Phase-change Solvents and Processes for Post-Combustion CO2

Subscriber access provided by Iowa State University | Library

Review

Phase-change Solvents and Processes for PostCombustion CO2 Capture- A Detailed Review Athanasios Papadopoulos, Fragkiskos Tzirakis, Ioannis Tsivintzelis, and Panos Seferlis Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/acs.iecr.8b06279 • Publication Date (Web): 11 Mar 2019 Downloaded from http://pubs.acs.org on March 17, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Phase-change Solvents and Processes for PostCombustion CO2 Capture- A Detailed Review

Athanasios I. Papadopoulos1,*, Fragkiskos Tzirakis1, Ioannis Tsivintzelis1,2, Panos Seferlis1,3 1Chemical

Process and Energy Resources Institute, Centre for Research and Technology-Hellas, 57001 Thermi, Greece,

2Department

of Chemical Engineering, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece

3Department

of Mechanical Engineering, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece

*Corresponding author: [email protected]

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract This work reviews research activities in the development of phase-change CO2 capture solvents and processes. The focus is on liquid-liquid phase-change solvents where the CO2-lean phase can be recycled to the absorber prior to regeneration hence the energy demands may be substantially decreased compared to non-phase change processes. The review briefly provides the basic chemical and physical principles required to understand the phase-change behavior in postcombustion CO2 capture systems. The reviewed work is further organized per different solvent type into major sections including experimental property measurement studies, experimental pilot plant studies, thermodynamic and kinetic modeling studies as well as process modeling and technoeconomic assessment studies. In all sections we provide details regarding experimental approaches, operating conditions of pilot plants, implementations in different industries and performance data. Key findings include operating observations and performance indicators regarding the investigated solvents and processes as well as a substantiated estimation of their technology readiness level. Keywords: CO2 capture, phase-change solvents, absorption/ desorption process, vapour-liquidliquid equilibrium


Page 2 of 61

Page 3 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1. Introduction Numerous emerging technologies for post-combustion CO2 capture are on the spot of many research activities. These include solid sorbents, membranes and new liquid absorbents. Among such technologies, absorption/desorption processes that employ a liquid absorbent/solvent have received a lot of attention, because conditions for both absorption and solvent regeneration are relatively easy to meet and the process can be easily retrofitted onto existing plants (Aaron and Tsouris, 2005). A major downside of such processes is the high energy (heat) required for solvent regeneration, which introduces significant energy penalties in the CO2 emitting plants where they are fitted. In recent years, intensive research efforts to address this shortcoming focused on the development of new solvents to replace the conventionally used monoethanolamine (MEA). While numerous different solvents have been investigated at an experimental or pilot scale1,2 very few have been tested at larger capacity units and lengthier campaigns. Among such solvents, mixtures such as 2-amino-2-methyl-1-propanol (AMP) and Piperazine (PZ), methyldiethanolamine (MDEA) and PZ or the proprietary solvent KS-1 appear to be promising as they indicated potential for regeneration energy close to 3.0 GJ/ton of CO2 captured (or slightly lower), compared to the approximately 4.0 GJ/ton of CO2 captured with the 30%wt MEA for 90% capture3,4. In order to be realistic about developing commercially feasible absorption/desorption systems it is necessary to find solvents that are able to achieve regeneration energy targets in the order of 2.0 GJ/ton of CO2 captured or below. Phase-change solvents represent a class of materials promising to deliver significant energy reductions compared to conventional solvents5. They consist of miscible solute-solvent mixtures (e.g. amines in water) which undergo phase separation upon a change in the processing conditions, e.g. upon reaction of CO2 with the solute or upon increasing the temperature after the reaction etc. This behavior results in the formation of a CO2-lean phase which may be separated through a non-thermal approach. Both the CO2-rich and CO2-lean phases may be in a liquid state, or the CO2-rich phase may be a solid (e.g. as in the case of amino-acid salt solvents)5. Phase-change solvents which exhibit liquid-liquid phase separation can readily exploit existing absorption/desorption systems through the incorporation of a mechanical separation step. On the other hand, the precipitation of a solid phase involves significant challenges in process design5. In this work, we investigate the case of liquid-liquid phase-change solvents where the CO2-lean phase is recycled to the absorber prior to regeneration hence the energy demands are substantially decreased compared to non-phase change processes because of the need to regenerate less amine solvent. There are currently different types of compounds that exhibit such behavior and represent potential CO2 capture candidates. Previously, Wang and Li5 as well as Mumford et al.6 published book chapters on such solvents which briefly introduce the phasechange concept, covering few published works. Furthermore, Zhuang et al.7 has presented the only review on phase-change solvents available to date, which selects few representative results for less than half of the currently available literature. Other than presenting a complete account of all existing efforts in phase-change solvents and processes, this work takes a systematic review approach. The reported developments are classified in the text in terms of the nature of the reported investigations, with main classes including experimental property measurement/characterization studies, experimental-pilot plant studies, thermodynamic/ kinetic


1


Page 4 of 61

modeling studies and process modeling/ technoeconomic assessment studies. Details and insights are organized in original tables and figures regarding experimental or model-based methods, materials, equipment sizes and conditions together with solvent and process selection criteria and key performance indicators. An estimated assessment of the technology readiness level (TRL) is further provided and substantiated, while significant achievements and observations are also summarized. Section 2 provides the basic chemical and physical principles required to understand the phase-change behavior in post-combustion CO2 capture systems. Sections 3-6 provide a detailed review of the works themselves. Section 7 provides a comparative analysis of key performance indicators for each solvent type and presents some concluding remarks. 2. Basic principles and concepts 2.1 CO2 chemical absorption in amine solutions The reaction of CO2 with primary and secondary alkanolamines, e.g. MEA or DEA, includes three main steps, i.e. the carbamate formation, the bicarbonate formation and the carbamate reversion8. CO2 + 2RNH2 ↔ RNH3+ + RNHCOO–

(1)

CO2 +

RNH2 + H2O ↔ RNH3+ + HCO3–

(2)

CO2 +

RNHCOO– + 2H2O ↔ RNH3+ + 2HCO3–

(3)

The carbamate formation is the dominating reaction in excess of amine conditions and takes place through the formation of a zwitterion8,9,10.If the carbamate formation is the dominating reaction, the maximum loading is 0.5 mole of CO2 per mole of amine, posing a stoichiometric limit to chemical absorption. However, a certain amount of carbamate hydrolysis occurs with all amines, e.g. MEA and DEA, and, consequently, the loading may exceed such value, particularly at high pressures11.The amine, water and hydroxyl anion can contribute to the deprotonation of the zwitterion10. Tertiary amines react with CO2, in presence of water, through a different mechanism. CO2 + R3N + H2O ↔ R3NH+ + HCO3–

(4)

The stoichiometric limit of such reaction is 1 mole of CO2 per mole of amine. Consequently, tertiary amines are often treated as another class of solvents. In this direction, Aleixo et al.12 classified potential DMXTM phase-change solvents in two broad categories: the former includes primary and secondary amines, while the latter includes tertiary amines. Sterically hindered amines, such as AMP, may be classified in another group of solvents. However, there is no agreement on the relevant chemical kinetic expressions13. A possible reaction scheme in aqueous AMP solutions is the formation of carbamate, through a zwitterion mechanism. The carbamate is rather unstable and undergoes a carbamate reversion reaction to bicarbonate11. Such behavior was confirmed by NMR spectroscopy14.


2

Page 5 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


For all types of amines (primary, secondary, tertiary and sterically hindered or not), the theoretical limit of CO2 absorption is, to a large extend, posed by the stoichiometry of the dominant reactions, since chemical absorption is more significant from physical absorption. However, physical absorption is always present and increases with increase of pressure, at least up to moderate pressures11. The reaction route in aqueous alkanolamine solutions may proceed through the so called “shuttle mechanism”15. According to this, initially, the dissolved CO2 undergoes a fast reaction towards the formation of carbamate. Being fast, such reaction occurs near the gas-liquid interface and, consequently, steepens the CO2 concentration profile enhancing the CO2 mass transfer rate. Since the carbamate reversion to bicarbonate is a slow reaction, the carbamate diffuses to the bulk of the liquid phase, where it reverts to bicarbonate and liberates amine that is able to react again with CO2 15,16. More recently, Zhang17 considered a similar route in order to explain the relatively fast reaction rates in heterogeneous systems consisting of two liquid phases, i.e. an upper liquid amine rich phase and a lower phase that is rich in water. As shown in Figure 1, initially, CO2 dissolves in the organic phase and, near the gas-liquid interface, undergoes a fast reaction with amine towards the formation of carbamate. The resulting ions migrate to the aqueous phase, where carbamate reversion and bicarbonate formation occur. The carbamate that is consumed through the bicarbonate formation in the aqueous phase is replaced as more amine is dissolved from the upper phase.

Figure 1: Illustration of the reaction route in biphasic systems17. 2.2 Liquid – Liquid phase separation 2.2.1 The LCST of amine – water mixtures Typical aqueous amine solutions undergo liquid-liquid phase separation at elevated temperatures presenting a Lower Critical Solution Temperature (LCST)18, as shown in Figure 2 for the ethylbutyl-amine – water system.


3


360 350

Temperature / K

340 330 320 310 300 290 280 270 260 0

0.1

0.2

0.3

0.4

Amine mole fraction

0.5

0.6

Figure 2: The LSCT behavior of the ethylbutylamine – water mixture (Experimental data reported by Goral et al.18). Such phase behavior mainly arises from hydrogen bonding interactions. Water, primary and secondary amine molecules have both proton donors and acceptors and, consequently, self- and cross-association hydrogen bonding interactions occur in aqueous mixtures. Tertiary amines have only proton acceptors and, consequently, only cross-association hydrogen bonding interactions occur in mixtures with water. At low temperatures, the development of crosshydrogen bonds results in homogeneous mixtures at all concentrations. The decrease of crossassociation interactions due to temperature increment renders the effect of hydrophobic interactions more pronounced and the mixture separates in two immiscible liquid phases. For amines that are not decomposed at high temperatures, further temperature increment at elevated pressures may result in miscibility enhancement and, thus, the phase behavior diagram presents a closed loop shape, in which both a Lower Solution Critical Temperature (LCST) and an Upper Critical Solution Temperature (UCST) are shown (Figure 3). 560 520

Temperature / K

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 61

480 440 400 360 320 280 0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

1-methylpoperidine mole fraction

Figure 3: 1-methylpiperidine + water liquid – liquid equilibrium (Experimental data reported by Goral et al.18).


4

Page 7 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2.2.2 Effect of CO2 - The liquid-liquid phase separation temperature The chemical absorption of CO2 in an aqueous amine solution and the subsequent production of new species have a non-negligible effect on the phase behavior. The produced multicomponent system may also present LCST, but the phase separation temperature is, to a large extent, affected by the CO2 content. If such phase separation temperature is low enough, then a threephase system (VLLE) may occur during the absorption process. Such behavior is observed in aqueous DEEA/MAPA solutions19,20. MAPA and water are found concentrated in the heavy, CO2 rich, liquid phase, while the CO2 lean phase contains mainly DEEA19. Only the CO2 rich phase needs to be regenerated resulting in relatively low heat requirements21. In order to improve process performance, the phase separation temperature of the CO2 loaded solution can be effectively changed by changing the composition of the solvent. This is usually performed using blended amines. The effective selection of amines and their mole ratio results in solvents with optimum phase separation temperature. Nevertheless, since the absorption of CO2 results in the production of many different species, the prediction of the phase separation temperature is difficult and usually the problem is solved after many trials and/or experimental measurements.

3. Lipophilic solvents as CO2 capture options 3.1 Main concepts Lipophilic or otherwise called Thermomorphic Biphasic Solvents (TBS) have been initially considered by Zhang22, then further developed by Agar et al.23,24 and by Tan25 . The analysis presented here is based on the work of Zhang17 and the corresponding publications, which have built upon and completed those earlier works. The key characteristic of lipophilic amines is that their loaded aqueous solutions (i.e. after reaction with CO2 in the absorber) exhibit a temperature-induced liquid-liquid phase separation (LLPS) in the range of 60oC-90oC8. The LLPS is therefore a key temperature to determine whether: a) The loaded amine exhibits phase separation. b) The new liquid phase appears at a temperature which is as low as possible to reduce regeneration thermal energy requirements. A typical flowsheet that corresponds to solvents exhibiting LLPS higher than the absorption conditions is shown in Figure 4.


5


Page 8 of 61

Figure 4: Typical, conceptual flowsheet of a phase-change process with LLPS higher than absorption conditions Most of the lipophilic amines studied by Zhang17 exhibit LCST at a concentration range of 1030%. The LCST is important because it determines the solubility in water of the regenerated rich-amine phase which is recycled to the absorber. The aqueous solubility of lipophilic amines decreases as the temperature increases, hence there is a higher likelihood to obtain single-phase aqueous amine solutions at lower temperatures. There are several cases where the rich-amine phase is not homogeneous (i.e. there is phase separation) at 40oC which is the typical absorption temperature. This is not a desirable condition for the operation of the absorber where the aqueous solvent should be homogeneous. It is therefore desirable to have solvents that exhibit: a) LCST higher than the absorption temperature (e.g. 40oC) b) LLPS as low as possible and closely above the LCST. Note that although Zhang17 screens solvents based on the above criteria, other works discussed in subsequent sections employ solvents with LLPS at 40oC (e.g. DEEA/MAPA) and present results that indicate low energetic regeneration requirements in pilot-plant experiments. Furthermore, Zhang17 proposes a distinction of lipophilic solvents as absorption activators and regeneration promoters based on their functionality in the absorption and desorption processes. The former exhibit fast reaction kinetics whereas the latter enable high regenerability (i.e. desorb as much CO2 as possible at the desired LLPS temperature)8,26. Based on this distinction, in the case of amine mixtures it is important to have: a) An activator with a high LCST. b) A promoter with a low LLPS.


6

Page 9 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3.2. Solvent screening and selection 3.2.1 Aqueous solutions of pure lipophilic amines Zhang17 investigates experimentally 44 amines as potential phase-change solvents and presents comparative results based on the following criteria: 1. 2. 3. 4. 5. 6. 7.

Cyclic capacity. Absorption rate. Regenerability. (a) LCST and (b) LLPS. CO2 loading capacity. Aqueous amine concentration range at which LLPS appears. Appearance of phenomena such as gel formation, precipitation, salts formation.

Figure 5 and Table 1 show the performance of some of the highest performing solvents identified in Zhang17. HA is used as a reference solvent because it is a relatively simple, linear, primary amine. The numbers used for Figure 5 are derived through visual observation from the corresponding diagrams reported in Zhang17. The bars provide comparable results among solvents for each criterion. Table 1 provides information regarding criteria 5-7 together with information regarding the amine functionality in the process. The CO2 loading capacity is reported qualitatively based on the system proposed in Zhang17.

Figure 5: Performance comparison of selected solvents with respect to criteria 1-4 using HA (Hexylamine) as a reference solvent. DPA: N, N-Dipropylamine, DsBA: Di-sec-butylamine, MCA: N-Methylcyclo hexylamine, DMCA: N,N-Dimethyl cyclohexylamine, MPD: N-Methyl piperidine, EPD: N-Ethyl piperidine. Blue bars pointing right indicate better performance than HA, red bars pointing left indicate worse performance than HA. Cyclic capacity is defined as the difference between CO2 rich and CO2 lean loading at 25oC and 90°C, respectively. This is without considering stripping gas for EPD and MDP, while stripping gas of 200ml/min N2 is considered at 75 oC for all the other amines. Regenerability is defined as the ratio of the difference between CO2 rich and CO2 lean loading over CO2 rich loading and is reported at 80 oC. Cyclic capacity and absorption rate are measured at 2.7 M. The regenerability of EPD is reported for a 3M concentration hence 2.7M and 4.5M concentrations don’t apply and the corresponding regenerability bars only provide an approximation. (Experimental data reported by Zhang17) The data in Figure 5 and Table 1 highlight quite clearly the amines that appear to have the highest performance and justify the ones finally selected by Zhang17. For example, DMCA has


7


Page 10 of 61

an overall above average performance in all criteria except for absorption rate and LCST. This is a clear indication that DMCA has slow kinetics and would operate best in a mixture with an activator. MCA on the other hand seems to be a very good activator with reasonable performance in all properties. Furthermore, DPA appears to have an overall good performance, but it precipitates hence it is not desirable. It is also worth noting that the LCST of the amines short-listed in Figure 5 and Table 1 are much lower than the 40oC at which the CO2 is absorbed in the absorption column hence a second liquid phase will exist in the absorber, which is undesirable. Table 1: Performance of selected solvents with respect to criteria 5-7 with information their potential function in CO2 capture. (Data obtained from Zhang17) Solventa

CO2 loading capacity (5)

LLPS concentration [M] (6)

Other issues (7)

Function

HA

Good

≥2.7

None

Activator

DPA

Good

≥2.7

Precipitation

Activator

DsBA

Excellent

All

None

Promoter

MCA

Good

NA

None

Activator

DMCA

Excellent

All

None

Promoter

MPD

Good

≥2.7

None

Promoter

EPD

Good

≥2.7

None

Promoter

Table 2 shows the structures of the above amines together with a discussion on how they affect their performance as phase-change solvents. Based on the analysis provided in Zhang17 the structure of the main carbon chain and the distance of the branches from the amine play a significant role in the properties of lipophilic amines. Cyclic, secondary amines and structures with branching to the carbon atom after the amine (i.e. α-carbon atom) exhibit favorable performance such as low viscosity and high absorption rates, while tertiary amines with similar characteristics exhibit high absorption capacity.


