait for signal stability in THF, or Pressure and THF depending on the used automation script] - in this case a Signal Stability test of THF which means that the system awaits a stable signal of THF according to different parameters. The Signal Stability test awaits a stable signal of Pressure and THF in some of the experiments to better grasp the kinetics of the system.

11. Dosing control [0] - zeroing the mass flow

12. New Temperature [25 °C] - set the temperature to 25 °C, more or less only needed when the initial temperature is 50 °C in order to cool the system.

13. Condition Set [Temperature below 25.1 °C] - Condition met when the reactor temperature is lower than 25.1 °C

14. Inactive - inactivates the thermal mode on the reactor

15. Stirrer speed [0 rpm] - stops the stirring

16. Stop log - the logging of data is stopped.

Table 1. ProFind Automation. Dashed lines indicates repeater.

Start log

User set parameters

Experiment temperature [25 °C or 50 °C]

Stirrer Speed [300 rpm]

Condition Set [Wait for signal stability in THF] Set initial dosing level [0]

1 Background calculations J

Stepping

Signal Control [Pump_MS3+(Stepping/100)] -

Timer 2 s

Valve Control [Open]

Timer 2 s

Reaching dosing level -

- Dosing Control [CTRL] -

-

Timer 2 s

Valve Control [Close]

Dosing Control [0] -

- Condition Set [Wait for signal stability in THF, or Pressure and THF depending - on the used automation script]^ *

New Temperature [25 °C]

Condition set [Temperature below 25.1 °C]

Inactive

Stirrer speed [0 rpm]

Stop log

Table 2. User set parameters

User set parameters

Name Expression Explanation

Initial dosing level 1-1 To be sure 0 = zero

Final pressure in bars, starts from vacuum at -1

Final pressure level 3.5 bar (0 bars absolute)

For the experiments conducted with focus on the kinetics the Final_pressure level where set to 5

Stepping in bar

Stepping 0.5 The stepping was set to 1 for some of the experiments in order reach higher pressures faster

Weight in grams. Not included in any ProFind-

Loaded amount 100

calculations. Included only in the log

Table 3. Background calculations

Background calculations

Name Expression Explanation

The experiment

Signal compare, Experiment_stop is lower

Stepping stops when this than 0.0001

condition is met

Pressure set val Place for

Pump_MS3- 100

ue calculation

Pressure in the

P Pressure A+l-NMP vap_pres reactor, starts with vacuum

Signal Compare, Break var lower than All of the

Reaching dosing specified following level conditions must be

Timer 150 s true to continue

Regulator specific

P reg 20

constant

Algorithm for stepping with

CTRL P reg^■ (Pressure_set_value_delayed-P)

pressure in the process

Stepping stop when

Break var Pressure_set_value-P

expression is met The experiment

Experiment stop Final_pressure_level - Pressure set value ends when this expression is met

STOP 1-1 To be sure 0 = zero

Access a precious

Pressure set val Previous(Pressure_set_value_delayed,0)^■ value. 99% of the ue delayed 0.99 + 0.01 ^■ Pressure set value previous and 1% of the current

/ 10 Cl 4290.3 \

( 10 214.11+Reactor_Temperature \ NMPs vapour

NMP vap pres pressure (Whitfield

^■ 0.0013328 et al, 1999)

CC-report is the program developed by ChemiSens AB used for the calculations of the data extracted from the experiments. A custom baseline is used when calculating the integrals of mass flow to reduce the effects of an offset in the measurements. See Figure 3. In the true heat flow case the custom baseline is used to minimize the influence of unwanted heat sources like the stirring. The two vertical black lines show the boundaries of the integral and the horizontal black line indicates the set baseline.

Equations Henry 's law

Henry's constant, J-C_{1 L} is calculated from a linear regression between partial pressure and the molar fraction or between partial pressure and concentration. In physical absorption, the concentration of a dissolved gas in a solvent increases in proportion to the partial pressure of the gas, (Equation 1.

Pi = Wi,L ' ^xi (Equation 1 for gaseous component i in solvent L. However in chemical absorption Henry's law does not hold because the chemical absorption isotherm is nonlinear. (Schlauer, 2008)

Because of this the Henry's constants are calculated for the pure physical solvent and then used for the mixtures with the sterically hindered amine to give an approximated heat of crystallization value. This makes it possible to compare the different systems to each other and to previously conducted article data.

Heat of solution

Change of enthalpy per mole when a substance dissolves to form a very dilute solution is called the enthalpy of solution. The dissolution can be both exothermic ( Δ H_sol < 0) and endothermic ( Δ H_sol > 0) (Jones, Atkins, 2000).

The heat of solution is calculated with |Equation 2, (Hans T. Karlsson, 2008). rtend f(_HeatFi_ow) dHeatFlow - V_R■ P_co, [Equation 2 u tstart ²

^sol

The absorbed amount of carbon dioxide was calculated according to (Equation 3, (Hans T. Karlsson, 2008). rtend /-(_Mass /_ow) dM ass Flow■ B, p_rn ■ V_D

n = ^tstart . ^c°2 (Equation 3

Mco₇ R^■ T

V_R = Reactor head space volume (Pure NMP: 160.6^■ 10^"6 m³

AMP 15 wt%: 159.2^■ 10^"6 m³, AMP 25 wt%: 158.2^■ 10^"6 m³)

B₁ = Mass flow conversion factor (0.0050 g^■ V^"1■ s^"1)

Pco₂ ⁼ Partial pressure of C02 (Pa)

M_Co₂ = Molar weight of C02 (44.01 g^■ mole^"1)

R = Gas constant (8.3145 m^3■ Pa^■ K^"1■ mole^"1)

T = Reactor temperature (K) n_s = Absorbed amount of C0₂ (mole)

Heat of crystallization

The heat of crystallization is calculated with |Equation 4 and (Equation 5, (Hans T. Karlsson, 2008). n_R = n_s — (Equation 4

H_A = Henry' s constant for the solubility of C0₂ in solvent (Pa^■ m^3■ mole^"1)

V_s = Volume of solvent in the sample (m³) n_R = Amount of C0₂ reacted with the amine (mole)

AH_cry (Equation 5 rtend f(_Heatpi_ow dHeatFlow - n_s■ AH_sol - V_R■ P_CO

tstart ²

2 K_A rtend f(_HeatFi_ow dHeatFlow - n_s■ AH_sol - V_R■ P_C07

t start ²

Results

Henry 's constants

In Figure 4 the solubility of C0₂ in pure physical solvent (NMP) is presented. The figure clearly indicates the difference between the scripts. The lines with squares and circles are results from the early version of the script and the lines with stars and triangles are results from the latest version of the script. The results from the new version shown as lines with in diamonds and crosses have a much closer fit to the article data from (Murrieta-Guevara and Rodriguez 1984).

In Table 4 below the results (slopes) from the linear regressions of the data in Figure 4 are presented. The slopes are the calculated Henry's constants. The NMP equilibrium constants are used in the heat of crystallization calculations for both the pure and mixed solvents.

Table 4. Henry's constants from the different scripts together with reference data from several articles. The NMP equilibrium data is from the latest version of the script.

Experimental data

In Figure 5 the solubility of C0₂ in some of the conducted experiments of mixtures of physical and chemical solvents (NMP+AMP) is plotted for all the mixtures and temperatures studied. In Figure 6 the same systems as in Figure 5 are plotted but with added quadratic polynomial regressions. Note that for AMP 25 wt% 25 °C the slope of the quadratic polynomial regression is approximated from AMP 25 wt% 50 °C in order to make it comparable due to the fact that only the four first equilibrium points were generated in the automation script (this in turn is because the latest version of the automation script needs more time to stabilize between pulses).

