Subscriber access provided by University of Sunderland
Separations
Comparative Kinetic studies of solid absorber catalyst (K/MgO) and solid desorber catalyst (HZSM-5)-aided CO2 absorption and desorption from aqueous solutions of MEA and blended solutions of BEA-AMP and MEA-MDEA Daniel B Afari, James Coker, Jessica Narku-Tetteh, and Raphael O. Idem Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/acs.iecr.8b02931 • Publication Date (Web): 26 Oct 2018 Downloaded from http://pubs.acs.org on November 5, 2018
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 44 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
1 2 3
Comparative Kinetic studies of solid absorber catalyst (K/MgO) and solid desorber catalyst (HZSM-5)-aided CO2 absorption and desorption from aqueous solutions of MEA and blended solutions of BEA-AMP and MEA-MDEA
4
Daniel B. Afari1, James Coker1, Jessica Narku-Tetteh1, Raphael Idem1*
5 6 7
1Clean
Energy Technologies Research Institute, Faculty of Engineering and Applied Science, University of Regina, Regina, Saskatchewan S4S 0A2, Canada
8
Abstract
9
The kinetic performance of a novel amine solvent blend BEA-AMP was compared with MEA and
10
blended MEA-MDEA in the presence and absence of a solid acid catalyst (HZSM-5) in the
11
desorber column of a bench-scale pilot plant. In addition, a total of seven solid base catalysts were
12
screened using a semi-batch reactor to select the one that is most suitable as catalyst for amine-
13
based CO2 absorption. The selected solid base catalyst, K/MgO, was placed in the absorber of the
14
pilot plant. Overall, three absorber-desorber catalytic scenarios were evaluated: blank-blank,
15
blank-HZSM-5 and K/MgO-HZSM-5. For the blank-blank and blank-HZSM-5 scenarios, the
16
novel solvent (4M BEA-AMP) outperformed 5M MEA and 7M MEA-MDEA blend despite BEA-
17
AMP having the lowest molarity. The rates of absorption and desorption for the blank-blank (non-
18
catalytic) scenario for BEA-AMP were 14.8 and 38.4 mol/m3min, respectively. For the blank-
19
HZSM-5 system, the rates were 18.1 and 45.6 mol/m3min, respectively. Absorption and desorption
20
rates of 29.7 and 62.4 mol/m3min, respectively were obtained for the K/MgO-HZSM-5 system.
21
These results reveal higher rates of absorption and desorption with the inclusion of solid base and
22
solid acid catalysts to the amine-based CO2 capture process. The results show that in the presence
23
of the amine, the electron-rich anion species in K/MgO easily attack the dissolved CO2. This
24
interaction ties the CO2 molecules to the surface of the catalyst, making them readily available for
25
nitrogen atom of the amine in the CO2 absorption process. This process is facilitated because of
26
the large pore size of K/MgO. With the desorber catalyst, easier proton donation by HZSM-5
27
results in weakening the N-C bond in carbamate thereby causing CO2 to break away.
28 29 30
Corresponding Author: R. Idem; Email:
[email protected]; Tel: 1-306-585-4470; Fax: 1-
31
306-585-4855 1 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
32
Nomenclature
33
AMP- 2-Amino-2-methyl-1-propanol
34
[𝐴] – Concentration of species, mol/dm3
35
BEA – Butyl ethanolamine
36
C – Concentration of species, mol/dm3
37
CO2 in – CO2 composition in the inlet gas
38
CO2 out – CO2 composition in outlet gas
39
𝐹i – Molar flow rate of species i mol min−1
40
MEA – monoethanolamine
41
MDEA – Methyldiethanolamine
42
MW – molecular weight
43
ri – rate of reaction based on a particular species, mol gcat−1 min−1
44
V – Volume of reactor, m3
45
V – Volumetric flowrate, slpm
46
W – Weight of catalyst, g
47
(𝑊/𝐹io) – Contact time, min
48
Xi – conversion of component i
49 50
Greek Letters
51
∆ - gradient
52
α - CO2 loading
53 54 2 ACS Paragon Plus Environment
Page 2 of 44
Page 3 of 44 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
55
Subscripts
56
Am – amine
57
Lean – solution lean in CO2
58
rich – solution rich in CO2
59
In – entering the reactor
60
Out – exiting the reactor
3 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
62
1. Introduction
63
Despite its huge contribution as a major source of energy, the use of fossil fuels could be
64
limited by fact that large quantities of greenhouse gases (GHGs), with CO2 being a major
65
contributor, are emitted. Among several existing technologies for mitigating global warming by
66
CO2 capture, post-combustion capture (PCC) emerges as the most advanced technology in use.1
67
The use of amine-based solvents in PCC has been a great area of research and has seen progress
68
to a very large extent. Nonetheless, its merits tend to be stifled by the huge energy penalty
69
associated with the capture process. Thus, current studies are focused more on reducing this energy
70
penalty than in other areas of challenge. The two main areas that are being understudied are process
71
optimization and solvent enhancement. Energy requirement is classified under the former.