8

Page 11 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 2: Chemical structure of amines and comments (Data from Zhang17). Solvent

Structure

Comments

HA

Linear primary amine, 6 carbon atoms, volatile, likely to exhibit high losses.

DPA

Linear secondary amine, 3 carbon atoms in each side, likely to precipitate and to exhibit volatility-induced losses.

DsBA

Linear secondary amine with branches at α- carbon atom, high reaction rate, potential for salt formation.

MCA

Cyclic secondary amine, 6 carbon atoms, with branch at α-carbon (cycle side), high reaction rate, ideal case.

DMCA

Cyclic tertiary amine, 6 carbon atoms with branch at α-carbon (cycle side), high CO2 capacity, ideal case.

MPD

EPD

Tertiary heterocyclic amine, 5 carbon atoms, no branching, low reaction rate, branching at an αcarbon atome could be desireable. Tertiary heterocyclic amine, 5 carbon atoms, no branching, low reaction rate, branching at an αcarbon atome could be desireable.

3.2.2 Aqueous solutions of blended lipophilic amines The previous analysis indicates that there is no single amine incorporating all the necessary properties to enable efficient CO2 capture. To address this issue Zhang17 tests few aqueous blends of activators and promoters, comprising combinations of amines from Table 1. The performance of promoters, activators and their blends is illustrated in Figure 6 against MEA. It is clear that the performance of blends between MCA, DMCA and DsBA is better than MEA or MDEA (Methyldiethanolamine) which are widely investigated CO2 capture solvents. It is worth noting that the reported results are for regeneration temperature of 70-80oC. At this temperature, all the amines perform much better than MEA as reflected in the residual loading which indicates that the CO2 remaining in the amine after desorption is much lower than that of MEA. Further detailed tests at different ratios and conditions are performed by Zhang17 for the MCA+DMCA blend which seems to have clear advantages in all criteria compared to the other tested solvents. Although mixtures of MCA with DsBA also perform very well, they are avoided because DsBA may potentially precipitate.


9


Page 12 of 61

Figure 6: Comparison of selected lipophilic solvents and their blends and MEA/MDEA with MEA (5M) based on important properties. Blue bars pointing right indicate better performance than MEA, whereas red bars pointing left indicate worse performance than MEA. Solvents are blended at a ratio 3:1. (Experimental data reported by Zhang17). 3.2.3 Important issues As in conventional amine solvents, phase-change solvents exhibit a series of different issues that need to be investigated and addressed efficiently prior to their use in absorption/desorption processes. Such issues are reported in Zhang17 and Zhang et al.27,28 and summarized in Table 3. Table 3: Overview of issues that appear in the use of phase-change solvents and investigated solutions. A/A Issue 1

2

3

4 5

Impact Investigated solutions - Liquid-liquid separation - Increase of solvent concentration. LCST lower than in absorber. - Addition of another amine to absorption - Inhomogeneous mixing. increase solubility. temperature - -Mass transfer limitation. - Avoidance of contaminants. - Reduction of absorption - Increase of pressure drop. Foaming temperature. - Limited mass transfer. - Increase of turbulence. - Use of water scrubber in absorber. - Reduction of absorption temperature. - Solvent losses. - Ensure single liquid phase in Solvent volatility - Increased OPEX. absorber through cooling before absorber. - Use of water scrubber in absorber. - Loss of solvent efficiency. - Use stable solvents as promoters Thermal degradation in regeneration. - Generation of undesired products. Oxidative - Generation of undesired - Use of inhibitors. degradation products.


10

Page 13 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Highly performing phase-change solvents proposed in Zhang17 exhibit LCST lower than the temperature of the absorption process which is undesirable. Zhang17 considers and experimentally investigates few different options to address this issue and concludes that the best option is the addition of an amine that can increase the solvent solubility in water. In particular, several alkanols, alkanolamines and diamines are tested, leading to the conclusion that AMP (2Amino-2-methyl-1-propanol) is able to increase the LCST above 40oC, when used at concentrations of 8-10 wt.%, without affecting absorption or desorption performance of the phase-change solvent. In the DMCA+MCA mixture AMP is used at a ratio of 3:1:1 (DMCA+MCA+AMP). Foaming is observed in the experimental work reported by Zhang17 in amines functioning as promoters but not in the activator MCA or in DMCA+MCA+AMP, at the ratio proposed above. Among the different options investigated, foaming losses are reduced by up to 80% with the use of a water scrubber in the absorption step while the removal of contaminants after solvent regeneration also had a significant defoaming effect. Losses due to amine volatility is an additional issue in phase-change solvents, as in all amine solvents and it contributes to more than 90% in the overall solvent losses of the selected mixture DMCA+MCA. Zhang17 notes that such losses may be reduced by approximately 60% by reducing the absorber temperature by approximately 10oC, whereas volatility related losses may be reduced by 70% by introducing a single liquid phase in the absorber by cooling the lean CO2 amine from the decanter to reduce the stream temperature below the LCST and/or by using a solubility enhancer. This is because a large percentage of such losses is due to the second, supernatant organic phase that is formed in the phase-change solvent. It should be noted here that among the tested solvents, volatility-related losses of MCA are minor because of its fast kinetics and its much higher solubility in water compared to DMCA. In lipophilic solvents thermal degradation is not as important as it is for conventional amines because the regeneration takes place at approximately 80-90oC, much lower than the typical 120oC of the MEA process. MCA is less stable than DMCA at higher temperatures hence it is important to have DMCA at a higher ratio in MCA+DMCA mixtures because DMCA is a promoter and it is the most important of the two components at the higher temperature regeneration process. Furthermore, the thermal degradation of the amines studied by Zhang17 results in new amines which are still active and able to react with CO2 and be regenerated. However, thermal degradation should generally be avoided because the new amines, albeit capable of capturing CO2, change completely the performance of the absorption/desorption process. Oxidative degradation is a phenomenon occurring in CO2 capture amines including phasechange solvents due to the presence O2 in the flue gases. DMCA is again more stable than MCA which degrades mainly to cyclohexanone-related products. In this respect, Zhang17 tests several inhibitors with KNaC4H4O6·4H2O (potassium sodium tartrate tetrahydrate) being the most effective in depressing the degradation rate of MCA.


11


Page 14 of 61

3.3 Technologies for intensification of solvent regeneration Lipophilic solvents enable regeneration at 80oC, which reduces the thermal energy requirements in the desorption column. Their desorption rate can be enhanced by non-thermal approaches. Zhang17 and Zhang et al.26,27,29,30 investigate several such technologies including nucleation, ultrasound, agitation and extractive regeneration. An overview of their main characteristics is shown in Table 4. Table 4: Overview of regeneration intensification technologies proposed by Zhang17 and Zhang et al.26,27,29,30. Technology

Energy consumption [GJ/tn CO2]

Suitability

Notes

Blends with regeneration promoter at a higher ratio. -

- Desorbed CO2 is 4-20% lower Nucleation than with N2 gas stripping, but desorption rate is comparable or faster. Ultrasound 2.0 - The absorbed CO2 is approx. up to 30% more than in nucleation. Blends with absorption - Combined agitation and activator at a higher Agitation 2.0 nucleation increases regeneration ratio. rate by 10-20% compared to using each separately. - With the proper solvent the Inert solvents that are of lipophilic amine can be removed similar polarity to the at temperatures of 50-70oC. lipophilic amine, have a Extractive - Volatility losses of the inert 3.5 large boiling point regeneration solvent is also an issue hence difference, and are pressurized extraction may be chemically stable. beneficial. Nucleation pertains to the use of different porous materials to enhance the formation of CO2 bubbles during desorption, i.e. during the point where due to high temperature bicarbonate and carbamate species dissociate and the loaded solution becomes supersaturated. Zhang17 notes that the porous particles are inert hence they do not influence the chemical equilibrium. The use of nucleation can be more beneficial when a regeneration promoter is used at a higher ratio in blends with activators because otherwise regeneration will not be as efficient. Ultrasonic regeneration is based on the idea of using high frequency sound waves to enhance bubble formation in the solution during regeneration. The underlying rationale is that the ultrasonic wave initiates microscopic cavities that appear as bubbles which facilitate the transfer of dissolved CO2 from the solution. Agitation is another approach that facilitates the formation of bubbles through cavitation phenomena that appear due to the reduction of the solution surface tension. Increased agitation speed is beneficial for bubble formation and also for mass transfer and desorption rate. Agitation enables enhanced regeneration efficiency even in cases when the absorption activator is at higher


12

Page 15 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ratio than the regeneration promoter. As noted in Table 4, a hybrid technology that combines the nucleation and agitation approaches is the best option. Extractive regeneration is another approach discussed by Zhang17 where an inert solvent which is immiscible with water is used to separate the lipophilic organic phase from the loaded solution hence enhancing CO2 desorption due to shifting of the chemical equilibrium. The solvent can then be separated using distillation. The selection of an inert solvent which is highly selective toward the utilized amine is of major importance in this approach whereas other important characteristics are also noted in Table 4. The approach is similar to a liquid-liquid extraction process which becomes more efficient when more than one extraction stages are used. Note that despite the much higher regeneration energy of 3.5 GJ/on CO2 required in this case, the very low temperature enables the use of industrial waste heat to cover the energy needs. 3.4 Proposed flowsheets Based on the above regeneration technologies Zhang17 and Zhang et al.26, 27,29,30 propose few different process flowsheets for which they provide theoretical calculations of the associated energy requirements. Their main features are shown in Table 5 while their energetic requirements are analysed in Figure 7. The first flowsheet, called “TBS 1”, employs MCA as the solvent with the highest ratio in the MCA+DMCA+AMP mixture and considers that the phase change takes place inside the regenerator, while the regenerated solvent enters the absorber as one homogeneous liquid phase. Table 5: Proposed flowsheets and basic characteristics (Zhang17; Zhang et al. 26,27,29,30). Flowsheet name

Solvent used

TBS 1

MCA+DMCA+AMP with MCA at higher ratio than DMCA

TBS 2 + Agitation

MCA+DMCA+AMP with DMCA at higher ratio than MCA

TBS 1 + Extraction

As in TBS 1with pentane as the inert solvent.

Differences from standard MEA-based absorption/desorption flowsheet - Same base flowsheet as MEA. - Phase-change occurs inside the desorber at 𝑇 =80oC . - Regenerated solvent cooled to 30-40oC to enter absorber as one liquid phase. - The standard packed-bed desorber is replaced by a CSTR. - Regeneration takes place inside the CSTR at the same temperature as TBS 1. - A slightly different version includes a liquidliquid phase separator with porous particles prior to the CSTR. - Regeneration is performed by a multi-stage counter-current extraction process. - A distillation column is used for inert solvent recovery from the amine solvent. - A two-column process is also used for the separation of CO2 and inert solvent vapors.

The second flowsheet is called “TBS 2 + Agitation” with a major difference being in the desorber which is replaced by a CSTR with an agitator. The use of porous particles is proposed


13


Page 16 of 61

inside the CSTR to enhance regeneration through nucleation. Pre-heating is also proposed to facilitate the heating of the rich solvent in the CSTR. Inter-cooled absorber is also proposed as it is known to enhance CO2 loading and prevent amine vaporization31,32. A slightly modified version of this flowsheet is also proposed where a liquid-liquid phase separator is placed before the CSTR to remove the second lean-CO2 phase and hence reduce the rich CO2 stream entering the CSTR. The separator is enhanced with porous particles to create nucleation, while the separation temperature is at 60oC. The third flowsheet is called “TBS 1 + Extraction” and is based on the use of an inert solvent to recover the amine mixture through a multi-stage, liquid-liquid extraction process. An additional process is also used to purify the CO2 from inert solvent vapors. Although this process has much higher energetic requirements than the other two (Table 5), the regeneration temperature is at 6070 oC which enables the use of industrial waste heat and the complete avoidance of costly utilities with beneficial effects on operating OPEX. However, the addition of a multi-stage extractor and a two-column process for purification of CO2 generates significant additional CAPEX compared to the two previous processes. Furthermore, the selection of an appropriate inert solvent which is highly selective toward the amine solvent becomes rather complex when the latter is a blend. Figure 7 shows that energy savings achieved with these processes are mainly in the stripping energy and the sensible heat, while any savings in the heat of reaction are purely due to employing different solvents.

Figure 7: Energy requirements for flowsheets proposed by Zhang17 compared to the standard MEA process. (Data from Zhang17). 3.5 Experimental set-ups Zhang17 and Zhang et al. 8,26,27,28,29,30,33 perform numerous different experiments for solvent and regeneration technology screening purposes. Table 6 shows the most important features of the


14

Page 17 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


employed experimental set-ups and conditions. The screening and selection of solvents is performed at four different equipment types of increasing size:    

Test tubes are used for the preliminary screening of the initial solvent set of 44 amines/ A bubble column rig is used for the more detailed investigation of fewer solvents which initially showed good performance An absorption column is used only for 1 reference (MEA) and 2 selected solvents (MCA and DMCA+MCA+AMP). A complete pilot plant comprising an absorption column and a CSTR reactor as a regenerator is used for 3 selected solvents (DMCA, A1, DMCA+A1). A1 is not named in the paper30.

Table 6: Overview of experimental work reported by Zhang17 and Zhang et al.8,26,27,28, 29,30,33 A/A Experiment Purpose

1

2

3

4

5 6

7

Equipment type Characteristics and Conditions Screening and selection of solvents - 6 mL aqueous amine solutions, 𝐶𝑠𝑜𝑙𝑢𝑡𝑖𝑜𝑛= Preliminary 0.9- 4.2 M. screening of Absorption: 𝑇=25oC, 𝐹𝐶𝑂2=25 mL/min. Test-tube rig absorption and desorption solvent - Desorption: 𝑇=40-90oC with stepwise performance heating in thermal bath. - Absorption: 40 mL aqueous amine solution, 𝐶𝑠𝑜𝑙𝑢𝑡𝑖𝑜𝑛=3-5 M at 𝑇=40 °C, 𝑃𝐶𝑂2= 4-100 Absorption and Bubble column kPa, 𝑡𝑐𝑜𝑛𝑡𝑎𝑐𝑡=2-5 hr. desorption for rig - Desorption: Initially N2 stripping and then solvent selection magnetic agitation at 250-1000 rpm and 𝑇 =70-80 °C. - 𝐷𝑖𝑛𝑡𝑒𝑟𝑛𝑎𝑙 = 0.04 m, 𝐻 = 1 m, Packing: Raschig rings 5-8 mm, Porosity: 0.36-0.88. Bench-scale - Tested solvents: 30 wt.% MEA, 5M MCA, Absorption absorption 5.5 M DMCA+MCA+AMP (3:1:1.5). column experiments - 𝐹𝑔𝑎𝑠=10-75 L/min (𝐶𝐶𝑂2=15 mol %), 𝐹𝑙𝑖𝑞𝑢𝑖𝑑 =60-200 mL/min. - Absorber: 𝐷𝑖𝑛𝑡𝑒𝑟𝑛𝑎𝑙 = 0.025 m, 𝐻 = 1.45 m, Packing: Sulzer Ex 316, Bench-scale - Regenerator: 500 mL CSTR, 𝑇=50-95oC Absorption absorption column- CSTR - Tested solvents: 30 wt.% MEA, 4M regeneration DMCA, 2M DMCA+2M A1. regenerator experiments - 𝐹𝑔𝑎𝑠=300 L/hr (𝐶𝐶𝑂2=15 mol %), 𝐹𝑙𝑖𝑞𝑢𝑖𝑑=0.8 kg/hr. Thermal Glass test-tubes - 10ml CO2 saturated amine solution at 𝑇 degradation in oil bath =120 °C for 5 weeks. - 200 mL amine solution at 𝑇=50 °C, 𝐹𝐶𝑂2 = Oxidative Glass bubble 2 mL/min, 𝐹𝑂2 = 98 mL/min, 𝐶𝑂2=98 vol. degradation column %, 𝑇 = 50 oC. Screening of solvent regeneration technologies Agitation for Heating plate - Regeneration: 50 mL CO2 loaded solution


15


regeneration intensification

stirrer, glass bubble column

8

Nucleation for regeneration intensification

Glass bubble column with different solids

9

Ultrasonic assisted desorption

Ultrasonic bath, test tubes

-

-

-

Screening of inert solvents for extractive regeneration

10

Double wall glass reactor

-

Page 18 of 61

at 𝑇=70-85 °C. Preparation of CO2 loaded solution: 320 mL aqueous amine solutions, 𝐶𝑠𝑜𝑙𝑢𝑡𝑖𝑜𝑛= 3-5 M 𝐶𝐶𝑂2=15 mol %, balanced with N2 at 𝑇=30 °C, 𝑡𝑐𝑜𝑛𝑡𝑎𝑐𝑡=3-5 hr. Solids tested: Silica bids, Al2O3 spheres, active carbon spheres, zeolite chips, PTFE boiling stones, ceramic Raschig rings and Berls-Saddles, molecular sieves. 10 mL loaded CO2 solution at 𝑇=30 °C, ultrasonic frequency 37 kHz. Extraction: 50 mL CO2 loaded solution extracted with the same volume of inert solvent, 𝑡𝑐𝑜𝑛𝑡𝑎𝑐𝑡=2-4 hr, solution agitated externally by magnetic stirrer at 600 rpm. Preparation of loaded solutions: 500 mL glass bubble column, amine aqueous solutions, 𝐶𝑠𝑜𝑙𝑢𝑡𝑖𝑜𝑛= 3-4 M, 𝐹𝑔𝑎𝑠 = 300 mL/ min, 𝐶𝐶𝑂2=30 mol%, balanced with N2 at 𝑇 =30 °C.