Comparison between experimental data and article data

Figure 7 compares the experimental data at 25 °C (solid lines) with article data (dashed lines) from Murrieta-Guevara and Rodriguez (1984). Note that the same physical solvent has been used but with different chemical solvents. The studied system (NMP+AMP) is comparable with the article data at 25 °C which make it a good candidate for C0₂ absorption in general.

Figure 8 compares the experimental data at 50°C (solid lines) with article data (dashed lines) from Murrieta-Guevara and Rodriguez (1984). Note that the same physical solvent has been used but with different chemical solvents. The studied system (NMP+AMP) is comparable with the article data at 50 °C which make it a good candidate for C0₂ absorption in general.

Figure 9 compares the experimental data at 25 °C with article data from Murrieta-Guevara, Rebolledo-Libreros, Trejo (1992), note that the same physical solvent has been used but with different chemical solvents. Two types of estimations, quadratic polynomial (solid lines) and linear (dashed lines), have been added in order to see possible high pressure effects. Even at higher pressure the studied system (NMP+AMP) is somewhat comparable with the article data at 25 °C and it is still a good candidate for C0₂ absorption in general.

Figure 10 compares the experimental data at 50 °C with article data from Murrieta-Guevara, Rebolledo-Libreros, Trejo (1992). Note that the same physical solvent has been used but with different chemical solvents. Two types of estimations, quadratic polynomial (solid lines) and linear (dashed lines), have been added in order to see possible high pressure effects. Even at higher pressure the studied system (NMP+AMP) is somewhat comparable with the article data at 50 °C and it is still a good candidate for C0₂ absorption in general.

Heat of absorption (AH_soi) and Heat of crystallization (AH_cry)

The results from the heat of absorption calculations are presented in Table 5 below. The runs have been included in the table together with a final mole ratio of C0₂ and amine in order to compare these with article data. For AMP 15 wt% at 25 °C the pressure steps were set to 1 instead of 0.5.

The a -value depends on how long the experiments have been conducted and with which script. Overall the heat of solution is fairly equal when comparing two separate runs, under the same conditions, while the heat of crystallization differs more depending on the difficulties in modeling the crystallization. The experimental heat of solutions is closets to the article data for 15 wt% MEA.

Table 5. Heat of absorption, where is the final mole ratio of C0₂/ Amine in the liquid phase

Experimental Data Article Data, Murrieta-Guevara (1992)

AHsol AHsol AHsol

[kJ/mole] [kJ/mole] [kJ/mole]

T AHsol AHcry

Compound NMP + NMP + NMP +

[^°C] [kJ/mole] [kJ/mole] 15 15 30 wt% wt% wt%

MEA DEA DEA

NMP 25 -10.1 - - - - - -

NMP 50 -11.6 - - - - - -

NMP,

25 -12.9 - - - - - - Equilibrium

NMP,

50 -14.9 - - - - - - Equilibrium AMP 15

25 -82.0 336 0.77 -26.9 - - 0.8 wt%, 1st

AMP 15

wt%, EQ

(Larger 25 -62.2 221 0.98 -17.9 - - 1.0 pressure

steps)

AMP 15

50 -69.1 257 0.54 -67.5 -42.3 -40.3 0.5 wt%, 1st

AMP 15

50 -84.8 306 0.59 -49.4 -35.8 -31.7 0.6 wt%, 2nd

AMP 25

25 -67.9 300 0.47 -67.5 -42.3 -40.3 0.5 wt%, 1st

AMP 25

25 -70.9 158 0.43 -89.1 -50.6 -53.1 0.4 wt%, EQ

AMP 25

50 -104 272 0.53 -67.5 -42.3 -40.3 0.5 wt%, 1st

AMP 25

50 -93.0 255 0.54 -67.5 -42.3 -40.3 0.5 wt%, 2nd

Precipitation

When the carbon dioxide reacts with the AMP amine at low temperatures, 25 °C, a white carbamate precipitation is formed. Fainter forms of precipitation are formed even at higher temperatures when a larger amount of AMP is present. Strong precipitations were formed in the mixture with 15 wt% AMP and 25 wt% AMP at 25 °C and fainter precipitations were sometimes formed at 50 °C. The precipitation formed in 25 wt% AMP at 25 °C was not as viscous as the one at 50 °C. Both the strong and faint precipitations were easily dissolved in water.

When pure AMP reacts with water a barely visible precipitation is formed, however it is clearly different from the more viscous carbon dioxide precipitation. In the faint precipitation there are some traces of a stronger precipitation on the bottom of the reactor due to the sedimentation effects. The stronger precipitations somewhat reduces the stirring speed depending on the viscosity of the precipitation. The strongest precipitations results in a coating of the reactor walls which no longer participates in the stirring. However, even with the strongest precipitation the stirring is still functional. The precipitation data is presented in Table 6 below.

Table 6. Precipitation formed in mixtures between NMP and AMP. Amount of AMP noted in columns. Scale from 0 to 4 describes the type of precipitation in addition to the text.

By studying the data from the experiments above it is possible to somewhat determine how much carbon dioxide have entered and absorbed in the solution just before the precipitation is formed. The time at which the precipitations form have not been possible to characterize exactly but it has been noted in the experiments that the true heat flow breaks its pattern when the precipitation is formed. By studying when the true heat flow notably differs from its pattern some interesting data could be extracted. This is depicted in Figure 11 (part of run) and Figure 12 (complete run) where the true heat flow breaks its pattern inside the green square. The solid black line shows the pressure in bars, the solid blue line shows the mass flow in volts and the red dashed line shows the true heat flow in watts. Table 7 shows intervals of the total amount of moles in the different fractions and the mole fraction just before the precipitation formation, ηχ is the total amount of moles in both the gas phase (no) and the solution phase (ns). The differences between the runs may depend on that the system is supersaturated before the precipitation forms 5 or that trace amount of impurities may start the crystallization earlier stage.

Table 7. Total amount of moles in the different fractions and the mole fraction, just before the precipitation formation, shown in intervals.

Figure 13 shows the system with 15 wt% AMP at 50 °C, precipitation is 0 formed for the experiment marked with stars between points 5 and 6 at approximately 0.225 MPa while no precipitation is formed in the experiment marked with squares. When the precipitation is formed it no longer takes place in the reaction which should make is possible for the next amine to react with new C0₂ or available C0₂-complex, R HCOOH, to form more precipitation until the remaining amine have been consumed. 5 ProFind automation script

In Figure 14, data extracted from an early version of the ProFind script is shown. The solid black line shows the pressure in bars, the solid blue line shows the mass flow in volts and the red dashed line shows the true heat flow in watts. This script seeks only stability in true heat flow between each pulse. The pressure does not have 0 enough time to reach equilibrium and more carbon dioxide could possibly be absorbed.

Compare this to Figure 15 where the script seeks stability in both pressure (squares) and true heat flow. This takes longer time between each pulse but ensure that equilibrium has been reached. The squares show that the pressure is not stable between each pulse.