72
There exist many amine-based solvents (primary, secondary and tertiary) used for CO2
73
capture. However, the most widely used is the primary amine, monoethanolamine (MEA) due to
74
its high kinetics and moderate absorption capacity. Its limitation is associated with the huge energy
75
penalty which has necessitated the development of many other solvents with better absorption and
76
desorption characteristics. Works done by Chakraborty et al.2, Liao et al.3, Mandal et al.4, Sun et
77
al.5, Choi et al.6, and Sutar et al.7 show a better performance of these amine-based solvents (single
78
and blended) over the conventional MEA. The selection of these amine solvents has been based
79
more on a trial and error approach and not on any rational scientific approach.
80
Recently, a more rational criterion for selection of amine-based solvent was developed by
81
Narku-Tetteh et al.8 and Muchan et al.9. A group of primary, secondary and tertiary amines were
82
studied to check for the effect of their side chain structures and number of hydroxyl groups on CO2
83
absorption and desorption.8 A combination of solvent properties and performance estimation
84
parameters such as CO2 absorption and desorption kinetics, equilibrium loading, heat duty, cyclic
85
capacity, pKa and heat of absorption were grouped into Absorption and Desorption parameters. A
86
novel bi-solvent aqueous amine blend constituting 2-butyl-aminoethanol (BEA) and 2-amino-2-
87
methyl-1-propanol (AMP) of equimolar concentration (2M each) outperformed other potential
88
solvents, including MEA and MDEA (known for its excellent desorption characteristics) based on
89
newly developed “Absorption and Desorption parameters”. A comparative kinetic study between
90
the novel solvent and conventional MEA as well as MEA-MDEA blend is studied in this work.
4 ACS Paragon Plus Environment
Page 4 of 44
Page 5 of 44 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
91
Catalysts account for faster kinetics in many reactions. The use of catalysts to enhance the
92
CO2 capture system is in its incipient stages and has not yet attained commercial implementation.
93
A number of catalysts have been employed in enhancing both CO2 absorption and desorption. The
94
use of inorganic catalysts to enhance CO2 absorption into aqueous solutions have been reported by
95
Sharma et al.10, Bandyopadhyay et al.11, Ghosh et al.12, Guo et al.13, Nicholas et al.14, and Phan et
96
al.15. Sivanesan et al.16 used tertiary amine nitrate salts in the presence of an aqueous tertiary amine
97
medium to enhance the CO2 absorption rate using the stopped-flow technique. The application of
98
solid acid catalysts in the desorber unit was introduced recently by Idem et al.17 This was in view
99
of reducing the conventional regeneration temperatures of about 120-140oC to temperatures below
100
100oC (between 85oC and 95oC).18 Shi et al.18 studied single MEA and blends of MEA-MDEA
101
and MEA-DEAB over two solid acid catalysts (HZSM-5 and γ-Al2O3) on a batch scale and
102
achieved a substantial reduction in heat duty with the introduction of these catalysts. He explained
103
his findings using the duplicative role of γ-Al2O3 as a HCO3- ion and the donation of protons by
104
HZSM-5 aiding in the breakdown of carbamate. This study was taken a step further to the bench-
105
scale pilot plant level by Akachuku19, who performed kinetic studies on MEA and MEA-MDEA
106
over these two solid acid catalysts and thus proved the results obtained by Shi et al.18 Other
107
researchers 20-22 have also reported on significant heat duty reduction by applying different solid
108
acid catalysts (SAPO-34, SO42-/TiO2, SO42-/ZrO2, MCM-41) to the CO2 desorption process.