4. The DMXTM process 4.1 Main concepts The DMXTM process is based on the DMX-1 solvent which enables a liquid-liquid phase split after reaction with CO2 and heating, prior to desorption34. The solvent is a blend of amines which has been selected and tested at multiple different levels, including: -

-

Experimental high-throughput screening of multiple different solvent options12,35 through an appropriate apparatus36. Experimental equilibrium investigations of the DMX-1 solvent together with other operating performance requirements such as mechanical separation (decantation)34, kinetics with respect to both CO2 and CO37, degradation and corrosion in the presence of CO234 or CO2 and CO37. Model-based technoeconomic assessment of process flowsheets34,35,38and application in different industries37,38,39. Experimental pilot plant investigations37,38.

The DMX-1 blend is developed by selecting amines based on the following criteria38: -

-

LCST is chosen higher than the maximum absorber temperature to avoid two liquid phases in the absorber. The blend concentration is chosen so that the concentration of the absorbed CO2 is significantly higher in the heavy, CO2-containing, liquid phase than in the light, aminecontaining phase in order to send the former phase into the absorber. The solvent exhibits low heat of reaction, high CO2 capacity and fast kinetics.


16

Page 19 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


4.2 Screening for DMX solvents In order to select appropriate solvents for the DMXTM process, Aleixo et al.12 investigate the phase behavior, at different temperatures, of more than 300 amines. Form such extensive screening tests, only amines that are miscible with water at room temperature and exhibit liquidliquid phase separation upon CO2 loading and/or increase of temperature are classified as potential DMXTM solvents. For such solvents, phase behavior diagrams are presented which reveal the solution behavior upon CO2 loading and increase of temperature. The results reveal that tertiary amines may exhibit liquid-liquid phase separation due to CO2 loading, but the effect of temperature is not pronounced in all cases. Temperature may facilitate the liquid-liquid demixing at lower amine concentrations and lower loading conditions. Considering primary and secondary amines, the investigated solutions remain homogeneous upon CO2 loading at low temperatures. Furthermore, all the studied systems of this kind that exhibit liquid-liquid phase separation due to increase of loading are also phase separated due to temperature increment. Interestingly, the phase separation of all potential DMXTM solvents, which are selected from the aforementioned screening tests, results in one amine rich phase that is almost free of CO2 and an aqueous phase with high CO2 loading. Finally, Aleixo et al.12 compare the CO2 solubility isotherms for various solvents. Such comparison reveales that the solvents, which present similar loading capacity, may show different cyclic loading capacity and, thus, may require different flow rates in the stripper. 4.3 Operational investigations for DMX-1 With respect to the DMX-1 solvent, Raynal et al.34 present a wide set of experimental and computational investigations regarding important operational issues prior to using the solvent in a pilot plant. The investigations pertain to liquid-liquid separation, degradation and corrosion. Furthermore, Dreillard et al.37 investigate the kinetic and corrosion behavior of DMX-1. Table 7 presents an overview of experiments and investigated conditions. Liquid-liquid separation experiments include a stirred-cell and a lab-scale decanter, supported by CFD simulations. The experimental results show a clear phase separation for DMX-1 in less than 1 minute, generating a transparent light phase and an opaque dense phase. The CFD calculations are able to reproduce the experimental results with good accuracy, indicating that it is an adequate tool for scaling-up purposes. To test the degradation of DMX-1, Raynal et al.34 perform laboratory tests in closed reactors at high temperature and pressure and in CO2 and O2 for a period of 15 days. Both fresh and degraded solvent are used to check the impact of impurities on interfacial properties and coalescence efficiency38. The tested temperatures range between 140-180oC, while the CO2 and O2 pressures are set to 20 and 4.2 bar, respectively. The temperature is considerably higher than the 120oC of the MEA process because, according to the authors, it enables the use of high pressure regeneration, hence the CO2 compression costs are reduced. MDEA


17


Page 20 of 61

(Methyldiethanolamine) is known to exhibit good stability hence it is used as a reference solvent for comparing the degradation of DMX-1. The results show that the degradation rate of DMX-1 is 7 times lower than that of MDEA. Further investigations of degradation, reported for a pilot plant in Raynal et al.38, indicate a 3 times lower rate than MEA. Corrosion tests are performed with DMX-1 in a pilot plant and in a dedicated autoclave under harsh conditions (see Table 7 for details). The obtained results indicate stainless steel corrosion rates of 2 μm/year and 10 μm/year for the corresponding technologies, respectively. Raynal et al.38 further note that the tests with both 304L stainless steel and carbon steel indicate similar, very limited corrosion hence the use of carbon steel columns is possible, resulting in an estimated 10 % reduction in CAPEX. Tests in two different pilot plants also unveil foaming issues that are addressed by changing of packings38. EX-type packing results in foaming in the absorber and replaced by Raschig-type packing which is less efficient but avoids intense foaming. DX-type packing also enables good hydrodynamic performance, although it is less efficient than the EX type. Pilot plant tests also indicate significant emission reductions up to 80% in VOC, NH3 and amines, however Raynal et al.38 notes that a downstream washing unit is necessary to ensure that the vented gasses are within environmental specifications. Table 7: Overview of experimental equipment, purpose of experiments and conditions used for the DMXTM process and solvent. Source

Experiment Purpose

Equipment type

Raynal et al.34

Liquid-liquid separation behavior

Stirred cell

Raynal et al.34

Liquid-liquid separation behavior

Lab-scale decanter

Raynal et al.34

Solvent stability/ degradation

Closed lab reactors

Dreillard Solvent stability/ et al.37 degradation

Solvent Degradation Rig (SDR)

Raynal et al.34

1) Dedicated autoclave 2) Pilot plant

Solvent corrosivity

Characteristics and Conditions Reported - Stirrer and cell diameters: 2.5 and 6 cm. - Stirrer rotating speed: 11000 rpm - Time to achieve two clear liquid phases: less than 1 min. - Pre-heated and pre-loaded DMX-1 solvent - 𝐹𝑙𝑖𝑞𝑢𝑖𝑑 = 20 kg/hr at controlled temperature - 𝑇=140-180 °C, 𝑃=3 up to 20 bar CO2 and/ or 4.2 bar O2. - Duration: 15 days Absorption conditions - 𝑇=60oC, 𝑃=3 bar, duration 60 min. Regeneration conditions - 𝑇=150oC, 𝑃=5.5 bar, duration 90 min. Pause conditions - 𝑇=60oC, 𝑃=5.5 bar, duration 90 min. - Overall duration: 340h 1) 𝑇=140-180 °C, 𝑃=20 bar CO2, 2) Stripper 𝑇𝑖𝑛𝑙𝑒𝑡 = 𝑇𝑜𝑢𝑡𝑙𝑒𝑡 = 105oC, - Both cases: Tests with 304L stainless steel and carbon steel coupons.


18

Page 21 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


- 𝑉𝑟𝑒𝑎𝑐𝑡𝑜𝑟= 375 cm3 - 𝑇=40oC,90oC - 𝑃𝐶𝑂2=100-800 mbar, stirrer speed 100 Dreillard Solvent kinetics Lewis cell 37 et al. rpm, duration 1 hr. - 𝑃𝐶𝑂= up to 5 bar, stirrer speed 200 rpm, duration 18-20 hr. - Up to 𝑇=150 °C, 𝑃= 5 bar Capture process Raynal Complete pilot assessment - Duration: 1500 hr of continuous et al.38 plant operation - 𝐹𝑔𝑎𝑠= 171.3 L/hr Capture process Dreillard Complete pilot - Hcolumn= 1m, Dcolumn=5 cm. assessment et al.37 plant - Duration: 1500 hr of continuous operation 37 The work of Dreillard et al. aims to identify the potential of using the DMXTM process in blast furnace gasses which include CO2 and CO at high pressures and lack of O2. The authors firstly investigate the solvent kinetics in the presence of these gases in a Lewis cell which is suitable for fluids exhibiting slow kinetics. The results show that 95% of CO2 is absorbed in approximately 1 hour, but after 20 hours only 5% of CO is absorbed. Results obtained from comparing DMX-1 with MEA show that as CO is injected in the cell the solubility of CO is higher in DMX-1 at the beginning. But as the experiment proceeds, the CO solubility in DMX-1 becomes much lower than the solubility in MEA. The reaction rate in MEA is therefore much lower than in DMX-1. While the above experiments take place with CO2-free solvents, CO2-loaded solvents are also considered. These experiments show that the amine-CO reaction is inhibited by the presence of CO2. The investigations conclude that at industrial scale the CO absorption is likely to be insignificant. This is important because CO can be recovered and utilized. Degradation experiments are also performed in a Solvent Degradation Rig (SDR) composed of 6 parallel, stirred reactors37. The tests take place at the operating conditions of the DMXTM process. Main conditions are shown in Table 7, whereas gas mixture compositions and flows are detailed in Dreillard et al.37. The tests indicate that MEA is greatly affected by the presence of O2, while the impacts of high temperature (150oC) and CO are insignificant. On the other hand, DMX-1 is resilient under these conditions and gases 4.4 Pilot plant results and technoeconomic studies The key characteristic of the DMXTM process lies in the placement of the mechanical separator after the intermediate rich/lean heat exchanger38 in an otherwise typical absorption/desorption process layout, without the regeneration intensifications proposed in Zhang17. Furthermore, the desorption process may operate up to 150oC (Raynal et al.38) and 5 bar pressure due to solvent resilience to degradation, in contrast to the lipophilic solvents of Zhang17 which operate at desorption temperatures of 60-90oC. Reynal et al.38 do not provide indications regarding the pilot plant capacity, but work with DMX-137 notes that the pilot plant gas flow is 171.3 L/hr. Pilot plant results indicate regeneration energy requirements of 2.5 GJ/ton CO2, with simulations indicating that they can reach 2.3 GJ/ton CO2. Corresponding evaluations of Zhang17 indicate 2.0 GJ/ton CO2 at best. Other important conclusions regarding the performance of the DMXTM process include a decrease in the energy from -11.6%pts (when MEA is used) to -9.1%pts when


19


Page 22 of 61

heat integration is considered within the DMXTM flowsheets, a 23% decrease in the cost of avoided CO2 and an increase of the Levelized cost of Electricity of 48% for DMXTM, which is much better than the 60% of the MEA-based process. There are several process characteristics that enable the above performance gains. The liquidliquid phase separation reduces the liquid flowrate that enters the stripper by approximately 2530%, while flashed CO2 is also collected directly from the decanter top which can also bypass the first compression stage due to the high-pressure process. Furthermore, carbon steel can be used as column material instead of the much more expensive stainless steel. These characteristics enable CAPEX and OPEX reductions in the compression, absorption and stripping sections38. It is worth noting that the decanter requires some extra auxilliary equipment such as tanks and pumps and the DMX-1 solvent is less reactive and more expensive than MEA. Despite such shortcomings, the overall DMXTM performance is better than MEA in major key indicators. Further pilot plant experiments are performed by Dreillard et al.37 for blast furnace gases which are rich in CO2 and CO and lack O2. The investigated case is called “Top Gas Recycle” (TGR), where blast furnace gases are sent to a CO2 capture process and purified gases that contain reducing gases (CO and H2) are re-injected in the blast furnace. The experiments are performed using the DMXTM and HICAPT+TM (40%wt MEA) processes. The latter includes a proprietary oxidation inhibitor on top of the increase ME concentration40. Main conditions and characteristics are reported in Table 7. Although pressures and temperatures are not reported in this case, it is assumed that for the DMXTM process they are the same as the ones reported for degradation experiments. The results indicate that the CO capture rate is around 0.03% for both solvents whereas the CO2 capture rate is 99.9%. In this respect, the solvents are highly selective toward CO2, while the CO can be recovered and valorized. Gomez et al.39 perform a technoeconomic evaluation of the DMX-1 solvent and the DMXTM process for flue gases emitted from a power plant, a waste incineration plant and a cement plant. The results are compared with the case of using the standard 30% MEA process for the corresponding plants. Figure 8 presents the CO2 capture cost for each case, indicating the benefits of the DMXTM process. The cost savings start from approximately 15% for the power plant and reach approximately 50% for the cement plant. For the cement plant there are two main reasons for cost reduction. Firstly, the costs of steam required for DMXTM are much lower than for MEA. Furthermore, the capital investment in the boiler required for steam generation in the DMXTM case is also lower.


20

Page 23 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 8: Performance comparison of the DMXTM and conventional MEA processes for different CO2 emission plants (Data from Gomez et al.39) Dreillard et al.37 perform a techno-economic analysis of the HICAPTTM (30 wt.% MEA), HICAPT+TM and DMXTM processes for blast furnace flue gases. The first case is the TGR, explained previously, while the other two are the BF and PWS. In the BF case, the purified furnace gases are burnt in a power station (instead of recycling them in the blast furnace), whereas in the PWS case the blast furnace gases are burned to the power station and the power station flue gases are then purified from CO2.The results for the total erected costs of the CO2 capture unit per case and process are shown in Figure 9. It appears that DMXTM exhibits much lower erection costs for the TGR and BF cases, whereas HICAPT+TM is only slightly better for the PWS case. Furthermore, the thermal energy requirements of DMXTM are lower by 20-25% compared to HICAPTTM and by 13-18% compared to HICAPT+TM. Finally, with the DMXTM it would be possible to produce CO2 at around 40€/ ton, while by better waste heat recovery in the emitting plant, this could drop to 27€/ ton.

Figure 9: Performance comparison of the HICAPTTM, HICAPT+TM and DMXTM processes for different CO2 emission characteristics (Data from Dreillard et al.37)


21


Page 24 of 61

5. The DEEA (2-(diethylamino)-ethanol)/ MAPA (3-(methylamino)-propylamine) solvent and process 5.1 Overview The DEEA/MAPA solvent is a binary mixture of amines which exhibits liquid-liquid phaseseparation upon reaction with CO2. The mixture strikes a good balance between reaction kinetic and CO2 loading requirements. MAPA combines a primary and a secondary amine group exhibiting high reaction rate41, while DEEA is a tertiary amine exhibiting a high absorption capacity1. The mixture has been investigated widely at various different levels including the following: -

-

Experimental measurement of important properties such as single component and mixture solubility in H2O42 and in H2O/CO243,44, speciation characteristics19, heat of absorption21,45 and freezing point depression46. Experimental, lab-scale investigation of absorption and desorption behavior20. Development of a UNIQUAC-based model47 for equilibrium data predictions48. Pilot plant absorption/desorption investigation, modeling49 and technoeconomic 50 assessment .

5.2 Experimental investigations and set-ups Hartono et al.42 generate experimental equilibrium-based data for MAPA/DEEA as shown in Table 8, while the data are also used to fit interaction parameters necessary for UNIQUAC. The corresponding data are then predicted and compared with the experimental results. The investigations result in several important insights which can be very useful in case of process level application of this solvent. DEEA is more volatile than MEA and MDEA, while it exhibits an azeotrope at a mole fraction of less than 0.1 in H2O at all the investigated temperatures. On the other hand, model predictions show an azeotrope for MAPA at water mole fraction of 0.7-0.9 at 40oC. This result is based on model predictions but is not investigated experimentally. MAPA is more volatile than DEEA in pure form but it exhibits a lower volatility than DEEA and MEA in aqueous solutions. The addition of MAPA in DEEA reduces the volatility of both components compared to having the corresponding pure aqueous forms. Information regarding equilibrium investigations performed by Arshad et al.43,44 is shown in Table 8. Aqueous DEEA solutions exhibits a high cyclic capacity and a low heat of absorption as noted in Arshad et al.21,45. Despite these advantages, the slow reaction kinetics may lead to a very large absorber. MAPA has fast kinetics and exhibits a high CO2 loading capacity at absorption temperature, but this high capacity is maintained at 120o C hence its cyclic capacity is low compared to DEEA and its heat of absorption is high21,45 with detrimental effects in regeneration energy requirements. MAPA reacts faster in mixtures with DEEA, while it remains the dominating reacting component at high temperatures. DEEA and MAPA mixtures exhibit high CO2 partial pressures at low CO2 loadings which may enable regeneration under pressure hence supporting reductions in compression costs.