The circles in Figure 14 represent pulse errors because the script for a short period of time believe that it has reached the specified break variable, no such errors are to be found in the latest version of the script, Figure 15, due to an added timer in each block of the script. As mentioned earlier there is great difference between the original script and the final version which can be seen in the Henry's constant graphs, Figure 4, where the triangle and star marks data from the latest version of the script, closer to the article data than the early version of the script. In Figure 15, the squares indicate point where pressure has reached equilibrium. This script also seeks stability in true heat flow between each pulse.

Figure 16 depicts that there is a leakage in the external valve because the pressure keep increasing when the external valve is closed but the valve on the mass flow controller is open (circles). This problem is solved by closing both valves in an earlier stage (stars). This is included in the latest version of the script.

Experimental data for the solubility of C02 in pure and mixed solvents

Table 8. Experimental data for the solubility of C02 in pure and mixed solvents, mole fraction, partial pressure, concentration.

xco₂ ^pco₂ ^Cco₂ ^xco₂ ^pco₂ ^Cco₂ rfemoiei rfemoiei

[MPa] [MPa]

L m³ J L m³ J

NMP, 25 °C NMP, 50 °<

0.00155 O.0295 Ο.ΟΙ60 Ο.ΟΟ295 Ο.Ο556 0.0307

0.00579 O.0683 O.0604 O.OO581 O.0831 0.0605

0.0133 0.II8 O.139 0.0105 O.125 0.110

0.0203 O.168 O.215 0.0153 O.174 0.161

0.0277 O.219 Ο.295 O.0250 Ο.275 0.266

0.0352 O.271 Ο.378 O.0341 Ο.375 0.366 0.0424 0.321 0.459

NMP, 25 °C, Equilibrium NMP, 50 °C, Equilibrium

0.00290 0.0230 0.0302 0.00219 O.0219 O.0227

0.00837 0.0597 0.0875 0.00585 O.0566 O.0609

0.0158 0.105 0.167 0.0108 O.105 0.113

0.0239 0.156 0.253 0.0160 O.155 O.168

0.0213 O.207 Ο.226

NMP + 15 wt% AMP, 25 °C NMP + 15 wt% AM ^[P, 50 c

O.0291 0.00755 0.311 0.0154 O.0242 0.162

O.0764 0.0921 0.858 O.0372 O.0680 0.401

O.0806 0.123 0.910 O.0496 0.113 0.542

O.0887 0.183 1.011 O.0593 O.163 0.654

O.0946 0.233 1.085 O.0635 O.204 0.704

0.102 0.298 1.178 O.0710 O.261 0.793

O.IO6 0.333 1.227 O.0772 O.313 0.868

0.112 0.383 1-305 O.0820 Ο.358 0.927

NMP + 25 wt% AMP, 25 °C, Equilibrium NMP + 25 wt% AM ^[P, 50 c

0.00825 0.00906 0.0864 O.0257 0.0170 0.275

0.0878 0.0147 1.000 O.0565 O.0472 0.622

0.0958 0.0336 1.101 O.0845 O.0585 0.958

0.1034 0.0778 1.198 O.107 O.169 1.243

O.112 O.220 1-315 0.117 Ο.273 I.382

0.122 Ο.325 I.447

O.I27 Ο.357 I.5IO

Experimental data for heat of solution ( A H_sol) and heat of crystallization ( AH_cry) in pure and mixed solvents

Table 9. Experimental data for heat of solution and heat of crystallization in pure and mixed solvents.

Compound T THF V_r P COzJinal H ^Hsol

[m³]

[m³] Pa^■ m³ r kj 1 r kj 1

[ ° ] [Ws] [MPa] [mole]

^•lO⁵ mole mole. mole]

10⁶

O.044

NMP 9-7

25 503 161 O.321 727 -10.1 -

7 4

NMP O.035 9-7

50 474 161 Ο.375 1050 -11.6 - 6 4

NMP,

0.024

Equilibriu 9-7

25 342 161 O.156 626 -12.9 -

111 7 4

NMP,

0.022

Equilibriu 9-7

50 361 161 O.207 920 -14.9 - O

111 4

AMP 15 9.8

9750 159 Ο.383 O.II8 626 -82 336

Wt%, 1^st 8

AMP 15

wt%, EQ, 25

1030 9.8

Larger 159 Ο.457 O.I65 626 -62.2 221

0 8

pressure

steps

AMP 15 Ο.Ο83 9.8

50 5800 159 Ο.358 920 -69 257

Wt%, 1^st 1 8 AMP 15 0.089 9.8

76ΟΟ 159 0.372 920 -84.8 306

Wt%, 2^nd 4 8

AMP 25 9.9

819O 158 0.412 0.120 626 -67.9 301

Wt%, 1^st 8

25

AMP 25 9.9

849Ο 158 0.0778 0.120 626 -70.9 158 wt%, EQ 8

AMP 25 I55O 9.9

158 0.319 0.149 920 -104 272

Wt%, 1^st 0 8

50

AMP 25 I33O 9.9

158 0.357 0.143 920 -93 255

Wt%, 2^nd 0 8

Discussion

The fact that the studied system is comparable with article data in other commonly used amine systems ensure that NMP+AMP can be used to successfully absorb C0₂.

Although the system did not reach the absorption possibilities of the commonly used MEA-system , the plausible low regeneration temperature of AMP could result in a more cost effective recirculation favoring AMP.

Henry 's constants

The calculated Henry's constants differ between the ProFind scripts used. The latest version of the script results in graphs with a closer fit to the article data from Murrieta-Guevara and Rodriguez (1984) with a slightly different slope. The latest version of the script ensures that equilibrium between both pressure and true heat flow is reached and should ensure that as much carbon dioxide as possible is absorbed in the solvents. These are the Henry's constants used in the rest of the calculations.

Carbon dioxide absorption

The trend in Figure 5 is that larger amount of amine and lower temperature equals better absorption. This trend is confirmed in Figure 7 and Figure 8 when comparing experimental data to article data from Murrieta-Guevara and Rodriguez (1984). The AMP solvent does not reach the absorption level of MEA but is close to DEA at the same wt%. The added quadratic polynomial regressions fit fairly well with the experimental data in Figure 6. Looking first at Figure 9 the absorption systems probably have a different appearance than the added estimations because the quadratic estimation is too steep and the linear estimation is too flat. In Figure 10 the quadratic estimations seems more reasonable and follow the other systems fairly well. To determine how the systems act at high pressure longer experimental runs are needed to make the system reach higher pressures, such runs can be expected to last for several days.

Heat of absorption and heat of crystallization

The calculations of heat of absorption and heat of crystallization differ depending on how the custom baseline is chosen in CC-report. All the values are negative because dissolution of gases is exothermic due to the fact that heating decreases the solubility of a gas.

In Table 5 the a-value depends on how long the experiments have been conducted and with which script. For the most part the conducted experiments have a fairly equal heat of solution when comparing two separate runs under the same conditions while the heat of crystallization differs more depending on the difficulties in modeling the crystallization. The experimental heat of solutions is closest to the article data for 15 wt% MEA.

Precipitation

When the precipitation forms this will alter the mass- and energy transfer in the system which is why there is a change in true heat flow, Figure 11 and Figure 12. Table 6 shows that stronger precipitation is formed at lower temperature and with more AMP. This is confirmed in Table 7 which also states that at higher temperatures more carbon dioxide can be dissolved before the precipitation forms.