109
Catalytic studies involving the use of solid mineral catalysts solely in the CO2 absorption process
110
is almost nonexistent in the literature. Recently, Shi et al.23 studied the addition of solid base
111
catalysts to DEA solvent to enhance the absorption process using CaCO3 and MgCO3 on a batch
112
scale. A reduction in overall reaction time by up to 14-28% and 11-28% were obtained for CaCO3
113
and MgCO3, respectively.
114
The present study evaluates the kinetic performance of a novel solvent blend, butyl (amino)-
115
ethanolamine
and
2-amino-2-methyl-1-propanol
(BEA/AMP)
116
Monoethanolamine (MEA) and blended monoethanolamine and n-Methyldiethanolamine
117
(MEA/MDEA) solvents and compares their kinetic performance with the addition of a solid acid
118
catalyst (HZSM-5) in the desorber section of a bench-scale pilot plant. Also, the performance of a
119
total of 7 solid base catalysts (Hydrotalcite, BaCO3, CaCO3, Cs2O/α-Al2O3, Cs2O/Al2O3, Ca(OH)2
120
and K/MgO) are evaluated and screened on a semi-batch scale with an equimolar mixture of the
121
novel solvent 4M BEA-AMP blend. The criterion for selecting the catalysts for screening was 5 ACS Paragon Plus Environment
with
conventional
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
Page 6 of 44
122
based on the H – scale, a measure of identifying solid base strengths as proposed by Tanabe et
123
al.24. With respect to this definition, they classified numerous solid base catalysts as either strong
124
or weak. Solid bases with H– values greater than +26 were classified as superbases. The selected
125
solid base catalysts for screening all had H– values greater than +26. These superbases have been
126
proven to be very reactive for most organic reactions including alkene isomerization and toluene
127
side-chain alkylation 25. Out of these, seven catalysts were selected and screened for amine-based
128
CO2 capture. They are BaCO3, CaCO3, Ca(OH)2, Cs2O/α-Al2O3 (hydrothermally treated),
129
Cs2O/Al2O3, K/MgO and Hydrotalcite. After screening, the selected solid base catalyst was
130
incorporated into the absorber section of the bench-scale pilot plant and its performance was
131
compared with two other configurations of (i) a blank case having no catalysts in both columns
132
(non-catalytic) and (ii) a semi-blank case with no solid base catalyst in the absorber but with solid
133
acid catalyst, HZSM-5 present in the desorber. The results are presented and discussed in this
134
paper.
135 136
2. Theory 2.1 Amine based CO2 reaction (Zwitterion mechanism).
137
Typical primary and secondary amines react rapidly with CO2 in a 2:1 mole ratio to form
138
carbamates and protonated amines via the zwitterion mechanism.26 This mechanism involves a
139
two-stage process where there is the formation of a ‘zwitterion’ intermediate (step 1) and
140
subsequent breakdown of this intermediate to form carbamate (step 2). BEA, which is a secondary
141
amine, reacts via this process to yield carbamate while AMP, a sterically hindered primary amine,
142
hydrolyzes to bicarbonate after the formation of its carbamate.27 These reactions are summarized
143
as follows:
144 145
BEA/AMP + CO2 → BEA + COO-/AMP+COO-
(1)
146
BEA+COO-/ AMP+COO- + BEA/AMP → BEACOO-/AMPCOO- + BEAH+/AMPH+
(2)
147
The carbamate reversion reaction for AMP to yield bicarbonate proceeds as:
148
AMPCOO- + H2O → AMP + HCO3-
149
2.2 Solid base catalysts
(3)
6 ACS Paragon Plus Environment
Page 7 of 44 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
150
The reaction between aqueous amine solvents and CO2 is exothermic leading to the release
151
of large amount of heat. These reactions can be viewed as acid-base reactions since CO2 is an acid
152
gas and amines are generally basic. With CO2 being an acid gas, there is the need for a basic
153
medium to enhance its rate of absorption into an aqueous solvent. One classification of a basic
154
medium is the introduction of solid base sorbents (solid base catalysts). The use of these base
155
sorbents (catalysts) can facilitate the rate of the transfer process by both physically increasing the
156
interfacial area for mass transfer and chemically by providing another reaction pathway with a
157
lower activation energy allowing for faster chemical absorption of CO2.28
158 159
3. Experimental Section
160
3.1 Chemicals and Materials
161
The
following
chemicals:
BaCO3(≥99%),
CH3COOCs(≥95%),
CaCO3
(≥99%),
162
KOH(≥99.99%),
163
Na2CO3(≥99%),
164
Butylethanolamine (≥98%) and Ludox HS 40 colloidal silica (40wt% suspension in H2O) were all
165
purchased from Sigma Aldrich Co. Canada. 2-amino-2-methyl-1-propanol (≥99%) was purchased
166
from Acros Organics. Ca(OH)2(≥99%) and NaOH (≥99%) were purchased from Fisher Scientific
167
Co. Canada. γ-Al2O3 and α-Al2O3 were purchased from Zeochem Inc., US while HZSM-5 was
168
purchased from Zibo Yinghe Chemical Company Limited, China. Pure CO2 and N2 gas tanks were
169
supplied by Praxair Inc., Regina, Canada. 15% CO2 (N2 balance) gas tanks for the Gas analyzer
170
calibrations was also purchased from Praxair Inc., Regina, Canada. Standard 1N Hydrochloric acid
171
used in titration experiments was purchased from Sigma Aldrich Co., Canada.