22

Page 25 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


With respect to heats of absorption, important insights derived from Arshad et al.21,45 indicate that the heat of absorption depends on the concentration of MAPA which is the component exhibiting faster kinetics. The mixture with higher MAPA concentration (2M in 2M DEEA) exhibits lower heat of absorption, but at temperatures of 120oC both the 1M and 2M MAPA mixtures exhibit very similar heat of absorption. Additional important observations from Arshad et al.21,45 pertain to the appearance of the liquid-liquid phase separation which depends on the CO2 loading and the temperature. The 2M MAPA/ 5M DEEA mixture exhibits 1 phase at 40oC and 2 phases at 80 and 120oC for loading ranges reported in Table 8. The leaner MAPA mixture exhibits phase-change only at 120oC. The lower, rich in CO2 phase is more viscous than the lean phase in both mixtures, while the loaded mixtures are also more viscous than their pure component constituents. Table 8: Purpose of investigations, equipment types, data types and corresponding components and conditions for the MAPA/DEEA solvent. Purpose of investigation/ Equipment type

Type of reported results Phase compositions.

- Evaluation of equilibrium - Ebulliometer apparatus - Liquid Chromatography-Mass Spectroscopy (LCMS) (Hartono et al.42)

- Evaluation of equilibrium - Thermally insulated jacked reaction calorimeter (Arshad et al.43,44) - Heat of absorption - Thermally insulated jacked reaction calorimeter (Arshad et al.21,45)

Activity coefficients. Freezing point depression. Pure component saturation pressure. Excess enthalpy.

Components and conditions DEEA/H2O: 𝑇=40-95°C, 𝑃=10-100 kPa MAPA/H2O: 𝑇=40-100°C, 𝑃=10-100 kPa DEEA/MAPA: 𝑇=80-120oC, 𝑃=0-60 kPa DEEA/MAPA/H2O: 𝑇= 40-110oC, 𝑃=6-111 kPa DEEA/H2O: 𝑇=50-95°C MAPA/H2O: 𝑇 =40-95°C DEEA/MAPA/H2O: 𝑇 =40-110°C DEEA/H2O: 𝑇=-30-0°C MAPA/H2O: 𝑇=-30-0°C DEEA/MAPA/H2O: 𝑇=-30-0°C DEEA: 𝑇=65-160oC, 𝑃=3-100 kPa MAPA: 𝑇=56-139oC, 𝑃=4-101 kPa DEEA/H2O: 𝑇=25°C

Pressure and solubility.

DEEA/CO2/H2O, MAPA/CO2/H2O, DEEA/MAPA/CO2/H2O: 𝑇=40°C, 80°C, 120°C, 𝑃=50-600 kPa

Differential and integral heat of absorption.

DEEA/CO2/H2O, MAPA/CO2/H2O, DEEA/MAPA/CO2/H2O: 𝑇=40°C, 80°C, 120°C, 𝑃=6 bar (CO2 feed pressure)

Liquid-liquid phase-change data.

5 M DEEA/ 2M MAPA (in CO2 and H2O): 2 liquid phases at 𝑇=80°C and 120°C and loadings 0.44-0.31. 5 M DEEA/ 1M MAPA (in CO2 and H2O): 2


23


Page 26 of 61

liquid phases at 𝑇=120°C and loading 0.23. - Absorption and desorption behavior - Glass volume graded desorption vessel with IR-CO2-analyzer. (Pinto et al.20)

Absorption rates, volume and mole 5 M DEEA/ 2 M MAPA (in CO2 and H2O) fraction 𝑃𝐶𝑂2= 1-20 kPa, 𝑇=40oC, 60oC, 80oC, 𝑃𝑡𝑜𝑡𝑎𝑙= 0-6 distribution, CO2 content in bar. phases, species concentrations, total pressure.

- Characterization of species - NMR spectrometer (Ciftja et al.19)

NMPRspectra.

DEEA/CO2/H2O, MAPA/CO2/H2O: 𝑇=25°C DEEA/MAPA/CO2/H2O: 𝑃=1-20 kPa

Freezing point depression.

DEEA/H2O: 𝐶𝑎𝑚𝑖𝑛𝑒=0-55 mass% MAPA/H2O: 𝐶𝑎𝑚𝑖𝑛𝑒=0-32.5 mass% DEEA/MAPA/H2O =𝐶𝑎𝑚𝑖𝑛𝑒=0-45 mass% (at various ratios). DEEA/CO2/H2O: 𝐶𝑎𝑚𝑖𝑛𝑒=12, 20, 20, 33 mass% MAPA/CO2/H2O: 𝐶𝑎𝑚𝑖𝑛𝑒=10, 20, 27 mass%

- Freezing point depression - Modified Beckmann apparatus (Arshad et al.46)

Ciftja et al.19 perform a quantitative speciation investigation of the DEEA/MAPA mixture using NMR spectroscopy. With respect to the liquid phases they find that the lower phase is rich in both CO2 and MAPA, while the lean-CO2 phase is rich in DEEA. The speciation study of Ciftja et al.19 indicates primary and secondary carbamate, dicarbamate and carbonate-bicarbonate in the mixed solvent. The upper phase contains very little carbamate and dicarbamate and no bicarbonate–carbonate. The 2M MAPA/ 5M DEEA mixture is investigated more thoroughly in absorption and desorption experiments conducted by Pinto et al.20. It is worth noting that this work reports liquid-liquid phase separation at 40oC, whereas Arshad et al.21,45 reports a single liquid phase at 40oC. Interesting insights are generated regarding the best way of exploiting this mixture at the process level. It is possible to operate the regeneration at 120oC and produce CO2 at 6-8 bars, which has direct energy saving benefits at the downstream CO2 compression process. As CO2 is generated at a high partial pressure, the regeneration can also be operated at a temperature lower than 100oC hence avoiding an increase in the regenerator pressure. This would allow operation with lower heat quality with positive effects in energy consumption and solvent degradation. This behavior appears because MAPA is first loaded very fast and then when DEEA starts loading, it is transferred to the rich-CO2 phase. When this happens the heat of absorption drops significantly, enabling energy savings. Such savings are added to the savings obtained from introducing only the rich-CO2 phase in the regenerator (i.e. reducing the processed solution amount) by recycling the lean-CO2 phase to the absorber, prior to regeneration. Arshad et al. 46 develops a method and measures the freezing points of the DEEA/MAPA solvent. The obtained results indicate that MAPA shows a stronger non-ideal behavior than DEEA which exhibits interactions with water similar to those of MEA. A model in the form of a correlation is also proposed for freezing points of unloaded systems as a function of solution


24

Page 27 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


concentration. Table 8 shows the MAPA/H2O and DEEA/H2O amine concentrations for which prediction are valid with the developed model. Prediction of freezing points can be used for the determination of water activity coefficients which are directly related to equilibrium calculations that affect process decisions. 5.3 Pilot plant experiments, modeling and design studies Pinto et al.49 investigates the performance of DEEA/MAPA in a pilot plant. The plant consists typically of an absorber, an intermediate heat exchanger and a desorber. For this solvent the liquid-liquid phase separator is placed before the intermediate heat exchanger, as shown in Figure 10. The absorber and stripper heights are 4.23 and 3.57 m respectively with internal diameters of 0.1 and 0.15 m. The gas and liquid flows are approximately 90m3/hr and 180 L/hr, respectively. The authors follow a systematic plant start-up procedure. The plant is initially loaded with the necessary solutions, it is left running for 1 day, the samples are analyzed with respect to total alkalinity and compared to standard lab solutions and the solution concentrations are adjusted by addition of appropriate solutions. The plant is left running for 1 more day and the procedure is repeated until the required concentration is reached. Initially the plant is run in manual mode until fluctuations in various operating parameters are stabilized.

-

-

Figure 10:Typical, conceptual flowsheet of a process exhibiting phase-separation at absorption conditions

During operation, the temperature at the absorber bottom is approximately 60-65oC with a loading of 0.9-1.04 mol CO2/mol MAPA, with the latter being the highest achieved loading. Pinto et al.49 notes that with better control over the lean amine and the inlet gas streams the loading could reach 1.15 mol CO2/mol MAPA. For the observed loading region the reboiler duty is at 2.2-2.4 GJ/ton CO2. These results are obtained for desorber bottom temperature between 93103oC, with temperatures being lower by 10-15oC at a height of above 1m. Phase separation as


25


Page 28 of 61

well as CO2 stripping are generally easy and fast, with the authors proposing that a 2-stage flash could be used to replace the stripper. No foaming or operational problems are observed, except for volatile losses. Some corrosion is observed, with the iron content being at the same levels as a corresponding conventional MEA process, with metal concentrations stabilizing after 3 weeks of operation. The detection of high concentrations of additional metals is eventually attributed to retrofitting work that takes place during the investigations, without properly removing welding leftovers. Based on experimental results Pinto et al.49 then develop a model to predict the partial CO2 pressure as a function of loading using the soft model in the CO2SIM simulator. The DEEA/MAPA mixture is considered as a one phase, single pseudo-component. The mass transfer in the absorber is also modelled using an enhancement factor model, and the DEEA/MAPA system is assumed to be in the fast pseudo-first order regime, modeled through a simple reaction rate dependent on loading, based on experimental data. The obtained results are in good agreement with the experimental data, with mass transfer rates under-predicted only at low CO2 partial pressures. The CO2SIM simulator is further used by Liebenthal et al.50 to evaluate the potential of integrating the DEEA/MAPA process with power plants. This work pertains to the integration of a CO2 capture plant and the downstream compression train for retrofitting of an existing power plant and for greenfield design of a new power plant. The obtained results indicate a reboiler heat duty of 2.4 GJ/ton CO2 and efficiency penalties of 6.3 and 5.9 %-points for the retrofit and greenfield cases, respectively. Arshad et al.48 present a detailed thermodynamic model for equilibrium predictions of the DEEA/MAPA solvent. The model is based on the extended UNIQUAC framework and includes energy interaction parameters for a large number of important molecular and ionic species pertaining to DEEA/MAPA/CO2/H2O mixtures. The model exhibits good agreement with available experimental data. 6. Other important solvents and mixtures 6.1 DEEA-based mixtures Table 9 provides a summary of the main experimental approaches used in the works reviewed in section 6. Xu et al.51 investigates 5 amines as potential phase-change solvent candidates in order to determine if they exhibit an LCST and their absorption and desorption rates with respect to loading and the corresponding CO2 capacities. They note that 2M DIPA (2M Diisopropylamine) shows the best performance, but none of the considered amines exhibit phase separation upon CO2 loading at 40oC. On the other hand, 3M TEA (Triethylamine) and 5M DAA (Diallylamine) results in two phases after desorption, with the latter maintained after cooling at room temperature. Performing similar experiments, Xu et al.52 investigate an additional set of 12 pure or mixed amine solvents in order to identify the ones that exhibit liquid-liquid phase change after absorption. In this work, DEEA is combined in mixtures with few different linear primary or


26

Page 29 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


secondary amines, apparently with the aim to enhance the mixture reaction kinetics. On the other hand, MAPA is also combined with few tertiary amines as a means of exploiting the fast MAPA kinetics and enhancing the mixture with co-solvents exhibiting high absorption capacity. The authors find that mixtures of DEEA with EEDA (N-Ethylethylenediamine), DAB (1,4Diaminobutane), DMBA (N,N-Dimethylbutylamine) and MAPA with TEA (Triethylamine) and DMBA (N,N-Dimethylbutylamine) exhibit phase change behavior. The authors conclude that the mixture 2M DEEA and 4M DMBA may be a suitable a CO2 capture candidate, but it should be noted here that both these amines are tertiary with potential implications on kinetics. It is also worth noting that based on the reported liquid-liquid phase characteristics, the rich CO2 phase of DEEA/DMBA has the highest loading, followed by DEEA/DAB with DAB also exhibiting good performance in other indicators such as absorption rate. Xu et al. 53 perform a detailed experimental investigation of the 1-4- Butanediamine which is the same as DAB reported in Xu et al.52, but it is now abbreviated as BDA. The aim of the investigation is to determine the absorption and desorption rates and loadings, while quantitative NMR is also used to determine the molecular and ionic species and their concentrations in different phases. The authors find that there is a critical CO2 loading of 0.313 mol/ mol amine above which the CO2 concentration is constant in the lower, CO2-rich phase. Furthermore, BDA has the highest concentration in the lower phase at the critical loading, while it does not change significantly in the upper or lower phases above the critical CO2 loading. The authors also find that the liquid-liquid phase separation is due to the limited solubility of DEEA in BDA and CO2 reaction products and the lower reaction rate of DEEA compared to BDA. The phase separation starts at 0.099 mole CO2/mole amine, whereas the equilibrium loading is at 0.505 mol CO2/ mol amine and 97.4% of the CO2 resides in the lower phase. Finally, it is observed that the 2M BDA/ 4M DEEA mixtures exhibits the best performance with: - 46% higher cyclic loading than 5M MEA, - 48% higher cyclic capacity than 5M MEA, - 11% higher cyclic efficiency than 5M MEA. Note that cyclic efficiency is defined as the ratio of cyclic capacity over rich loading. Table 9: Main features of the experimental work performed in the publications reported in this section. Source Xu et al.51,52

-

Xu et al53

-

Experiment Purpose Absorption and desorption rates, Loading and capacity Absorption and desorption capacities, Rich and lean loadings, Composition of liquid phases after absorption

Main equipment type

Characteristics and Conditions Reported

Glass reactor in water batha

- 𝑉𝑟𝑒𝑎𝑐𝑡𝑜𝑟= 150 ml - 𝐹𝑔𝑎𝑠= 463ml/min - 𝑇= 40oC, 80oC

Same as above, Ion chromatography, NMR

-

𝑉𝑟𝑒𝑎𝑐𝑡𝑜𝑟= 100 ml 𝐹𝑔𝑎𝑠= 463ml/min (absorption), 𝐹𝑔𝑎𝑠= 874ml/min (desorption), 𝑇= 40oC, 90oC


27


-

Reaction rate constants, Reaction mechanism

Xu et al. 54 Xu et al.55

- Phase composition - Cyclic capacity and loadings, reaction products - Absorption rates

Xu et al.56

-Absorption rate and loadings

Wang et al.57 Wang et al.58

- Constructed from Stainless steel tube, 11cm height, 1.2 cm diameter, enclosed in glass tube of 31 cm diameter. Wetted wall - 𝐹𝐵𝐷𝐴/𝐷𝐸𝐸𝐴= 5 L/min column (WWC) - 𝑇= 25oC, 40oC, 60oC - 𝑃=3-35 kPa - 𝑇= 25oC, 40oC, 60oC (density and Density and solubility viscosity - 𝑇= 25oC, 30 oC, 40oC, 50oC, 60oC apparatuses, glass, (viscosity) jacketed stirred - 𝐶𝐵𝐷𝐴= 1.95-14.3 mol% reactor for - 𝐶𝐷𝐸𝐸𝐴= 2.01-19.3 mol% solubility - 𝐶𝐵𝐷𝐴/𝐷𝐸𝐸𝐴= 3.63-16.7 mol% - Ion chromatograph - Experimental ranges and conditions are y generally similar to the previous - Same as (1) and investigations reported above. NMR - WWC - Constructed from stainless steel tubes, 9.1 cm height, 1.26 cm outer diameter - WWC - 𝐹𝑔𝑎𝑠= 500 ml/min, 25 ml/sec - 𝑇= 30oC, 60oC

- Mass transfer coefficients -Absorption and desorption behavior - Composition of phases after absorption

Ye et al.59,60

-

-

WWC

Absorption - Laboratory bubbling column Desorption - High pressure Parr reactor - NMR

- Absorption loading - Phase behaviour - Reaction mechanism

Zhou et al.61

Luo et al.62

Density, viscosity, solubility of N2O

Page 30 of 61

Laboratory bubbling reactor NMR

CO2 solubility Solubility - Thermostatic in mixture, distribution of VLE reactor components, Absorption compositions of capacity, analysis phases, of components cyclic loading - Sealed flask, and cyclic separating

- Same as Wang et al. (2017a) Absorption - Vbub.col.=150 ml - 𝐹𝑔𝑎𝑠= 105 ml/min - 𝑇= 30oC Desorption - Vreactor=500ml - Used CO2 loaded samples from absorption - 𝑇= 80oC -

Height 25 cm, diameter 2.5 cm 𝐹𝑔𝑎𝑠= 60 ml/min Vliquid.=25 ml 𝑇= 50oC

Solubility - Vreactor=400 ml - T=40oC-120oC - 𝑃𝐶𝑂2= up to 400 kPa Absorption capacity, analysis of components - Size details are not reported - Conditions same as above


28

Page 31 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


capacity -

Wang et al.63

-Kinetic behavior -Absorption rate

funnel, Adsorption packed column, Thermal conductivity detector. -Laboratory bubble reactor -NMR -WWC

-Solubility measurement apparatus - Laboratory absorption stirring cell - Phase separation - Laboratory and composition regeneration - Corrosion studies apparatus - Laboratory apparatus for corrosion study - 6-port semi-batch - Absorption absorption capacity apparatus - NMR

- Vreactor=50 ml - 𝐹𝑔𝑎𝑠= 500 ml/min - WWC Constructed from stainless steel tubes, 9.1 cm height, 1.26 cm outer diameter - T=60oC

Machida -Solubility of CO2 64,65,66 et al.