As can clearly be seen in Figure 13 the system with 15 wt% AMP at 50 °C follow each other quite well until a pressure of about 0.225 MPa where one of the graphs derails from the predicted path. It is around this point that the precipitation is formed. For the other 15 wt% AMP at 50 °C system no precipitation is formed and the graph continues with no change in the slope. The formed precipitation no longer takes place in the reaction and should make it possible for the next amine to react with new C0₂ or available C0₂-complex, R HCOOH, to form even more precipitation until the remaining amine have been consumed. No precipitation is formed in other commonly used amine systems.

The regeneration temperature depends on the amount of AMP amine added in the solution. For 15 wt% the regeneration temperature can be expected to be around > 50 °C because the runs at 50 °C resulted in no precipitation at all and only one resulted in a very faint precipitation. For 25 wt% the regeneration temperature can be expected to be > 50 °C because all experiments resulted in precipitation..

ProFind automation script

The latest version of the ProFind automation script is at a state where it can be run without any alteration during the process. It is fully automated after loading the chemicals and start parameters and there are no longer any risks for human error when the script has been launched. There are no longer any pulse errors present and consistent data can be generated in future experiments.

Absorption of CO? by a composition comprising AMP and TEGDME

The experiments used to study the absorption of C0₂ by a composition comprising AMP and TEGDME are listed in Table 10 below.

Table 10. Composition of the amine solutions used in the experiments. All solutions were prepared by weight. Key: R18 = AMP.

TEGDME R18 (wt%) Temp Experiment no. (wt%) (^°C)

100 - 25 1.1

100 - 25 1.2

100 - 50 2.1

100 - 50 2.2

85 15 25 3.1

85 15 25 3.2

85 15 25 3.3 TEGDME R18 (wt%) Temp Experiment no.

(wt%) (^°C)

85 15 50 4.1

85 15 50 4.2

75 25 25 5.1

75 25 25 5.2

75 25 50 6.1

75 25 50 6.2

100 - 25 9.1

100 - 25 9.2

100 - 50 10.1

100 - 50 10.2

In the program ChemiCall v2 both the configuration and the process to be used was chosen. The configuration is loaded and the variables in the process script are edited before each run. The variable changes are:

1. The reactor temperature was constant during the experiment, either 25 °C or

50°C.

2. The file name in which all the variables from the calorimeter are saved.

3. The final charged pressure level.

The prepared mixture for each experiment is poured into the reactor. The reactor is then lowered into the CPA202s water bath and the appropriate connections are plugged in. All the experiments are run at Isothermal mode, starting with a vacuum pressure of approximately -0,95bar.

The automated ProFind script used in the thesis is used to reduce human error. Every condition in the script has to be met in order for the script to continue. An overview of how the script works follows: 1 Start the script.

2 Experiment temperature, in this thesis 25 °C or 50°C, is set. The stirrer speed is set at 300 rpm. The pressure and true heat flow has to be stabile within the limits of 0.02 bar and 0.02W respectively.

3 Conditions met.

4 The mass flow meters valve opens, until a certain pressure is reached. In this thesis the pressure stepping was 0.5bar, starting from -0.95 bars up to the final pressure.

5 Equilibrium conditions are met.

6 Step 4-6 are repeated until the final pressure of 3 bars or 4 bars is reached.

7. The temperature is lowered if needed to 25 °C and the stirrer is stopped. The thermal mode is changed to inactive.

The process is based on the solution reaching equilibrium before each injection of C0₂. Equilibrium meaning that both pressure and heat flow in the system have to be within a maximum difference of 0.02 bars and 0.02W respectively. These conditions have to be met during a time lapse of 10 min for the pressure and 100 s for the heat flow. When all the conditions are met the valve opens and C0₂ flows into the system until the step pressure loading is reached. For the experiments conducted the end pressure was 3 bars or 4 bars with each pressure loading step being 0.5 bars. When the system has reached equilibrium after its last pressure loading step, the thermal mode is changed to Inactive at the end of the script.

When the experiment has finished and the INACTIVE sign is shown the reactor is ready to be cleaned. The air valve is opened so that the systems pressure is equalized with its surroundings. The solution and its precipitation are disposed of properly. The chemicals used in this thesis are water soluble so the reactor is washed thoroughly with deionized water and dried before the next experiment.

Save and deciphering information

CCReport is a computer software program developed by Chemisens AB which is used for analyzing and evaluating an experiment. In this thesis all the operating conditions were recorded as a function of time. Figure 17 is an example of how the three examined measurements: pressure, true heat flow and mass flow meter are shown in the program CCReport. For example, in Figure 18 the total heat of absorption for each pressure loading step was the integration of the heat flow over its time period. A horizontal custom baseline is chosen for each integral and the two vertical lines show the boundaries of the integration.

Equations used

The amount of CO₂ absorbed

The total amount of C0₂ that entered the reactor can be calculated from knowing the mass flow signal from the mass flow controller. The mass flow meter signal was integrated over time during the pressure loading step. With a conversion factor, ki, the total amount of C0₂ in the reactor could be calculated.

The amount of C0₂ in the reactors head space, in the gas phase, was calculated with the Ideal Gas Law by knowing the pressure differences in each stage as well as the volume of the reactors head space.

With the knowledge of the total amount of C0₂ that enters the reactor and the amount of C0₂ in the reactor head space, the amount of C0₂ absorbed, by the

reactant could be calculated, see Equation 1. All the calculations were performed using mass units instead of molar units due to the fact that the volume can be subject to change but not the mass. n

V_gas = Reactor head space (ml)

/ = Mass flow conversion factor 0,001659(,g/V^■ s)

P_CQ2 = Partial pressure of C0₂ (bar)

Mco₂ = Molar weight of C0₂ 44,01 (g /mol)

R = Gas constant 83,43(mZ^■ bar /K^■ mol)

Henry 's law

Henry^'s law is a gas law that states that at constant temperature, the molar fraction a soluted gas in a solution is proportional to its partial pressure over that solution. This is true for physical absorption and therefore Henry's constant is calculated only for the experiments with 100 wt% of the solvent, TEGDME. Henry's constant was calculated according to Equation 2:

P_Co₂ = H^■ x_Co₂ (Equation 2)

H = Henry's constant

= Partial pressure of C0₂ (MPa)

xco₂ ⁼ Mole fraction of C0₂

Heat of absorption

The heat transfer to and from the absorption process is also important. The highest energy costs for a plant are in the regeneration process. The energy needed in the regeneration is almost the same amount of the heat of absorption. The heat of absorption for this thesis consists of three separate heat factors:

• Heat of solution, the heat produced when C0₂ dissolves in the solvent TEGDME.

• Heat of reaction, the heat that is produced due to the reaction that takes place, the C0₂ reacts with the reactant AMP.

• Heat of crystallization, the heat that is produced if precipitations are formed.