- T=40oC-90oC - 𝑃𝐶𝑂2= 1 to 101 kPa

Hu et al.67

Absorption apparatus - Cell inner diameter 100 mm, depth 100 mm - CCO2=0-1200 mol/m3 - T=25-45oC (absorption) - T=95-125oC (desorption)

Kim et al.68

Barzagli et al.69

- Loadings - Glass cylinder - Absorption - Pilot plant efficiency - NMR - Working capacity

Zhuang - VLE data and Clements70 Tzirakis et al.71

Zhang et al.72

- T=40oC - 𝐹𝑔𝑎𝑠= 200 ml/min Batch absorption/ desorption experiments - Cylinder 56 mm diameter and 300 mm height - T=40oC, 110oC Absorption/ desorption pilot plant - Cylinders of 56 mm diameter and 400 mm height - T=40-50oC, 110-120oC - 𝐹𝑔𝑎𝑠= 29.0 dm3 h−1

- Batch mode gas- Vequipment=660 ml, 1245 ml liquid apparatus

- VLLE data

- Glass bubbling cylinders

-

- VLE data - Absorption rate - Mass transfer resistance

- Glass bubbling reactor - Double-stirred cell reactor - NMR

-

Vequipment=250 ml 𝐹𝑔𝑎𝑠= 0-40 ml/min T=40-90oC 𝑃𝐶𝑂2= 10- 80 kPa Vequipment=50 ml, 500 ml 𝐹𝑔𝑎𝑠= 50 ml/min, 1 L/min T=40-60oC 𝑃𝐶𝑂2= 1.3 and 13 kPa


29


Page 32 of 61

- Glass bubbling - Vequipment=50 ml, 500 ml reactor - 𝐹𝑔𝑎𝑠= 50 ml/min, 1 L/min Shen et - Double-stirred 73 al. cell reactor - T=40-60oC - Ubbelohde - 𝑃𝐶𝑂2= 1.3 and 13 kPa viscometer asame apparatus used in Xu et al52with no further details reported - VLE data - Viscosity - Absorption rate - Mass transfer resistance

Xu et al. 54 perform a thorough investigation of the kinetics for the BDA/DEEA mixture as well as for each one of the individual components. The termolecular mechanism interprets the kinetics of the experimental data for BDA well, whereas the second-order reaction rate constants are determined for both BDA and DEEA. It is also observed that the reaction of CO2 with BDA/DEEA mixtures can be regarded as two parallel reactions of CO2 with BDA and DEEA, respectively. Xu et al.55 further measure the density and viscosity and N2O solubility in BDA, DEEA and BDA/DEEA at various concentrations. A major motivation is to derive data of physical solubility and diffusivity of CO2 in these solvents. However, due to the chemical interaction of CO2 with these solvents, N2O is used instead because it does not react with these amines, while it is very similar with CO2 in terms of molecular volume, configuration and electronic structure. The authors find that N2O solubility increases proportionally to the increase of BDA concentration whereas it decreases with increasing DEEA concentration. In BDA/DEEA it also decreases with increasing DEEA concentration in the mixture. Finally, Xu et al.56 perform a thorough investigation only of the lower, CO2-rich phase of the 2M BDA/ 4M DEEA mixture, motivated by the very high amount of CO2, as reported in Xu et al.54. The investigation employs almost all the experimental infrastructure used in the previous papers to measure several important indices and characteristics. The authors find that the cyclic capacity and loading are the same for the lower phase as in the mixture, while the absorption rates are higher for the lower phase for both fresh and half-loaded solutions, at CO2 pressure over 10kPa. Furthermore, the reaction products of the lower phase have more BDA bicarbamate, less BDA, less BDA carbamate and twice as much carbonate/bicarbonate than the mixture. Wang et al.57 further investigate the 4M DMBA + 2M DEEA and 2M DMBA + 4M DEEA mixtures in order to determine absorption capacity and mass transfer during reaction. Experiments are performed in a wetted wall column. They initially find that after CO2 absorption at 30oC the rich solution is observed in the lower layer, whereas the lean solution in the upper. Compared with MEA and K2CO3 the authors find that the absorption rates are in the following order: 5M MEA > 4M DMBA + 2M DEEA > 2M DMBA + 4M DEEA > 40%K2CO3. It is further observed that the overall reaction rate is controlled by the liquid film mass transfer resistance. Overall, the 4M DMBA + 2M DEEA solution exhibited higher absorption rate and reaction stability. Wang et al.58 investigate 1M DETA/ 4M DEEA with respect to mass transfer characteristics in a wetted wall column. Species analysis shows that the main species in the upper phase solvent is DEEA. DEEA and DETA are identified in the lower phase, indicating that DEEA gradually transfers to the lower phase during reaction with CO2. Compared to MEA and DEEA the


30

Page 33 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


absorption rates are in the following order: 5M MEA> 5M DETA> 1M DETA+4M DEEA> 5M DEEA. At temperature of 60oC CO2 is mainly at the lower 70% of the solution (i.e. this is the part that would be taken to the desorber in a potential process). Finally,. the total mass transfer coefficient is significantly affected by temperature and CO2 loading but insignificantly by gas flow changes. 6.2 Mixtures with polyamines Ye et al.59,60 perform a wide investigation of 50 mixtures to ascertain their absorption and desorption behavior as phase-change solvents and to select the one exhibiting the most favorable overall performance. The type of solvents investigated include molecules with at least one primary or secondary amine to act as the absorption activator and molecules with at least a tertiary amine to act as regeneration promoter. The absorption activator candidates include mainly linear or cyclic amines, whereas the regeneration activators are either DEEA or DMCA. Alkanolamines are not used as activators because they do not enhance the absorption behavior. However, they can be used as regeneration promoters because they have good solubility in water and their high boiling points enable thermal stability. The main characteristics of the absorption and desorption conditions are reported in Table 9. Polyamines such as DETA (diethylenetriamine), TETA, TEPA (tetraethylenepentamine) or PEHA (pentaethylenehexamine), used as activators in mixtures investigated in desorption experiments, are blended with DEEA or DMCA in ratios of 1:4 to avoid viscosity and precipitation issues. Among the tested activators, 14 are monoamines with carbon chains in the range C5-C12. An important observation for the monoamines is that the CO2 loading of the mixtures presents a maximum at activators with 7 carbon atoms and then declines. For activators containing more than 8 carbon atoms the absorption rates decrease and gel-like solids form which are insoluble up to 60 oC. Among the cyclic activators only mixtures with COtA (Cyclooctyl-amine) exhibit a high CO2 loading. With respect to promoters, higher CO2 capacities are observed for the DEEAbased mixtures than for the corresponding DMCA ones. Among the tested mixtures, the 5 mol/kg TETA/DEEA (1:4) exhibits approximately (under certain assumptions detailed in Ye et al.59): -

40% higher cyclic loading capacity than MEA, 15% lower heat of absorption than MEA, 50% lower sensible heat than MEA, stripping heat 30% lower than MEA, overall energy requirements 30% lower than MEA.

Furthermore, this mixture exhibits low viscosity, volatility, risk of precipitation and foam tendency. In general, activators with 3-4 amine atoms and 4-6 carbon atoms exhibit high absorption and desorption capacities without the previously mentioned operational issues. Building on the above findings, Ye et al.60 propose and investigate diethylenetriamine (DETA)/ N,N,N’,N’,N’’-pentamethyldiethylenetriamine (PMDETA)/H2O mixture. The first component contains two primary and one secondary amine, acting as an activator while the second acts as a


31


Page 34 of 61

regeneration promoter with three tertiary amines. The authors perform lab absorption/ desorption experiments to determine CO2 absorption capacity, phase transition behavior, and CO2 desorption pressure. NMR experiments are then performed to identify and quantify the species in both CO2-rich and CO2-lean phases and to determine mechanisms and reaction pathways. The findings indicate that the proposed mixture (tested in two different concentrations) exhibits liquid-liquid phase separation at absorption temperature. It further exhibits a cyclic capacity of up to 2.5 mol CO2/ kg of total solvent compared to 1.25 mol CO2/ kg of total solvent of MEA, and at higher CO2 partial pressure at stripping. With respect to reaction mechanisms it is observed that during absorption CO2 goes to the rich phase where DETA reacts quickly, followed by the protonation of PMDETA and the formation of 𝐻𝐶𝑂3 ―/𝐶𝑂23 ― species. In desorption these species accept protons produced by heating the protonated PMDETA. This enables the release of CO2 and the formation of a new lean oily phase caused by the limited solubility of molecular PMDETA in the rich aqueous phase. The work of Ye et al.60 is further extended by Zhou et al.61 who investigate various molar ratios of mixture PMDETA/DETA with respect to CO2 capture capacity and reaction mechanism. Findings indicate that mixture PMDETA/DEAT (4:1) exhibits phase separation at 50oC with a loading of 0.613 mol CO2/mol of total amine. Practically all loading is observed at the lower phase which occupies 57% of the total volume. The upper phase only contains PMDETA. DETA is the absorption accelerator but also contributes to the CO2 absorption capacity of the mixture. 6.3 Mixtures with non-amine compounds 6.3.1 DETA (Diethylenetriamine)/ Sulfolane Luo et al.62 investigate the use of a DETA/Sulfolane mixture as a biphasic solvent and perform experimental work to determine several parameters pertaining to absorption and desorption behavior, as reported in Table 9. Sulfolane is selected because it has been previously used as an aprotic polar solvent in the physical absorption of acid gas and exhibited high stability and low corrosivity. However, due to being aprotic, sulfolane cannot replace H2O entirely since there is no ionization in sulfolane itself. As a result, it is likely that the reaction of CO2 and DETA would not occur without H2O. Sulfolane changes the vapor-liquid equilibrium but does not affect the DETA-CO2 reaction mechanism. The phase separation appears at a loading of 0.35 moles CO2/mole amine. The upper phase contains all the DETA with most of the absorbed CO2. The liquid-liquid phase separation is due to the limited solubility of the ionic products of DETA with CO2 in sulfolane. In particular, 60% of the mixture containing 30% highly loaded DETA, 10% sulfolane and 60% H2O is routed to the stripper for desorption, while 40% of the mixture comprising 80% sulfolane and 20% H2O is recycled away from the desorber. The cyclic loading of this mixture is 35% higher than that of 30 wt% MEA. Wang et al.63 investigate DETA/Sulfolane with the aim to characterize kinetic behavior as well as CO2 absorption rate and theoretical regeneration energy requirements. The authors find that the 2M DETA/ 3M Sulfolane mixture exhibits higher absorption rate than the (1M, 4M) mixture as well than 5M MEA and 5M DETA solutions. Data regarding the phase-change mixture behavior are reported at 60oC, while regeneration energy is estimated at 120oC. Compared to 5M


32

Page 35 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


MEA, the mixture enables a decrease in sensible heat requirements of 59% and in vaporization heat of 19%. The overall regeneration heat is estimated at 3.23 GJ/ ton CO2 captured. The viscosity of this mixture is also found to be up to 20 times lower than the other investigated amines. 6.3.2 Amine/ ether mixtures Machida et al. 61 investigate mixtures EAE (2-ethylamino ethanol)/DEGDEE (Diethylene Glycol Diethyl Ether), EAE/ DEGDME (Diethylene Glycol Dimethyl Ether) and MAE (2-Methylamino ethanol)/ DEGDME. VLLE measurements are performed at absorption temperature in a lab solubility measurement apparatus. All three solvents exhibit phase separation at 40oC. The phase separation pressure is higher for DEGDME than for DEGDEE, with the former exhibiting lower hydrophobicity than the latter. Further investigation shows that as the hydrophobicity of the ether-based solvent increases and the hydrophobicity of the amine-based solvent decreases, the phase separation pressure decreases too. Machida et al.65 test mixtures of 60 wt.% DEGDEE with 30 wt.% BAE (2-Butylamino ethanol), or AMB (2-Amino-1-methoxybutane), or EAE or DAP (DL-1-Amino-2-propanol) or MAE or AEE (2-(2-Aminoethoxy) ethanol) or MEA in 10 wt.% water. The authors perform phase separation tests in the presence of CO2 and find that EAE exhibits phase-separation at PCO2=101 kPa and T= 40oC. Mixture EAE/DEGDEE is then investigated in temperatures 40-60oC at various PCO2 up to 101 kPa with VLL behavior exhibited at all temperatures and at higher pressures as the temperature increases. Finally, the mixture is compared with DMX-1 and MCA/DsBA (Zhang et al.29), indicating lower CO2 solubility at 40oC and at high pressures but higher solubility at lower pressures. Machida et al.66 investigate the EAE/DEGEE solvent using solubility measurements to perform theoretical regeneration energy assessment. The investigation takes place for a case of desorption operation at 80oC with an overall required energy of 2.93 GJ/ton CO2. This is further reduced to 1.6 GJ/ton CO2 through the use of a heat pump to recover sensible and reaction heat from 60oC to 50oC using a heat exchange solvent at 40oC. A key feature of this work is that the two phases are not separated but are sent to the desorber together. In cases where the phases are separated prior to the desorber, the latent heat is reduced but the mixing of the two phases after desorption produces entropy. In this new process the mixing entropy is exploited to promote desorption. The authors note that the whole process works as a heat engine which pumps CO2 by receiving high temperature heat in the desorber and losing low temperature heat at the absorber. 6.4 Non-aqueous mixtures Hu67,74,75,76,77 proposes a biphasic solvent consisting of 20-40% (by volume) activated agent and a solvent. The details are not disclosed but based on Kim et al.68 the biphasic solvent is a mixture of amine and alcohol. Phase separation occurs between 30oC and 45oC, while the biphasic solvent can be regenerated at 95oC. The author finds that regeneration energy requirements are 15% that of benchmark MEA solution, while the capital costs are 80% less than MEA. Kim et al.68 follow the approach of Hu74 and investigate the absorption behavior at 40oC of MEA and DEA in mixtures with 1-heptanol, 1-octanol and isooctanol in concentrations between 10-40 wt.%. The authors also perform speciation analysis. They find that the alcohols appear in the


33


Page 36 of 61

upper phase, and that amine and amine carbamate are present in the lower phase after absorption of CO2. The loadings of the amine-water solutions are higher than those of the amine-alcohol solutions. In non-aqueous MEA solutions the amine concentration does not affect the CO2 loading. In non-aqueous DEA solutions the CO2 loading is high at high amine concentrations. Barzagli et al.69 investigate MMEA)(2-(methylamino) ethanol)/DEGDEE and EMEA (2(ethylamino)ethanol)/ DEGDEE. Experiments include batch absorption screening of solvents, speciation experiments and cyclic absorption/ desorption experiments in a pilot plant. The authors observe that upon CO2 capture the proposed mixtures exhibit phase-separation and the speciation analysis indicates that the lighter phase contains the DEGDEE with a small amount of free amine while the heavier phase comprises ionic couple amine carbamate and protonated amine with a small amount of the diluent. The authors further find that EMEA-DEGDEE is best because of its high absorption efficiency up to 97.6%. Furthermore, for the same mixture the solubility results are used to estimate 40% lower sensible heat requirements compared to 30 wt.% MEA, due to the absence of water. Zhuang and Clements70 investigate MEA and MAE (2-methylaminoethanol) in isooctanol and 1heptanol in lab-scale experiments. MEA exhibits phase separation at absorption temperature in the alcohols but MAE does not exhibit phase separation at the entire temperature tested, i.e. up to 90oC (both amines are tested at 30 wt.%). The CO2-rich to CO2-lean volume ratio is 28/100. The lower phase is the CO2-rich one, which is to be sent to regenerator in a potential process. The upper phase is the CO2-lean one, which is to be sent to the absorber. 6.5 Cyclic and acyclic solvents In recent work, Tzirakis et al.71 perform VLLE experiments and report measurements for mixtures MCA/DMCA/AMP, MCA/DMCA, MCA/DMCA/Sulfolane and DEEA/MAPA. The experimental results are used by Perdomo et al.78 for the development of a predictive equation of state (EoS) using the SAFT-γ Mie79,80 group-contribution (GC) approach. They derive parameters for the prediction of the VLLE behavior of solvents in a GC sense, meaning that parameters are fitted for functional groups instead of entire molecules, hence making them transferrable to molecular structures which are different than the structures which were used to fit them81,82,83. The GC concept makes the approach very generic and applicable to a large variety of solvents for prediction of important solvent-CO2-water mixture properties such as loading and enthalpy of reaction at different temperatures and pressures. The GC idea is also exploited in Papadopoulos et al.84 who build on the computer-aided molecular design (CAMD) framework of Papadopoulos et al.1 to support the optimization-based design of novel phasechange solvents considering thermodynamic, sustainability and reactivity performance criteria. Finally, experimental VLLE data are used by Kazepidis et al.85 for the systematic design of phase-change CO2 capture flowsheets, illustrated for MCA, where structural modifications (e.g. stream splits, side feeds etc.) of conventional flowsheets exhibit important trade-offs between equipment sizes and process energetic performance. Zhang et al.72 investigate experimentally mixtures TETA/DEEA, TETA/DMCA, TETA/PMDETA, TETA/BDMAEE, and TETA/DEAPD. The investigations aim to determine the VLE behavior of the solvents in terms of absorption capacity, while regeneration energy requirements exploit experimental data for theoretical prediction of regeneration energy requirements. TETA/DMCA is selected for further investigation due to its high absorption