•

The heat of absorption was calculated according to Equation 3 : ΛΗ^ = ^ "->«>-^ _(Equatlon 3)

n_C0₂abs

V_gas = Reactor head space (ml)

P_Co₂ = Partial pressure of C0₂ (bar)

= Absorbed amount of C0₂ (mol)

Results

The results in this thesis are presented both as integral and differential heats of absorption data. The reason for this is to be able to see the true impact that C0₂-loading has to the heat of absorption. The differential data gives a broader understanding of the different parameters, among them acid gas loading, that influence the heat of absorption. (I.Kim, F. Svendsen, 2011) In the program CCReport the data from the experiments was plotted as a function of time. These data is the basis for the calculations performed in this thesis. Henry 's constant

The solubility of C0₂ in pure solvent, TEGDME, is presented in Figure 19 at 25°C, and in figure 20 at 50°C. The article data used to compare the results of Henry's constant are plotted with dashed bold lines. Figure 21 shows all the experiments conducted with pure solvent, TEGDME with linear regressions. Of importance is that the article data from Henni (2005) is on average 18% higher than those of Sciamanna (1988) at 25°C, were Sciamanna used an automated procedure. Table 11 below shows the results from a linear regression of the solubility of C0₂ in pure solvent, TEGDME at 25°C and 50°C. In Figure 21 the line for Henni (2005) for 50°C, has been calculated from linear regression of the articles three temperatures (25°C, 50°C and 60°C). In Figure 21 the line for Kodama (2011) at 40°C, has been calculated from quadratic regression due to the fact that the solubility data was for higher mole fractions and pressure. From this quadratic regression a linear regression calculated Henry's constant.

Table 11. Linear regressions of the Solubilty of C0₂ in pure TEGDME at 25°C and 50°C.

R²

Ex. No 1.1 0,9856

Ex. No 1.2 0,997

Ex. No 2.1 0,9766

Ex. No 2.2 0,9389

Ex. No 9.1 0,8889

Ex. No. 9.2 0,996 Ex. No. 10.1 0,9989

Ex. No 10.2 0,9898

Table 12. Henry's constant, from linear regression (Table 11) of the experimental data in figure 21 including article data from several references.

T (°C) H (MPa)

TEGDME Ex. no. 1.1 25 3,620

this invention

TEGDME Ex. no. 1.2 25 3.402

this invention

TEGDME Ex. no. 2.1 50 4,90

this invention

TEGDME Ex. no. 2.2 50 4,846

this invention

TEGDME (extra)Ex. 25 4,514

no. 9.1

this invention

TEGDME (extra)Ex. 25 3,483

no. 9.2

this invention

TEGDME (extra) Ex. 50 5,379

no. 10.1

this invention

TEGDME (extra) Ex. 50 5,822

no. 10.2

this invention A. Henni et al. 2005 25 4,4

A. Henni et al. 2005 40 6,0

A. Henni et al. 2005 50 7,6 Calculated from a linear regression of the articles 3 temperatures

(25°C, 40°C, 60°C).

A. Henni et al. 2005 60 8,9

S. F. Sciamanna and 25 3,4

Scott Lynn

1988

D. Kodama et al. 2011 40 5,237 Calculated from a linear regression.

Solubility, experimental data

The experimental data for the solubility of C0₂ with 15 wt% AMP is plotted in Figure 22 and with 25 wt% AMP in Figure 24, for both 25°C and 50°C. Figure 23 shows the data with 15 wt% AMP with cubic regression and Figure 25 shows the data with 25 wt% AMP with cubic regression.

The experimental data for the solubility of C0₂ at 25°C is plotted in Figure 26 and at 50°C is plotted in Figure 28, for both 15wt% and 25wt% AMP. Figure 27 shows the data at 25°C with cubic regression and Figure 29 shows the data at 50°C with cubic regression for 15wt% AMP and for 25wt% AMP. Figure 29 shows the solubility of C0₂ in AMP for all the experiments, with cubic regression. To note is that ex. no. 3.3 had a final pressure of 4 bars and has 8 equilibrium points compared to the rest of the data that has 6 equilibrium points. Experimental solubility data compared to literature data

Figure 31 shows a comparison of the solubility experimental data of C0₂ in mixtures of AMP) and TEGDME (average for each wt% AMP) with literature data from Murrueta-Guevara and Rodriguez (1984) at 25°C. Figure 32 shows the same results but for 50°C. The dotted lines represent the article data for MEA and the dashed lines represent the article data for DEA, both at 5.1wt% and 14,3 wt% respectively. The lines are all cubic regressions. To note is that experiment, no. 6.1 and 6.2 both had a final pressure of 4 bars, resulting in 8 equilibrium points.

Heat of absorption, experimental data

The heat of absorption was calculated for the mixtures of AMP and TEGDME according to equation 3. These results have been plotted against the C0₂-loading in Figure 33 with 15 wt% AMP and Figure 35 with 25 wt% AMP, as differential calculations of the heat of absorption. The integral plots of the heat of absorption can be seen in Figure 34 for 15 wt% AMP and Figure 36 with 25 wt% AMP, with cubic regressions. The C0₂-loading is the molar amount of CO₂ absorbed divided by the total molar amount amine AMP. To note is that for the system examined the maximum CO₂- loading is 0,5mol CCVmol amine according to the reaction stoichiometry.

The results from the calculations of the heat of absorption, temperature and C0₂-loading can be seen in Table 13. The partial pressure of the experiments has been plotted against CCVloading in Figure 37 with 15 wt% AMP and Figure 38 with 25 wt% AMP, to examine what effect higher temperature has to the system. Figure 37 and 38 best fit was with cubic regression.

Table 13. The results from all the experiments conducted with the mixtures of AMP and TEGDME.

Experiment no.3.1 (15wt% AMP) Experiment no.3.2 (15wt% AMP) a a

Pco2 (molC0₂/mol ΔΗ^ ΔΗ_ίηί Pco2 (molC0₂/mol AH_diff ΔΗ_¾ (MPa) Xco2 Am) (kj/mol) (kj/mol) (MPa) Xco2 Am) (kj/mol) (kj/mol)

0,0038 0,0087462 0,0338034 125,78842 125,78842 0,0037 0,0082514 0,0318554 127,36442 127,36442

0,0078 0,0256857 0,1009995 133,26398 130,76199 0,0065 0,0244098 0,0957973 131,78327 130,31387

0,0079 0,0509923 0,2058552 127,0844 128,88875 0,0091 0,0494325 0,1991065 131,14669 130,74599

0,0098 0,0792818 0,3298934 130,49836 129,49395 0,0123 0,0796946 0,3315529 131,83728 131,18193

0,0566 0,1074134 0,4610362 96,269358 120,04314 0,0604 0,107458 0,4609628 98,620919 122,04081

0,1546 0,1248868 0,5467382 20,752394 104,47918 0,1525 0,1252806 0,5483666 19,793603 105,74371

Experiment no.3.3 (15wt% AMP)

a

Pc02 (molC0₂/mol ΔΗ^ ΔΗ_ιηί

(MPa) Xco2 Am) (kj/mol) (kJ/mol)

0,0045 0,0098712 0,0377343 120,16124 120,16124

0,0104 0,0268032 0,1042418 130,02088 126,45181 0,01 1 1 0,052512 0,2097683 125,17124 125,80761

0,0149 0,0846789 0,3501525 124,12133 125,13154

0,0654 0,1 103906 0,4696651 93,261425 1 17,02176

0,1527 0,1277076 0,554128 18,833409 102,05541

0,2233 0,1425151 0,6290566 17,214582 91,949797

0,273 0,1559428 0,6992764 14,442911 84,166726

Experiment no. 4.1 (15wt% AMP) Experiment no. 4.2 (15wt% AMP)

PC02 (molC0₂/mol Δ¾ PC02 (molC0₂/mol AH_d ΔΗ,_Γ

Xc02 XCQ2

0,0121 0,0071196 0,027475 125,17542 125,17542 0,0087 0,0072234 0,0278719 1 19,01545 119,01545

0,0169 0,0175827 0,0685755 138,37777 133,08821 0,013 0,0186897 0,0729586 130,66307 126,21341

0,0215 0,0317235 0,1255336 135,84347 134,33835 0,018 0,0344802 0,1368005 133,22484 129,4855

0,0263 0,0500985 0,2020806 137,93314 135,70004 0,0222 0,0536028 0,2169666 133,312 130,89934

0,0335 0,0719497 0,2970543 138,24955 136,51516 0,0297 0,0768475 0,318886 134,42816 132,02718

0,0746 0,0960862 0,4072984 115,1 1928 130,72391 0,0724 0,1008265 0,4295466 1 16,26686 127,96698

Experiment no. 5.1 (25wt% AMP) Experiment no. 5.1 (25wt% AMP) a a

Pco2 (molC0₂/mol ΔΗ^ ΔΗ_ίηί p_∞2 (molC0₂/mol AH_diff ΔΗ_¾ (MPa) Xco2 Am) (kj/mol) (kj/mol) (MPa) x_C02 Am) (kj/mol) (kj/mol)

0,0086 0,0080626 0,0202994 1 10,5571 1 10,5571 0,0087 0,0071622 0,0180221 124,32869 124,32869

0,0149 0,0208967 0,0533019 101,77077 105,1 1693 0,0142 0,022543 0,0576168 125,22438 124,94422

0,02 0,043981 1 0,1 148927 121,76521 1 14,04161 0,0192 0,04496 0,1 176087 125,03743 124,99176

0,0257 0,0955736 0,2639109 123,8011 1 19,55234 0,0195 0,0965757 0,2670614 122,01537 123,32611

0,0321 0,1250355 0,3568909 118,52684 1 19,28517 0,031 0,1 166365 0,3298601 189,8378 135,98861

0,06 0,1489223 0,4370017 101,66879 1 16,05575 0,0679 0,1400215 0,4067636 102,94978 129,74223

Experiment no. 6.1 (25wt% AMP) Experiment no. 6.2 (25wt% AMP) a a

0,0132 0,0062028 0,0155871 93,688793 93,688793 0,0095 0,0075228 0,0189315 122,06455 122,06455

0,0172 0,0146356 0,0370926 164,52932 134,76063 0,01 18 0,0188218 0,04791 16 77,032304 94,826047

0,0224 0,0294745 0,0758423 123,69982 129,10938 0,0153 0,0342737 0,0886408 136,47429 113,96284

0,0274 0,0486327 0,1276593 132,59464 130,52405 0,0192 0,0559987 0,1481607 129,16786 120,07108

0,0316 0,0776716 0,2103045 98,622646 1 17,98748 0,0231 0,0915248 0,2516244 128,75477 123,64167

0,0391 0,1 1754 0,3326308 119,89845 1 18,69024 0,0317 0,1227199 0,3493849 87,949678 113,65479

0,0572 0,1437769 0,4193476 49,104711 104,30068 0,0782 0,1453073 0,4246241 27,940209 98,467013

0,1292 0,1648187 0,4928304 5,092772 89,508421 0,1762 0,1627939 0,4856609 4,9108476 86,70907

Precipitation

In all the experiments with the reactant AMP a precipitation was formed almost immediately after the first C0₂ loading. The precipitation started as small flakes in the solution, but towards the end of the experiment the solution became more viscous. Table 14 described the precipitation formed.

Table 14. Precipitation formed in the mixtures of TEGDME and AMP. Scale from 0 to 3 where 0 is no precipitation. 25 °C 50 °C

100wt% 0, No precipitation 0, No precipitation

TEGDME

15wt% AMP 1 , Yes clear precipitation 2, Yes strong precipitation

25wt% AMP 2, Yes strong precipitation 3, Yes very strong (thick) precipitation

Discussion

The absorption system in this thesis is comparable to the article data from Murrieta-Guevara and Rodriguez, which show promising absorption levels at both compositions of AMP. This ensures that the mixtures of AMP and TEGDME are able to absorb C0₂.

Henry 's constant

The results from experiments no. 1.1 and 1.2, for Henrys constant, show a close fit with the reference data from Sciamanna (1988) at 25 °C, but with a somewhat higher slope. Experiment 9.2, at the same temperature, is a closer fit to Sciamanna (1988), with a more linear slope. Experiment no. 9.1 has probably been affected by the pressure not being able to stabilize at the beginning of the run and show a large linear difference and should not be seen as accurate data for Henry's constant.

The experimental results for experiment, no. 2.1 and 2.2, for Henry's constant at 50 °C, are close to those of Kodama (2011) at 40 °C and far from those of Henni at 40 °C, 50 °C and 60 °C. Experiment no. 10.1 and 10.2 have very good linear regressions. Their results are somewhat over Kodama (2011) which is preferable since the Kodama (2011) reference data is at 40°C. The ex.no.10.2 is closest to the data from Henni (2005) at 40°C. The data from Henni (2005) at 50 °C and 60 °C is higher than all the experimental results conducted at 50°C which was expected.

The results for Henry's constant are comparable with the reference data Sciamanna (1988), Henni (2005), Kodama (2011). Due to some leaks being found in the experimental setup the extra experiments (no. 9.1, 9.2, 10.1, 10.2) of the pure solvent are to be believed the most reliable, with the exception of ex. no. 9.1. Solubility, experimental data

The experimental results of the solubility of C0₂ with 15 wt% AMP are coherent to their temperature of 25 °C and 50 °C. As the temperature increases from 25 °C to 50 °C, the solubility of C0₂ in the amine AMP decreases. Experiment no. 3.3 with more equilibrium points, shows that the amount of C0₂ that is dissolved starts to decrease at higher pressures, as the slope of the curve gets bigger. To see exactly how higher pressure would affect the solubility of AMP more experiments would have to be conducted.

In Figure 24 the experimental data with 25 wt% AMP are scattered for the 2 temperatures and show no apparent pattern. The best fit for the data was a cubic regression as shown in Figure 25.

At 25 °C, in Figure 26 the solubility data is fairly gathered at the beginning and starts to divide clearly at the mole fraction point of 0.1. By increasing the amount of amine, AMP the solubility of C0₂ increases. Figure 27 shows the cubic regression of the previous mentioned experimental data, which was the best fit. The same conclusion can be made at 50 °C, see Figure 28. Figure 29 shows the data at 50 °C with cubic regression for 15 wt% AMP and cubic regression for 25 wt% AMP, this was their best fit.

Experimental solubility data compared to article data

The patterns from the experimental solubility data are confirmed by the data in the article from Murrieta-Guevara and Rodriguez (1984). The absorption levels for the experimental data of 15 wt% AMP, for both 25 °C and 50 °C, are close to those of 14,3 wt% MEA. The absorption levels of 25 wt% AMP exceed all the article data, for both 25°C and 50°C. It is desirable to do the experiments at higher pressures for a better comparison.

Heat of absorption experimental data

A pattern appears in the results of the figures of the heat of absorption as a function of C0₂ loading. The heat of absorption increases from equilibrium point 1 to 2, to then descend to equilibrium point 3 to and increase again at point 4. The high points of this pattern is most likely due to the extra heat produced when the precipitation is formed, the heat of crystallization. It is hard to draw a clear conclusion and in order to obtain better and more accurate results for the heat of crystallization further experiments should be performed with smaller pressure steps. Smaller pressure steps should increase the possibility to follow the reaction more closely and hopefully see a division between the heat of crystallization and the heat of solution. In the integral and differential plots the heat of absorption starts by increasing to then decrease as the C0₂ loading increases.