34

Page 37 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


capacity but also due to the fact that 98% of the absorbed CO2 appears in the lower phase. For this mixture a kinetic mechanism is also propose. Initially the reaction proceeds with the zwitterion mechanism between CO2 and TETA, whereas when TETA is consumed DMCA facilitates the regeneration of TETA for sustaining CO2 absorption. The regeneration energy requirements are 40% lower than 5M MEA reaching 2.07 GJ/ton CO2 captured at a lean loading of 0.45 mol/ mol and rich loading at 0.75 mol/mol. Shen at al.73 investigate experimentally several cyclic and acyclic solvents and mixtures aiming to identify structure-property relationships which can help guide the selection of phase-change solvents. In particular, ten primary/ secondary amines and eight tertiary amines with different alkalinity (pKa) and hydrophobicity (Log P) are considered in order to determine structural effects on absorption capacity, phase separation behavior, kinetics, and thermodynamics. The work reports experimental VLE data, viscosity, absorption rates, enhancement factors and total mass transfer resistance. Experimental data are also exploited for theoretical prediction of the regeneration energy requirements for solvents DEAPD, DEEA, TMPDA, DMCA and mixtures TETA/DEAPD, TETA/DEEA, TETA/TMPDA, TETA/DMCA. The lowest reaction enthalpy and regeneration energy are 45.8% and 52% lower than MEA for TETA/TMPDA for lean and rich loadings of 0.25 and 0.75 mol/mol, respectively. The work finds that the maximum CO2 absorption capacity is achieved with a tertiary amine of a pKa around 9.8. The total mass transfer resistance of the rich solution decreases with the increase in the Log P value of tertiary amines. The sensible heat decreases with increase of Log P while the latent heat increases with increase of Log P of tertiary amines. 7. Comparative analysis and concluding remarks Tables 10 and 11 provide an overview of the key characteristics of the reviewed solvent categories with respect to their use in phase-change processes. Apparently, only the lipophilic and the DMXTM solvents enable the placement of the liquid-liquid phase separator after the intermediate heat exchanger (HEX) in the absorption/desorption flowsheet, implying that it is possible to achieve a single liquid phase in the absorber. The rest of the solvents appear to exhibit an LLPS at absorption conditions and after a certain loading is reached hence the absorber possibly operates with 2 liquid phases. Furthermore, there are large differences in regeneration temperatures too. Solvents operating at lower than 100oC enable the use of waste heat from the emitting processes, hence their use enables the minimization of energetic requirements from auxiliary resources. Table 10: Main process-related features for the solvents selected/investigated in each reviewed category A/A

Solvent mixture

LLPS

1

DMCA+MCA+AMP (5.5 M, 3:1:1.5)17

60oC-90oC

2

DMX-1 (blend)38

>90oCb

3

DEEA/MAPA

40oC

Placement of Desorption phase separator conditions 60oC-80oC, After intermediate depending on HEX separation method After intermediate Up to 150oC and 5 HEX bar Before 90oC on average or


35


(5M, 2M)50 4 5 6 7 8

BDA/DEEA (2M, 4M)55

40oC

TETA/ DEEA (5 mol/kg, 1:4) 72 PMDETA/DETA (5M, 4:1)

40-50oC 40-50oC

60

DETA/ Sulfolane/ H2O (20 wt.%, wt. 40%, wt. 40%)61

40oC

DETA/ Sulfolane (2M, 3M) 63

Data for 60oC

9

EAE/DEGEE/H2O (30:60:10 wt%)66

40oC

10

Amine-alcohol67

30-45oC

11

12

13

14 15

MEA and DEA with 1Observed heptanol, 1-octanol and above 23oC isooctanol (10-40 wt%)68 - MMEA/DEGDEE (28.1 wt%, 71.9 wt%) - 50 oC - EMEA/DEGDEE (29.9 - 40 oC wt%, 70.1 wt%)69 - MEA/1-heptanol or - 40 oC isooctanol (30. Wt% MEA) - No phase - MAE/ 1-heptanol or separatio isooctanol (30. Wt% n MEA)70 - 2M DMBA+ 4M DEEA 30oC - 4M DMBA+ 2M DEEA57 DETA/ DEEA (1M/ 4M)58 40oC

Page 38 of 61

intermediate HEX 120oC and pressure at 6-8 bars Before intermediate 90oC a HEX Before 80oC-120oC intermediate HEX Before 120oC intermediate HEX Before intermediate 120oC a HEX NA VLL at all temperatures, no phase separator used Before intermediate HEXa NA

120oC 80oC

95oC NA

Before - 120 oC intermediate HEX - 110 oC Before intermediate HEXc

NA

NA

NA

NA NA After intermediate 16 MCA/DMCA/AMP (3M)71 40oC 90oC HEX Before 17 TETA/DMCA (4M, 1:3)72 40-60oC 100-120oC intermediate HEX Before 18 TETA/TMPDA (4M, 1:3)73 40-60oC 100-120oC intermediate HEX aThis is assumed here due to the LLPS, it has not been reported in the corresponding papers bReported by Zhang et al.29 cPlacement of phase-separator refers only to the first of the two mixtures Table 11 illustrates that three out of all the phase-change solvents evaluated have reached a technology readiness level of 4-5, whereas the remaining are still at a level of property measurements in bench-scale equipment and have not yet been tested in larger pilot plants. The


36

Page 39 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


capacity indicators are used as a means of comparing the different studies and apparatuses or plants. Solvents 2 and 3 have been shown in pilot plant studies to enable regeneration energy requirements in the order of 2.2-2.5GJ/ton CO2, whereas solvent 1 is also estimated to be able to reach a similar energetic performance. It is worth noting that capital and operating cost estimates for the DMXTM process indicate that economic gains are tightly related to the type of emitting plant and specifications of flue gases. Process energetic investigations have been performed for solvents 1 and 3. For the remaining solvents, the operating gains are based on bench-scale experiments. ]Table 11: Key process performance indicators, operational issues, Technology Readiness Level (TRL) and process capacity indicators for each solvent category. Numbering corresponds to solvents of Table 11. A/A

Key performance indicators Operational issues

TRL

3 1

Volatility losses, corrosion by MCA mainly

Regeneration energy: 2 GJ/ton CO2a

2

Reboiler duty: 2.3-2.5 GJ/ton CO2b

Foaming

3

Reboiler duty: 2.2-2.4 GJ/ton CO2

Volatile losses, corrosion similar to MEA

4

5

6

- 46% higher cyclic loading than 5M MEA, - 48% higher cyclic capacity than 5M MEA, - 11% higher cyclic efficiency than 5M MEA. - 40% higher cyclic loading capacity than MEA, - 15% lower heat of absorption than MEA, - 50% lower sensible heat than MEA, - stripping heat 30% lower than MEA, - overall energy requirements 30% lower than MEA. - 0.613 mol CO2/mol of total amine (for

3-4

4-5 5

Capacity indicators -

Absorption column 𝐹𝑔𝑎𝑠=10-75 L/min 𝐹𝑙𝑖𝑞𝑢𝑖𝑑=60-200 mL/min - Absorption/ stirred tank regenerator pilot plant - 𝐹𝑔𝑎𝑠=100-400 L/hr - 𝐹𝑙𝑖𝑞𝑢𝑖𝑑=0.8kg/hr - Complete pilot plant - 𝐹𝑔𝑎𝑠=171.3 L/hr - Complete pilot plant - 𝐹𝑔𝑎𝑠=90 m3/hr - 𝐹𝑙𝑖𝑞𝑢𝑖𝑑=180L/hr - Lab equipment - 𝐹𝑔𝑎𝑠= up to 874 ml/min

-

2-3

-

2

- Lab equipment - 𝐹𝑔𝑎𝑠= 105 ml/min

-

2

- Lab equipment - 𝐹𝑔𝑎𝑠= 60 ml/min


37


7

8

9

10

11

12

13

14 15 16 17

PMDETA/DETA-5M, 4:1) - Cyclic capacity 100% that of MEA (For 2 mol kg−1DETA/ 3 mol kg−1PMDETA in water) - 35% higher cyclic loading than 30wt% MEA - Regeneration heat 3.23 GJ/ton CO2 - Viscosity 20 times lower than the other investigated amines - Absorption rate 2-fold higher than 5m MEA - 2.93 GJ/ton CO2 - 1.6 GJ/ton CO2 with heat pump - Energy requirements15% that of MEA - Capital costs 80% less than MEA. - CO2 loading capacity 4-6 times higher than 20% (v/v) MEA Loadings of amine-water solutions are higher than those of amine-alcohols solutions tested 40 % lower sensible heat requirements compared to 30 wt.% MEA. CO2-rich to CO2-lean volume ratio is 28/100 (i.e. only a small fraction would go to regenerator) Absorption rates lower than MEA but higher than 40%K2CO3 Absorption rates lower than 5M MEA but higher than DEEA 40% better loading than MEA 40% lower regeneration energy than MEA (2.07 GJ/t CO2)

Page 40 of 61

-

2

-

Lab equipment Vreactor=400 ml

-

2

-

Lab equipment Fgas=500 ml/min

-

2-3

-

Lab equipment

Very limited corrosion and volatile losses

2-3

-

Lab equipment Vliquid=900 ml

-

2

-

Lab equipment Fgas=200 ml/min

-

Absorption/ desorption pilot plant Fgas= 29 dm3/hr Fliquid= 0.1-0.3 dm3/hr

Assumed reduced equipment corrosion due to absence of water

3

-

Lab equipment Vequipment=660 or 1245 mL

-

2

-

2

-

Lab equipment Vequipment= 150 ml 𝐹𝑔𝑎𝑠= 500 ml/min

-

2

-

Lab equipment 𝐹𝑔𝑎𝑠= 500 ml/min

-

2

-

Lab equipment 𝐹𝑔𝑎𝑠= 0-40 ml/min

-

2

-

Lab equipment 𝐹𝑔𝑎𝑠= 1 L/min


38

Page 41 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


45.8% and 52% lower reaction enthalpy (45 kJ/mol) 18 and heat duty (1.83 GJ/t CO2) than MEA aEstimated for TBS1+ agitation flowsheet b2.5 GJ/ton CO based on pilot plant data 2

-

-

2

Lab equipment 𝐹𝑔𝑎𝑠= 1 L/min

Overall it appears that almost all studies focus on lab- or pilot-scale testing of the solvents. Furthermore, there are very limited studies reporting VLLE data at various temperatures and pressures which can be used for modeling purposes. In few occasions model-based investigation of phase-change behavior or process flowsheets has been reported for the solvents summarized in Table 12. In most modeling attempts, models are developed based on direct regression and fitting of experimental data on an empirical equation. A notable exception is the work of Arshad et al.48 who attempt to derive interaction parameters for the UNIQUAC model in order to be able to predict activity coefficients for a broader range of conditions for the DEEA/MAPA mixture. These findings indicate that there is significant scope in the development of general activity coefficient or equation of state (EoS) models which enable predictions of the phase-change behavior of a wider set of solvents. Such work is presented by Perdomo et al.78 who show the derivation of parameters for the prediction of the VLL behavior of solvents using the SAFT-γ Mie79,80 GC approach. Finally, there is a clear need to identify advanced process flowsheets which exploit the capabilities of phase-change solvents and further improve their capital and operating cost requirements through process modifications. To this end, Zhang17 illustrated the advantages of such flowsheets through empirical modifications. Damartzis et al.31,32 showed that optimization-based synthesis of process flowsheets for CO2 capture can systematically identify structural and operational features of high performance. Kazepidis et al.85 expand this work to the systematic design of phase-change CO2 capture flowsheets, illustrated for MCA, where structural modifications of conventional flowsheets exhibit important trade-offs between equipment sizes and process energetic performance. To this end, phase-change CO2 capture technologies can take advantage of advance process systems engineering approaches which account for a holistic assessment of capture costs, starting from the material level and reaching all the way to the capture, utilization and sequestration, supply chain network level. A notable such approach is proposed for non-phase-change processes by Hasan et al. 86,87,88 who employ optimization to select source plants, capture processes, capture materials, CO2 pipelines, locations of utilization and sequestration sites, and amounts of CO2 storage. Table 12: Summary of solvent cases where investigations of wider content are observed

Solvent type

Lipophilic

Experimental property measurement studies - Agar et al.23,24 - Tan25 - Zhang et al. 8,26,,28,29,30,33

Experimental Thermodynamic/ -pilot plant kinetic modeling studies studies

Process modeling/ technoeconomic assessment

- Zhang et al.30 - Zhang17

- Tan25 - Zhang et al.29 - Zhang17

-

- Zhang17


39


DMXTM

DEEA/ MAPA BDA/ DEEA

al.12

- Aleixo et - Raynal et al.35

- Raynal et al.38 - Dreillard et al.37

- Hartono et al.42 - Arshad et al.21, 43,44,45,46

- Ciftja et al.19 - Pinto et al.20 - Xu et al. 51,52,54,55,53,56

MMEA/DEG - Barzagli et al.69 DEE Cyclic and acyclic -Tzirakis et al.71 solvents

- Pinto et al.49

-

-

- Pinto et al.49 - Arshad et al.48 - Xu et al. 55

Page 42 of 61

- Raynal et al.34,35,38 - Gomez et al.39 - Dreillard et al.37 - Liebenthal et al.50 -

- Barzagli et al.69

-

-

-

- Perdomo et al.78 - Papadopoulos et al.84

- Kazepidis et al.85

Other than improving the costs of conventional chemisorption processes, phase-change solvent and process technologies may also compete with other CO2 capture processes. Hasan et al. 86 provide an economic comparison regarding the cost of capture of MEA-based chemisorption, adsorption- and membrane-based processes. They note that MEA-based chemisorption is suitable for lower flue gas CO2 compositions of 5-10%, which are typical of power plants. For flue gas of 15-20% in CO2 or higher, adsorption-based technologies enable capture cost reductions compared to MEA, which start from 20% and may increase up to 45% as the CO2 concentration in the flue gas increases. Gomez et al.39 show in Figure 8 that for a cement plant with 15% CO2 in the flue gas, the DMXTM capture cost reduction may be 50% compared to MEA. With respect to membrane technologies, chemisorption-based processes using conventional solvents are competitive in terms of energetic requirements for flue gases with CO2 mole fraction lower than 0.1789 (to reach 90% CO2 purity and recovery). This assumes that some conventional solvents may reach regeneration energy demands in the order of 3.2 GJ/CO2 capture. With phase-change solvents reaching regeneration energy demands close to 2.0 GJ/ton CO2 capture, they may compete with membrane processes for industrial flue gases with CO2 mole fraction of about 0.2789. These are clear indications that phase-change solvents may become competitive with other technologies for industries other than power plants. It should be further noted that this comparison does not account for the flue gas processing capacity of each technology, where chemisorption processes are suitable for plants with very high flue gas flowrates. Phase-change processes exhibit significant economic incentives for scaling up and wider implementation, but they are not without challenges. As in all cases of chemisorption solvents, phase-change solvents may be volatile and corrosive. The economic and environmental impacts pertaining to solvent losses and equipment maintenance should also be taken into account during their assessment. Furthermore, their oxidation products should also be investigated with respect to emissions that are generated from capture plants. These are issues which have been highlighted in few cases (e.g. see Zhang17) and potential solutions have been proposed, but more research is needed in this direction. To this end, Papadopoulos et al. 1,90 proposed a framework for holistic sustainability assessment of CO2 capture solvents, including phase-change solvents, ACS Paragon Plus Environment

40

Page 43 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


while Papadokonstantakis et al. 91 extended this framework to account for holistic sustainability of process flowsheet modifications in view of different solvents. Note here that other materials used in CO2 capture such as membranes also exhibit environmental impacts when viewed from a life-cycle perspective92. Further important aspects in phase-change solvents include the lower critical solution and phase-separation temperatures of their aqueous mixtures. These parameters determine the placement of the phase separator and operating characteristics of the process flowsheet, such as selection of appropriate packing materials. Precipitation and salt formation are also important issues which must be identified and avoided early on, during solvent screening. Abbreviations Symbol AMP BDA BF C CAPEX COtA CST CSTR D DAA DAB or DBA DEEA DETA DIPA DMBA DMCA DPA DsBA EEDA EPD F H HA HEX LCST MAPA MCA MDEA MEA MPD OPEX P PEHA PWS PZ T t TBS

Explanation 2-amino-2-methyl-1-propanol 1,4-Butanediamine Blast furnace Concentration Capital expenditutres Cyclooctyl-amine Critical solution temperature Continuous stirred tank reactor Diameter Diallylamine 1,4-Diaminobutane 2-(Diethylamino)-ethanol Diethylenetriamine Diisopropylamine N,N-Dimethylbutylamine N,N-Dimethyl cyclohexylamine N, N-Dipropylamine Di-sec-butylamine N-Ethylethylenediamine N-Ethyl piperidine Flow Height Hexylamine Heat exchanger Lower critical solution temperature 3-(Methylamino)-propylamine N-Methylcyclo hexylamine Methyldiethanolamine Monoethanolamine N-Methyl piperidine Operational expenditures Pressure Pentaethylenehexamine Power station Piperazine Temperature Time Thermomorphic biphasic solvent


41


TEA TEPA TETA TGR UCST V

Page 44 of 61

Triethylamine Tetraethylenepentamine Thriethylenetetramine Top gas recycle Upper critical solution temperature Volume

Acknowledgements This work has received funding from the European Union’s Horizon 2020 research and innovation program under the grant agreement 727503 - ROLINCAP – H2020-LCE-20162017/H2020-LCE-2016-RES-CCS-RIA. This research has also been co‐financed by the European Union and Greek national funds through the Operational Program Competitiveness, Entrepreneurship and Innovation, under the call RESEARCH – CREATE – INNOVATE, project T1EDK-02472.