After the maximum C0₂-loading 0,5mol C0₂/mol amine, only physical absorption takes place. This can be seen as the heat of absorption decreases rapidly after the maximum loading point is reached and is more apparent when looking at the differential plots.

Precipitation

In all the experiments conducted with the amine, AMP, a white precipitation was formed. The precipitation increased when the temperature was increased but also when the concentration of the amine increased. After the formation of the precipitation, the amine shifts the equilibrium reaction so that more amine is able to absorb more C0₂. These observations have not been mentioned in any other known amine systems.

During experiment no. 6.1 and 6.2 the slurry was very thick. The precipitation was so extensive that it stuck to the reactors walls and bottom. This could possibly have altered the results for these experiments, due to the fact that all measurements from the reactor come from the bottom of the reactor. If the precipitation increased so much that the stirrer was unable to stir the solution completely, the measurements that are closest to the stirrer would be different to those at the bottom at the reactor.

References

Articles Forster, P., V. Ramaswamy, P. Artaxo, T. Berntsen, R. Betts, D.W. Fahey, J. Haywood, J. Lean, D.C. Lowe, G. Myhre, J. Nganga, R. Prinn, G. Raga, M. Schulz and R. Van Dorland, 2007: Changes in Atmospheric Constituents and in Radiative Forcing. In:

Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M.Tignor and H.L. Miller (eds.)]. Cambridge, United Kingdom and New York, NY, USA: Cambridge University Press.

Kierzkowska-Pawlak H. (2007): Pressure swing absorption of carbon dioxde in n- methyl-2-pyrrolidone solutions. Polish Journal of Chemical Technology, 9, 2, 106-109.

Melzer, W.-M.; Schrodter, F.; Knapp, H. (1988): Solubilities of methane, propane and carbon dioxide in solvent mixtures consisting of water, Ν,Ν-dimethylformamide, and N- methyl-2-pyrrolidone. Fluid Phase Equilibria, 49, 167-186.

Murrieta-Guevara F, Rebolledo-Libreros E., Trejo A. (1992): Solubility of Carbon Dioxide in Binary Mixtures of N-Methylpyrrolidone with Alkanolamines . J. Chem. Eng. Data, 37, 4-7.

Murrieta-Guevara F., Rodriguez A.T. (1984): Solubility of Carbon Dioxide, Hydrogen Sulfide, and Methane in Pure and Mixed Solvents. J. Chem. Eng. Data, 29, 456-460. (Journal include reference to article: Rivas, O. R., Prausnk J.M. (1979): AICE J. 1979, 25, 975.)

Park J.Y., Yoon S. J., Lee H. (2003): Effect of Steric Hindrance on Carbon Dioxide Absorption into New Amine Solutions: Thermodynamic and Spectroscopic Verification through Solubility and NMR Analysis . Environ. Sci. Technol. 2003, 37, 1670-1675.

Samanta A., Bandyopadhyay S.S. (2008): Absorption of carbon dioxide into aqueous solutions of piperazine activated 2-amino-2-methyl-l-propanol. Chemical Engineering Science 64, 1185-1194.

Sanchez S., Pacheco R., Sanchez A., La Rubia M. D. (2005): Thermal Effects of CO₂ Absorption in Aqueous Solutions of 2-Amino-2-methyl-l -propanol. American Institute of Chemical Engineers (AIChE) Journal, Vol. 51, No. 10, 2769-2777. Sayari A., Belmabkhout Y. (2010): Stabilization of Amine-Containing C02 Adsorbents: Dramatic Effect of Water Vapor, Journal of the American Chemical Society (ACS Publications), 132, 6312-6314. Shou Xu, Yi-Wei Wang, Frederick D. Otto, Alan E. Mather (1996): Kinetics of the reaction of carbon dioxide with 2-amino-2methyl-l -propanol solutions. Chemical Engineering Science, Vol. 51, No. 6, 841-850.

Thambimuthu K., Davison J., Murlidhar G. (2002): CO 2 Capture and Reuse. ECN: IPCC, 31-52.

Whitfield J.K., Ramsey G.H., Adams N.H., N es CM., Gillum D.E. (1999):

Demonstration of n-methyl pyrrolidone (NMP) as a pollution prevention alternative to paint stripping with methylene chloride, Journal of Cleaner Production 7, 331-339.

S. Bouzalakos, M. Merdedes Maroto-Valer (2010) Overview of carbon dioxide (CO 2) capture and storage technology University of Nottighamn, UK, Woodhead Publishing Limited

Christian Azar, Kristian Lindgren, Eric Larson and Kenneth Mollersten (2006) Carbon Capture and Storage from Fossil Fuels and Biomass- Costs and Potential Role in Stabilizing the Atmosphere

Masahiko Tatsumi, Yasuyuki Yagi, Kouji Kadono, Kazhuiko Kaibara, Masaki Iilima, Tsyyoshi Ohishi Tanaka, Takuya Hirata, Ronald Mitchell (2010) New

Energy Efficient Processes and Improvements for Flue Gas CO 2 Capture

CCJ (2012) Alstom and Dow progress on amines, Nov 04 2012, Carbon Capture Journal

CCJ (2007) Mitsubishi's carbon capture technology, Nov 20 2007, Carbon Capture Journal

Gary T. Rochelle (2009) Amine Scrubbing for C0₂ Capture, Science Magazine Vol 325 (25 September 2009), 1652-1654

W.l.Echt (1997) Chemical Solvent-Based Process For Acid Gas Removal In Gasification Applications, 1-13 J. H. Meldon (2011) Amine screening for flue gas CO ₂ capture at coal-fired power plants. Should the heat of desorption be high, low or in between? Current opinion in Chem. Eng. 1, 55-63

S.Xu, Y.-W. Wang, F.D. Otto and A.E. Mather, (1992) Representation of equilibrium solubility properties of CO₂ with aqueous solutions of 2-amino-2-methyl-l -propanol, Chemical Engineering and Processing 31, 7-12

Guido Sartori and David W. Savage (1983) Sterically Hindered Amines for CO, Removal from Gases) Ind. Eng. Chem. Fundam. 22,239-249

Inna Kim, Hallvard F. Svendsen (2011) Comparative study of the heats of absorption of post-combustion C0₂ absorbents, International Journal of Greenhouse Gas Control 5, 390-395

S.K. Dash, A.N. Samanta, S.S. Bandyopadhyay (2012) Experimental and theoretical investigation of solubility of carbon dioxide in concentrated aqueous solution of 2- amino-2 -methyl- 1 -propanol and piperazine J.Chem. Thermodynamics 51. 120-125

Steven F. Sciamanna, Scott Lynn (1988) Solubility of Hydrogen Sulfide, Sulfur Dioxide, carbon Dioxide, Propane, and n-Butane in Poly (glycol ethers) Ind.Eng.Chem.Res. 27. 492-499 Amr Henni, Paitoon Tonitiwachikul and Amit Chakma (2005) Solubilities of carbon Dioxide in polyethylene Glycol Ethers. The Canadian Journal of Chemical Engineering Vol 83. 358-361

D. Kodama, M. Kanakubo, M. Kokubo, S. Hashimoto, H. Nanjo, M. Kato (2011) Density, viscosity, and solubility of carbon dioxide in glymes Fluid Phase Equilibria 302, 103-108

F. Murrieta-Guevara, A. T. Rodriguez (1984) Solubility of Carbon Dioxide, Hydrogen Sulfide, and Mixed Solvents J. Chem. Eng. Data 29. 456-460 A. Dermirbas (2004) Combustion characteristics of different biomass fuels Progress in Energy and Combustion Science 30. 219-230

Books

Jones L., Atkins P. (2000): Chemistry: Molecules, Matter and Change, USA: W. H. Freeman, 538.