(1)

(2) (3) (4) (5) (6) (7) (8) (9) (10)

8. References Papadopoulos, A. I.; Badr, S.; Chremos, A.; Forte, E.; Zarogiannis, T.; Seferlis, P.; Papadokonstantakis, S.; Adjiman, C. S.; Galindo, A.; Jackson, G. Computer-Aided Molecular Design and Selection of CO2 Capture Solvents Considering Thermodynamics. Mol. Syst. Des. Eng. 2016, 1, 313–334. Papadopoulos, A. I.; Seferlis, P. Process Systems and Materials for CO2 Capture: Modelling, Design, Control and Integration; Papadopoulos, A. I., Seferlis, P., Eds.; John Wiley & Sons: Chichester, 2017. Aaron, D.; Tsouris, C. Separation of CO2 from Flue Gas: A Review. In Separation Science and Technology; 2005; Vol. 40, pp 321–348. https://doi.org/10.1081/SS200042244. Mangalapally, H. P.; Hasse, H. Pilot Plant Experiments for Post Combustion Carbon Dioxide Capture by Reactive Absorption with Novel Solvents. In Energy Procedia; 2011; Vol. 4, pp 1–8. https://doi.org/10.1016/j.egypro.2011.01.015. Wang, X.; Li, B. Phase-Change Solvents for CO2 Capture. In Novel Materials for Carbon Dioxide Mitigation Technology; Shi, F., Morreale, B., Eds.; Elsevier: Amsterdam, 2015; pp 3–22. Mumford, K. A.; Smith, K. H.; Stevens, G. W. Phase Change Solvents for CO2 Capture. In Energy Efficient Solvents for CO2 Capture by Gas-Liquid Absorption; Budzianovsky, W. M., Ed.; Springer, 2017; pp 99–117. Zhuang, Q.; Clements, B.; Dai, J.; Carrigan, L. Ten Years of Research on Phase Separation Absorbents for Carbon Capture: Achievements and next Steps. International Journal of Greenhouse Gas Control. 2016. https://doi.org/10.1016/j.ijggc.2016.04.022. Zhang, J.; Qiao, Y.; Agar, D. W. Improvement of Lipophilic-Amine-Based Thermomorphic Biphasic Solvent for Energy-Efficient Carbon Capture. In Energy Procedia; 2012; Vol. 23, pp 92–101. https://doi.org/10.1016/j.egypro.2012.06.072. Hasib-Ur-Rahman, M.; Larachi, F. CO2 Capture in Alkanolamine-RTIL Blends via Carbamate Crystallization: Route to Efficient Regeneration. Environ. Sci. Technol. 2012, 46 (20), 11443–11450. https://doi.org/10.1021/es302513j. ZHANG, P.; SHI, Y.; WEI, J.; ZHAO, W.; YE, Q. Regeneration of 2-Amino-2-Methyl-1Propanol Used for Carbon Dioxide Absorption. J. Environ. Sci. 2008, 20 (1), 39–44. https://doi.org/10.1016/S1001-0742(08)60005-4.


42

Page 45 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(11) (12) (13) (14) (15) (16)

(17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27)

(28)


Saha, A. K.; Bandyopadhyay, S. S.; Biswas, A. K. Kinetics of Absorption of CO2into Aqueous Solutions of 2-Amino-2-Methyl-1-Propanol. Chem. Eng. Sci. 1995, 50 (22), 3587–3598. https://doi.org/10.1016/0009-2509(95)00187-A. Aleixo, M.; Prigent, M.; Gibert, A.; Porcheron, F.; Mokbel, I.; Jose, J.; Jacquin, M. Physical and Chemical Properties of DMXTM Solvents. In Energy Procedia; 2011; Vol. 4, pp 148–155. https://doi.org/10.1016/j.egypro.2011.01.035. Gabrielsen, J.; Svendsen, H. F.; Michelsen, M. L.; Stenby, E. H.; Kontogeorgis, G. M. Experimental Validation of a Rate-Based Model for CO2 capture Using an AMP Solution. Chem. Eng. Sci. 2007, 62 (9), 2397–2413. https://doi.org/10.1016/j.ces.2007.01.034. Chakraborty, A. K.; Astarita, G.; Bischoff, K. B. CO2 absorption in Aqueous Solutions of Hindered Amines. Chem. Eng. Sci. 1986, 41 (4), 997–1003. https://doi.org/10.1016/00092509(86)87185-8. Astarita, G.; Savage, D. W.; Longo, J. M. Promotion of CO2 Mass Transfer in Carbonate Solutions. Chem. Eng. Sci. 1981, 36 (3), 581–588. https://doi.org/10.1016/00092509(81)80146-7. Versteeg, G. F.; Kuipers, J. A. M.; Van Beckum, F. P. H.; Van Swaaij, W. P. M. Mass Transfer with Complex Reversible Chemical Reactions-II. Parallel Reversible Chemical Reactions. Chem. Eng. Sci. 1990, 45 (1), 183–197. https://doi.org/10.1016/00092509(90)87091-6. Zhang, J. Study on CO2 Capture Using Thermomorphic Biphasic Solvents with EnergyEfficient Regeneration, TU-Dortmund, 2013. Góral, M.; Shaw, D. G.; Maczyński, A.; Wiśniewska-Gocłowska, B.; Oracz, P. IUPACNIST Solubility Data Series. 96. Amines with Water Part 2. C7-C24 Aliphatic Amines. J. Phys. Chem. Ref. Data 2012, 41 (4). https://doi.org/10.1063/1.4755953. Ciftja, A. F.; Hartono, A.; Svendsen, H. F. Experimental Study on Phase Change Solvents in CO2 Capture by NMR Spectroscopy. Chem. Eng. Sci. 2013, 102, 378–386. https://doi.org/10.1016/j.ces.2013.08.036. Pinto, D. D. D.; Zaidy, S. A. H.; Hartono, A.; Svendsen, H. F. Evaluation of a Phase Change Solvent for CO2 Capture: Absorption and Desorption Tests. Int. J. Greenh. Gas Control 2014, 28, 318–327. https://doi.org/10.1016/j.ijggc.2014.07.002. Arshad, M. W.; Von Solms, N.; Thomsen, K.; Svendsen, H. F. Heat of Absorption of CO2in Aqueous Solutions of DEEA, MAPA and Their Mixture. In Energy Procedia; 2013; Vol. 37, pp 1532–1542. https://doi.org/10.1016/j.egypro.2013.06.029. Zhang, X. Studies on Multiphase CO2 Capture Systems, TU-Dortmund, 2007. Agar, D. W.; Y., T.; X., Z. CO2 Removal Processes by Means of Absorption Using Thermomorphic Biphasic Aqueous Amine Solutions. 2008/015217, 2008. Agar, D. W.; Tan, Y. H. Extractive Regeneration of Loaded Amine Solutions at Very Low Temperature. 10 2008 007 087.4., 2008. Tan, Y. H. Study of CO2 Absorption into Thermomorphic Biphasic Quasi-Immobilised Amines, TU-Dortmund, 2010. Zhang, J.; Qiao, Y.; Agar, D. W. Intensification of Low Temperature Thermomorphic Biphasic Amine Solvent Regeneration for CO 2 Capture. Chem. Eng. Res. Des. 2012, 90 (6), 743–749. https://doi.org/10.1016/j.cherd.2012.03.016. Zhang, J.; Qiao, Y.; Wang, W.; Misch, R.; Hussain, K.; Agar, D. W. Development of an Energy-Efficient CO2capture Process Using Thermomorphic Biphasic Solvents. In Energy Procedia; 2013; Vol. 37, pp 1254–1261. https://doi.org/10.1016/j.egypro.2013.05.224. Zhang, J.; Nwani, O.; Tan, Y.; Agar, D. W. Carbon Dioxide Absorption into Biphasic Amine Solvent with Solvent Loss Reduction. Chem. Eng. Res. Des. 2011, 89 (8), 1190– 1196. https://doi.org/10.1016/j.cherd.2011.02.005.


43


(29) (30) (31) (32)

(33) (34)

(35) (36) (37) (38)

(39)

(40) (41) (42) (43)

(44)

Page 46 of 61

Zhang, J.; Misch, R.; Tan, Y.; Agar, D. W. Novel Thermomorphic Biphasic Amine Solvents for CO 2 Absorption and Low-Temperature Extractive Regeneration. Chem. Eng. Technol. 2011, 34 (9), 1481–1489. https://doi.org/10.1002/ceat.201100099. Zhang, J.; Agar, D. W.; Zhang, X.; Geuzebroek, F. CO2 Absorption in Biphasic Solvents with Enhanced Low Temperature Solvent Regeneration. In Energy Procedia; 2011; Vol. 4, pp 67–74. https://doi.org/10.1016/j.egypro.2011.01.024. Damartzis, T.; Papadopoulos, A. I.; Seferlis, P. Process Flowsheet Design Optimization for Various Amine-Based Solvents in Post-Combustion CO2 Capture Plants. J. Clean. Prod. 2016, 111. https://doi.org/10.1016/j.jclepro.2015.04.129. Damartzis, T.; Papadopoulos, A. I.; Seferlis, P. Optimal Design of Solvent-Based PostCombustion CO2 Capture Plants Using Phase-Change Solvents. In Computing and Systems Technology Division 2015 - Core Programming Area at the 2015 AIChE Annual Meeting; 2015; Vol. 2. Zhang, J. F.; Chen, J.; Misch, R.; Agar, D. W. Carbon Dioxide Absorption in Biphasic Amine Solvents with Enhanced Low Temperature Solvent Regeneration. Chem. Eng. Trans. 2010, 21, 129–174. Raynal, L.; Alix, P.; Bouillon, P. A.; Gomez, A.; Le Febvre De Nailly, M.; Jacquin, M.; Kittel, J.; Di Lella, A.; Mougin, P.; Trapy, J. The DMXTM Process: An Original Solution for Lowering the Cost of Post-Combustion Carbon Capture. In Energy Procedia; 2011; Vol. 4, pp 779–786. https://doi.org/10.1016/j.egypro.2011.01.119. Raynal, L.; Bouillon, P. A.; Gomez, A.; Broutin, P. From MEA to Demixing Solvents and Future Steps, a Roadmap for Lowering the Cost of Post-Combustion Carbon Capture. Chem. Eng. J. 2011, 171 (3), 742–752. https://doi.org/10.1016/j.cej.2011.01.008. Porcheron, F.; Gibert, A.; Mougin, P.; Wender, A. High Throughput Screening of CO 2 Solubility in Aqueous Monoamine Solutions. Environ. Sci. Technol. 2011, 45 (6), 2486– 2492. https://doi.org/10.1021/es103453f. Dreillard, M.; Broutin, P.; Briot, P.; Huard, T.; Lettat, A. Application of the DMXTM Capture Process in Steel Industry. Energy Procedia 2016, 114, 2573–2589. Raynal, L.; Briot, P.; Dreillard, M.; Broutin, P.; Mangiaracina, A.; Drioli, B. S.; Politi, M.; La Marca, C.; Mertens, J.; Thielens, M. L.; et al. Evaluation of the DMX Process for Industrial Pilot Demonstration - Methodology and Results. In Energy Procedia; 2014; Vol. 63, pp 6298–6309. https://doi.org/10.1016/j.egypro.2014.11.662. Gomez, A.; Briot, P.; Raynal, L.; Broutin, P.; Gimenez, M.; Soazic, M.; Cessat, P.; Saysset, S. ACACIA Project – Development of a Post-Combustion CO2 Capture Process. Case of the DMXTM Process. Oil Gas Sci. Techol. – Rev. IFP Energies Nouv. 2014, 69, 1121–1129. Lemaire, E.; Bouillon, P.; Lettat, K. Development of HiCapt+TM Process for CO2 Capture from Lab to Industrial Pilot Plant. Oil Gas Sci. Technol. –Rev. IFP Energles Nouv. 2014, 69, 1069–1080. Kim, I.; Svendsen, H. F. Comparative Study of the Heats of Absorption of PostCombustion CO2 Absorbents. Int. J. Greenh. Gas Control 2011, 5 (3), 390–395. https://doi.org/10.1016/j.ijggc.2010.05.003. Hartono, A.; Saeed, M.; Ciftja, A. F.; Svendsen, H. F. Binary and Ternary VLE of the 2Amino-2-Methyl-1-Propanol (AMP)/Piperazine (Pz)/Water System. Chem. Eng. Sci. 2013, 91, 151–161. https://doi.org/10.1016/j.ces.2013.01.015. Arshad, M. W.; Svendsen, H. F.; Fosbøl, P. L.; Von Solms, N.; Thomsen, K. Equilibrium Total Pressure and CO2solubility in Binary and Ternary Aqueous Solutions of 2(Diethylamino)Ethanol (DEEA) and 3-(Methylamino)Propylamine (MAPA). J. Chem. Eng. Data 2014, 59 (3), 764–774. https://doi.org/10.1021/je400886w. Arshad, M. W.; Fosbøl, P. L.; Von Solms, N.; Svendsen, H. F.; Thomsen, K. Equilibrium


44

Page 47 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(45)

(46) (47) (48) (49) (50)

(51) (52) (53) (54) (55)

(56) (57) (58)

(59)


Solubility of CO2in Alkanolamines. In Energy Procedia; 2013; Vol. 51, pp 217–223. https://doi.org/10.1016/j.egypro.2014.07.025. Arshad, M. W.; Fosbøl, P. L.; Von Solms, N.; Svendsen, H. F.; Thomsen, K. Heat of Absorption of CO2in Phase Change Solvents: 2-(Diethylamino)Ethanol and 3(Methylamino)Propylamine. J. Chem. Eng. Data 2013, 58 (7), 1974–1988. https://doi.org/10.1021/je400289v. Arshad, M. W.; Fosbøl, P. L.; Von Solms, N.; Thomsen, K. Freezing Point Depressions of Phase Change CO2solvents. J. Chem. Eng. Data 2013, 58 (7), 1918–1926. https://doi.org/10.1021/je3013167. Abrams, D. S.; Prausnitz, J. M. Statistical Thermodynamics of Liquid Mixtures: A New Expression for the Excess Gibbs Energy of Partly or Completely Miscible Systems. AIChE J. 1975. https://doi.org/10.1002/aic.690210115. Arshad, M. W.; von Solms, N.; Thomsen, K. Thermodynamic Modeling of Liquid–liquid Phase Change Solvents for CO2capture. Int. J. Greenh. Gas Control 2016, 53, 401–424. https://doi.org/10.1016/j.ijggc.2016.08.014. Pinto, D. D. D.; Knuutila, H.; Fytianos, G.; Haugen, G.; Mejdell, T.; Svendsen, H. F. CO2post Combustion Capture with a Phase Change Solvent. Pilot Plant Campaign. Int. J. Greenh. Gas Control 2014, 31, 153–164. https://doi.org/10.1016/j.ijggc.2014.10.007. Liebenthal, U.; Di D. Pinto, D.; Monteiro, J. G. M. S.; Svendsen, H. F.; Kather, A. Overall Process Analysis and Optimisation for CO2Capture from Coal Fired Power Plants Based on Phase Change Solvents Forming Two Liquid Phases. In Energy Procedia; 2013; Vol. 37, pp 1844–1854. https://doi.org/10.1016/j.egypro.2013.06.064. Xu, Z.; Wang, S.; Liu, J.; Chen, C. Solvents with Low Critical Solution Temperature for CO2capture. In Energy Procedia; 2012; Vol. 23, pp 64–71. https://doi.org/10.1016/j.egypro.2012.06.045. Xu, Z.; Wang, S.; Zhao, B.; Chen, C. Study on Potential Biphasic Solvents: Absorption Capacity, CO2 Loading and Reaction Rate. Energy Procedia 2013, 37, 494–498. https://doi.org/DOI: 10.1016/j.buildenv.2010.02.002. Xu, Z. C.; Wang, S. J.; Chen, C. H. CO2 Absorption by Biphasic Solvents: Mixtures of 1,4-Butanediamine and 2-(Diethylamino)-Ethanol. Int. J. Greenh. Gas Control 2013, 16, 107–115. Xu, Z.; Wang, S.; Chen, C. Kinetics Study on CO2 Absorption with Aqueous Solutions of 1,4-Butanediamine, 2-(Diethylamino)-Ethanol, and Their Mixtures. Ind. Eng. Chem. Res. 2013, 52 (29), 9790–9802. Xu, Z.; Wang, S.; Chen, C. Solubility of N2O in Density and Viscosity of Aqueous Solutions of 1,4-Butanediamine, 2-(Diethylamino)-Ethanol, and Their Mixtures from (298.15 to 333.15) K. J. Chem. Eng. Data 2013, 58 (6), 1633–1640. https://doi.org/10.1021/je301371p. Xu, Z.; Wang, S.; Qi, G.; Liu, J.; Zhao, B.; Chen, C. CO 2 Absorption by Biphasic Solvents: Comparison with Lower Phase Alone. Oil Gas Sci. Technol. – Rev. d’IFP Energies Nouv. 2014, 69 (5), 851–864. https://doi.org/10.2516/ogst/2013155. Wang, L.; An, S.; Li, Q.; Yu, S.; Wu, S. Phase Change Behavior and Kinetics of CO2absorption into DMBA/DEEA Solution in a Wetted-Wall Column. Chem. Eng. J. 2017, 314, 681–687. https://doi.org/10.1016/j.cej.2016.12.033. Wang, L.; An, S.; Yu, S.; Zhang, S.; Zhang, Y.; Li, M.; Li, Q. Mass Transfer Characteristics of CO2absorption into a Phase-Change Solvent in a Wetted-Wall Column. Int. J. Greenh. Gas Control 2017, 64, 276–283. https://doi.org/10.1016/j.ijggc.2017.08.001. Ye, Q.; Wang, X.; Lu, Y. Screening and Evaluation of Novel Biphasic Solvents for Energy-Efficient Post-Combustion CO2 Capture. Int. J. Greenh. Gas Control


45


(60) (61) (62) (63) (64) (65) (66)

(67) (68) (69) (70) (71) (72) (73) (74) (75) (76) (77) (78)

(79)

Page 48 of 61

2015, 39, 205–214. https://doi.org/10.1016/j.ijggc.2015.05.025. Ye, Q.; Zhu, L.; Wang, X.; Lu, Y. On the Mechanisms of CO2 Absorption and Desorption with Phase Transitional Solvents. Int. J. Greenh. Gas Control 2017, 56, 278–288. https://doi.org/10.1016/j.ijggc.2016.11.027. Zhou, X.; Liu, F.; Lv, B.; Zhou, Z.; Jing, G. Evaluation of the Novel Biphasic Solvents for CO2 Capture: Performance and Mechanism. Int. J. Greenh. Gas Control 2017, 60, 120– 128. https://doi.org/10.1016/j.ijggc.2017.03.013. Luo, W.; Guo, D.; Zheng, J.; Gao, S.; Chen, J. CO2 Absorption Using Biphasic Solvent: Blends of Diethylenetriamine, Sulfolane, and Water. Int. J. Greenh. Gas Control 2016, 53, 141–148. https://doi.org/10.1016/j.ijggc.2016.07.036. Wang, L.; Yu, S.; Li, Q.; Zhang, Y.; An, S.; Zhang, S. Performance of Sulfolane/DETA Hybrids for CO2 Absorption: Phase Splitting Behavior, Kinetics and Thermodynamics. Appl. Energy 2018, 228, 568–576. Machida, H.; Esaki, T.; Oba, K.; Tomikawa, T.; Yamaguchi, T.; Horizoe, H. Phase Separation Solvent for CO2 Capture. Energy Procedia 2017, 114, 823–826. Machida, H.; Oba, K.; Tomikawa, T.; Esaki, T.; Yamaguchi, T.; Horizoe, H. Development of Phase Separation Solvent for CO2 Capture by Aqueous (Amine+ Ether) Solution. J. Chem. Thermodyn. 2017, 113, 64–70. Machida, H.; Ando, R.; Esaki, T.; Yamaguchi, T.; Horizoe, H.; Kishimoto, A.; Akiyama, K.; Nishimura, M. Low Temperature Swing Process for CO2 Absorption-Desorption Using Phase Separation CO2 Capture Solvent. Int. J. Greenh. Gas Control 2018, 75, 1–7. https://doi.org/10.1016/j.ijggc.2018.05.010. Hu, L. CO2 Capture from Flue Gas by Phase Transitional Absorption; 2009. Kim, Y. E.; Park, J. H.; Yun, S. H.; Nam, S. C.; Jeong, S. K.; Yoon, Y. Il. Carbon Dioxide Absorption Using a Phase Transitional Alkanolamine-Alcohol Mixture. J. Ind. Eng. Chem. 2014, 20 (4), 1486–1492. https://doi.org/10.1016/j.jiec.2013.07.036. Barzagli, F.; Mani, F.; Peruzzini, M. Novel Water-Free Biphasic Absorbents for Efficient CO2capture. Int. J. Greenh. Gas Control 2017, 60, 100–109. https://doi.org/10.1016/j.ijggc.2017.03.010. Zhuang, Q.; Clements, B. CO2 Capture by Biphasic Absorbent–absorption Performance and VLE Characteristics. Energy 2018, 147, 169–176. https://doi.org/10.1016/j.energy.2018.01.004. Tzirakis, F.; Tsivintzelis, I.; Papadopoulos, A. I.; Seferlis, P. Experimental Measurement and Assessment of Equilibrium Behaviour for Phase Change Solvents Used in CO2 Capture. Chem. Eng. Sci. 2019, 199, 20–27. https://doi.org/10.1016/J.CES.2018.12.045. Zhang, S.; Shen, Y.; Shao, P.; Chen, J.; Wang, L. Kinetics, Thermodynamics, and Mechanism of a Novel Biphasic Solvent for CO2Capture from Flue Gas. Environ. Sci. Technol. 2018, 52 (6), 3660–3668. https://doi.org/10.1021/acs.est.7b05936. Shen, Y.; Jiang, C.; Zhang, S.; Chen, J.; Wang, L.; Chen, J. Biphasic Solvent for CO2 Capture: Amine Property-Performance and Heat Duty Relationship. Appl. Energy 2018, 230, 726–733. https://doi.org/10.1016/j.apenergy.2018.09.005. Hu, L. Phase Transitional Absorption Method. US 2007/0237695 A1, 2007. Hu, L. Phase Transitional Absorption Method. US 7541011 B2, 2009. Hu, L. Methods and Systems for Deacidizing Gaseous Mixtures. US 7,718,151 B1, 2010. Hu, L. Self-Concentrating Absorbent for Acid Gas Separation. US 7,846,407 B2, 2010. Perdomo, F. A.; Papadopoulos, A. I.; Galindo, A.; Adjiman, C. S.; Jackson, G.; Seferlis, P. Design of Phase-Change Solvents for Post-Combustion CO2 Capture Based on Molecular Thermodynamics, Frontiers of Molecular Engineering. In Frontiers of Molecular Engineering; Chicago, 2018. Dufal, S.; Papaioannou, V.; Sadeqzadeh, M.; Pogiatzis, T.; Chremos, A.; Adjiman, C. S.;


46

Page 49 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(80)

(81)

(82) (83) (84)

(85)

(86) (87)

(88)

(89) (90)

(91)


Jackson, G.; Galindo, A. Prediction of Thermodynamic Properties and Phase Behavior of Fluids and Mixtures with the SAFT-γ Mie Group-Contribution Equation of State. J. Chem. Eng. Data 2014, 59 (10), 3272–3288. https://doi.org/10.1021/je500248h. Hutacharoen, P.; Dufal, S.; Papaioannou, V.; Shanker, R. M.; Adjiman, C. S.; Jackson, G.; Galindo, A. Predicting the Solvation of Organic Compounds in Aqueous Environments: From Alkanes and Alcohols to Pharmaceuticals. Ind. Eng. Chem. Res. 2017, 56 (38), 10856–10876. https://doi.org/10.1021/acs.iecr.7b00899. Papadopoulos, A. I.; Tsivintzelis, I.; Seferlis, P.; Linke, P. Computer-Aided Molecular Design: Fundamentals, Methods, and Applications. Reedijk, J. (Ed.) Elsevier Reference Module in Chemistry, Molecular Sciences and Chemical Engineering.; Waltham, MA: Elsevier., 2018. https://doi.org/10.1016/B978-0-12-409547-2.14342-2. Lymperiadis, A.; Adjiman, C. S.; Galindo, A.; Jackson, G. A Group Contribution Method for Associating Chain Molecules Based on the Statistical Associating Fluid Theory (SAFT- γ). J. Chem. Phys. 2007, 127 (23). https://doi.org/10.1063/1.2813894. Lymperiadis, A.; Adjiman, C. S.; Jackson, G.; Galindo, A. A Generalisation of the SAFTγ Group Contribution Method for Groups Comprising Multiple Spherical Segments. Fluid Phase Equilib. 2008, 274 (1–2), 85–104. https://doi.org/10.1016/j.fluid.2008.08.005. Papadopoulos, A. I.; Shavalieva, G.; Perdomo, F. A.; Seferlis, P.; Papadokonstantakis, S.; Adjiman, C. S.; Galindo, A.; Jackson, G. Systematic Design of Phase-Change Solvents for Post-Combustion CO2 Capture Based on Advanced Thermodynamics and Holistic Sustainability Assessment. In AICHE Annual Meeting; AICHE: Pittsburgh, 2018; p 58h. Kazepidis, P.; Papadopoulos, A. I.; Seferlis, P.; Adjiman, C. S.; Perdomo, F. A.; Galindo, A.; Jackson, G. Optimal Design of Post Combustion CO2 Capture Processes Based on Phase-Change Solvent. In Proceedings of the 29th European Symposium on Computer Aided Process Engineering; Kiss, A. A., Zondervan, E., Lakerveld, R., Özkan, L., Eds.; Elsevier: Eindhoven, 2019. Hasan, M. M. F.; First, E. L.; Boukouvala, F.; Floudas, C. A. A Multi-Scale Framework for CO2capture, Utilization, and Sequestration: CCUS and CCU. Comput. Chem. Eng. 2015, 81, 2–21. https://doi.org/10.1016/j.compchemeng.2015.04.034. Hasan, M. M. F.; Boukouvala, F.; First, E. L.; Floudas, C. A. Nationwide, Regional, and Statewide CO2Capture, Utilization, and Sequestration Supply Chain Network Optimization. Ind. Eng. Chem. Res. 2014, 53 (18), 7489–7506. https://doi.org/10.1021/ie402931c. Hasan, M. M. F. Multi-Scale Process Systems Engineering for Carbon Capture, Utillization, and Storage: A Review. In Process Systems and Materials for CO2 Capture: Modelling, Design, Control and Integration; Papadopoulos, A. I., Seferlis, P., Eds.; John Wiley and Sons, 2017; pp 231–247. Belaissaoui, B.; Willson, D.; Favre, E. Membrane Gas Separations and Post-Combustion Carbon Dioxide Capture: Parametric Sensitivity and Process Integration Strategies. Chem. Eng. J. 2012, 211–212, 122–132. https://doi.org/10.1016/j.cej.2012.09.012. Papadopoulos, A. I.; Shavalieva, G.; Papadokonstantakis, S.; Seferlisa, P. A Framework for the Integration of Holistic Sustainability Assessment in Computer-Aided Molecular Design. In Proceedings of the 29th European Symposium on Computer Aided Process Engineering; Kiss, A. A., Zondervan, E., Lakerveld, R., Özkan, L., Eds.; Elsevier: Eindhoven, 2019. Papadokonstantakis, S.; Badr, S.; Hugerbuhler, K.; Papadopoulos, A. I.; Damartzis, T.; Seferlis, P.; Forte, E.; Chremos, E.; Galindo, A.; Adjiman, C. S.; et al. Towards Sustainable Solvent-Based Post Combustion CO2 Capture: From Molecules to Conceptual Flowsheet Design. In Sustainability of Products, Processes and Supply Chains: Theory and Applications, Computer-aided Molecular Engineering 36; You, F., Ed.; Elsevier,


47


(92)

Page 50 of 61

2015; pp 279–305. Troy, S.; Schreiber, A.; Zapp, P. Life Cycle Assessment of Membrane-Based Carbon Capture and Storage. Clean Technol. Environ. Policy 2016, 18 (6), 1641–1654. https://doi.org/10.1007/s10098-016-1208-x.


48

Page 51 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1: Illustration of the reaction route in biphasic systems17 154x100mm (300 x 300 DPI)



Figure 2: The LSCT behavior of the ethylbutylamine – water mixture (Experimental data reported by Goral et al.18) 144x113mm (300 x 300 DPI)


Page 52 of 61

Page 53 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3: 1-methylpiperidine + water liquid – liquid equilibrium (Experimental data reported by Goral et al.18) 140x117mm (300 x 300 DPI)



Figure 4: Typical, conceptual flowsheet of a phase-change process with LLPS higher than absorption conditions 329x150mm (300 x 300 DPI)


Page 54 of 61

Page 55 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5: Performance comparison of selected solvents with respect to criteria 1-4 using HA (Hexylamine) as a reference solvent. DPA: N, N-Dipropylamine, DsBA: Di-sec-butylamine, MCA: N-Methylcyclo hexylamine, DMCA: N,N-Dimethyl cyclohexylamine, MPD: N-Methyl piperidine, EPD: N-Ethyl piperidine. Blue bars pointing right indicate better performance than HA, red bars pointing left indicate worse performance than HA. Cyclic capacity is defined as the difference between CO2 rich and CO2 lean loading at 25oC and 90°C, respectively. This is without considering stripping gas for EPD and MDP, while stripping gas of 200ml/min N2 is considered at 75 oC for all the other amines. Regenerability is defined as the ratio of the difference between CO2 rich and CO2 lean loading over CO2 rich loading and is reported at 80 oC. Cyclic capacity and absorption rate are measured at 2.7 M. The regenerability of EPD is reported for a 3M concentration hence 2.7M and 4.5M concentrations don’t apply and the corresponding regenerability bars only provide an approximation. (Experimental data reported by Zhang17) 246x68mm (300 x 300 DPI)



Figure 6: Comparison of selected lipophilic solvents and their blends and MEA/MDEA with MEA based on important properties. Blue bars pointing right indicate better performance than MEA, whereas red bars pointing left indicate worse performance than MEA. Solvents are blended at a ratio 3:1. (Experimental data reported by Zhang17). 242x96mm (300 x 300 DPI)


Page 56 of 61

Page 57 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 7: Energy requirements for flowsheets proposed by Zhang17 compared to the standard MEA process. (Data from Zhang17). 216x144mm (300 x 300 DPI)



Figure 8: Performance comparison of the DMXTM and conventional MEA processes for different CO2 emission plants (Data from Gomez et al.39) 176x102mm (300 x 300 DPI)


Page 58 of 61

Page 59 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 9: Performance comparison of the HICAPTTM, HICAPT+TM and DMXTM processes for different CO2 emission characteristics (Data from Dreillard et al.37) 194x114mm (300 x 300 DPI)



Figure 10:Typical, conceptual flowsheet of a process exhibiting phase-separation at absorption conditions 268x150mm (300 x 300 DPI)


Page 60 of 61

Page 61 of 61 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


208x196mm (300 x 300 DPI)


Phase-change Solvents and Processes for Post-Combustion CO2

Recommend Documents