Sterner O. (2003): Drivmedel, avgaser och andra luftfororeningar, Forgiftningar och miljohot, Lund: Studentlitteratur, 230-233.

Arthur L. Kohl, Richard B. Nielsen, (1997): Gas purification (Fifth Ed) ISBN:978-

00- 88415-220-0 Other literature

Descamps C. (2004): Etude de la capture du C0₂ par absorption physique dans les systems de production d' electricite bases sur la gazeification du charbon integree a un cycle combine. Ecole des Mines de Paris. (Doctoral thesis)

Elwell L.C., Grant W.S (2005): Technical Overview of Carbon Dioxde Captrue Technologies for Coal-Fired Power Plants. MPR Associates Inc., Alexandria Virginia,

1- 15.

IEA GHG (2000b): Leading options for the capture of CO₂ emissions at power stations, Report number PH3/14, IEA Greenhouse Gas R&D Programme, Stoke Orchard, Cheltenham, UK. Persson O. (2008): Thermochemical investigation of carbon dioxide absorption using reaction calorimetery , Lund university: Department of Chemical engineering, 1-28. (Master thesis.)

Schlauer J. (2008): Absorption, I. Fundamentals, Lurgi GmbH, Weinheim: Wiley-VCH Verlag GmbH & Co. KGaA, page 1-25. Topham, S. (2005): Carbon dioxide. Imperial Chemical Industries, Weinheim: Wiley- VCH Verlag GmbH & Co. KGaA, 1-19. IEO201 1 (201 1),7¾e International Energy Outlook 2011 was prepared under the general direction of John Conti, Assistant Administrator for Energy Analysis (john.conti@eia.gov, 202-586-2222), and Paul Holtberg, Team Leader, Analysis Integration Team (paul.holtberg@eia.gov, 202-586-1284), 9-16, 139-144. Also available at: http://www.eia. gov/forecasts/ieo/pdf/0484(201 D.pdf

B. Metz, O.R. Davidson, P.R. Bosch, R. Dave, L.A. Meyer (eds) (2007) IPCC, 2077: Climate Change 2007: Mitigation. Contribution of Working Group III to the Fourth Assesment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, United Kingdom and New York, NY, 102-116. Metz, B.O. Davidson, H. C. de Coninck, M. Loos, and L. A. Meyer (2005) IPCC, 2005: IPCC Special Report on Carbon Dioxide Capture and Storage. Prepared by Working Group III of the Intergovernmental Panel on Climate Change [Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA

IEAGHG (201 1) The International Energy Agency, Potential for Biomass and Carbon Dioxide Capture and Storage

Kelly Thambimuthu (2002) IPCC workshop on carbon dioxide capture and storage. CO₂ Capture and Reuse, CANMET Energy Technology Centre, Natural Resources Canada John Davison,IEA Greenhouse Gas R&D Programme Cheltenham, United Kingdom Murlidhar Gupta CANMET Energy Technology Centre Natural Resources Canada, 52pages

Electronic

U.S. Department of Energy (2008): Enhanced Oil Recovery/C0₂ Injection, Fossil Energy Office of Communications, Available from: http://fossil.energy.gov/programs/oilgas/eor/ [ 1 October 2010]. SSE (2011), Scottish and Southern Energy, SSE (2011) Ferrybridge, The carbon capture project, Project information. Available from: [30 November 2011] http: vvvvvv.ss

plate.aspx?pageid=3957

ESRL (2013) Earth System Research Laboratory, Carbon Dioxide at NOAA 's Mauna Loa Observatory reaches new milestone: Top 400ppm. Available from [10 May 2013] http://www.esrl.noaa.gov/news/2013/CQ2400.html.

Oral

Hans T. Karlsson (2008), Department of Chemical Engineering at Lund University. Olov Sterner (2009), Department of Chemistry (KILU) at Lund University.

Claims

1. A method for removing carbon dioxide (C0₂) from a gas stream comprising the steps of:

providing a gas stream containing C0₂; and

contacting the gas stream with a composition comprising:

2-amino-2-methyl-l-propanol (AMP); and

to remove the C0₂ from the gas stream.

2. The method according to claim 1, wherein contacting the gas stream with the composition forms a solid that is separated from the composition.

3. The method according to claim 2, further comprising the step of recycling the composition separated from the solid back to the contacting step.

4. The method according to claim 2 or claim 3, further comprising the step of heating the solid to release C0₂ and regenerate AMP.

5. The method according to claim 4, further comprising the step of recycling the AMP separated from the heating step back to the contacting step.

6. The method according to claim 4 or claim 5, wherein the solid is heated at a temperature below 100 °C.

7. The method according to any one of claims 4 to 6, wherein the solid is heated at a temperature between about 70 to 80 °C.

8. The method according to any preceding claim, wherein the gas stream is contacted with the composition at a temperature between about 25 to 50 °C.

9. The method according to any preceding claim, wherein the gas stream is contacted with the composition at a temperature of about 25 °C.

10. The method according to any preceding claim, wherein the gas stream is contacted with the composition at a temperature of about 50 °C.

11. The method according to any preceding claim, wherein the solvent is NMP.

12. The method according to any preceding claim, wherein the solvent is TEGDME.

13. The method according to any preceding claim, wherein the C0₂-containing gas stream is a flue gas stream.

14. The method according to any preceding claim, wherein the composition comprises 15 to 25 wt.% AMP.

15. The method according to any of the claims 4 to 7, wherein the released C0₂ is received at elevated pressure, such as about 6 bars.

16. Use of a composition comprising:

2-amino-2-methyl-l-propanol (AMP); and

to remove C0₂ from a gas stream.

17. The use according to claim 16, wherein AMP and C0₂ react to form a solid that is separated from the composition.

18. The use according to claim 17, wherein the composition separated from the solid is recycled to remove C0₂ from the gas stream.

19. The use according to claim 17 or claim 18, wherein the solid is heated to release C0₂ and regenerate AMP.

20. The use according to claim 19, wherein the AMP separated from the heating step is recycled back to the composition to remove C0₂ from the gas stream.

21. The use according to claim 19 or claim 20, wherein the solid is heated at a temperature below 100 °C.

22. The use according to any one of claims 19 to 21, wherein the solid is heated at a temperature between about 70 to 80 °C.

23. The use according to any one of claims 16 to 22, wherein the gas stream is contacted with the composition at a temperature between about 25 to 50 °C.

24. The use according to any one of claims 16 to 23, wherein the gas stream is contacted with the composition at a temperature of about 25 °C.

25. The use according to any one of claims 16 to 23, wherein the gas stream is contacted with the composition at a temperature of about 50 °C.

26. The use according to any one of claims 16 to 25, wherein the solvent is NMP.

27. The use according to any one of claims 16 to 25, wherein the solvent is TEGDME.

28. The use according to any one of claims 16 to 27, wherein the C0₂-containing gas stream is a flue gas stream.

29. The use according to any one of claims 16 to 28, wherein the composition comprises 15 to 25 wt.% AMP.

30. The use according to any one of claims 19 to 22, wherein the released C0₂ has an elevated pressure, such as about 6 bars.