Monoethanolamine
Al(NO3)3.9H2O(≥98%),
Mg(OH)2(≥95%),
(≥98%),
Methyl
Mg(NO3)2.6H2O(≥98%),
diethanolamine
(≥98%),
2-
172 173
3.2 Catalyst Preparation
174
The preparation procedure from Cavani et al.29, Climent et al.30, Diez et al.31, and
175
Gorzawski et al.32 were followed for the preparation of each catalyst with slight modifications.
176
The preparation of Hydrotalcite was done following the procedure outlined by Cavani et al.29 using
177
the co-precipitation method.
7 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
178
Cs2O/α-Al2O3 and Cs2O/Al2O3 were prepared by the procedure outlined in Gorzawski et
179
al.32 with little modifications. α-Al2O3 and Al2O3 beads were crushed and impregnated with
180
prepared solutions of caesium acetate. For α-Al2O3, prior to impregnation with caesium acetate, it
181
was hydrothermally treated in a Parr reactor at 500oC for 2 hours and dried for 12 hours. Upon
182
impregnation with Caesium acetate, both α-Al2O3 and Al2O3 products were stirred for
183
approximately 2 hours and finally calcined at 900oC for 2 hours to decompose the caesium acetate.
184
The steps outlined by Diez et al.31 were employed in the preparation of K/MgO. All
185
catalysts were pelletized to the desired size by passing through appropriate sieves after being
186
pressed using a 4 cm internal diameter die set under a hydraulic press.
187 188
3.3 Catalyst Characterization
189
The Brunauer-Emmett-Teller (BET) Surface Area, Pore Volume, and Average Pore Size
190
measurements and X-ray Diffraction (XRD) characterization experiments were performed at the
191
Chemical and Biological Engineering Department Laboratory at the University of Saskatchewan,
192
Saskatoon. The Temperature Programmed Desorption (TPD) and Scanning Electron Microscope
193
(SEM) experiments were conducted at the Clean Energy Technologies and Research Institute
194
(CETRI) laboratory at the University of Regina. ICP analysis was conducted at the Environmental
195
Analytical Laboratories of the Saskatchewan Research Council in Saskatoon.
196
For the BET analysis, the instrument used was BET ASAP 2020 from Micromeritics,
197
(Georgia, USA). The sample was degassed at 150oC for 5 hours. Nitrogen gas was used during
198
analysis. BJH method was employed to calculate the surface area, pore volume and pore size
199
obtained from the adsorption and desorption isotherms. 46 relative pressure points were recorded
200
to give the Isotherm plots.
201
Powder X-Ray Diffraction (XRD) experiments were performed on a Rigaku Ultima IV X-
202
Ray diffractometer, equipped with a Cu source (1.54056 Å), a CBO optical, and a scintillation
203
counter detector available at the Saskatchewan Structural Science Centre of the University of
204
Saskatchewan, Saskatoon. The measurements were carried out on the multipurpose attachment in
205
the parafocusing mode, with a Kβ filter (Ni foils) inserted into the receiving slit box. The intensity
206
data were obtained over a 2θ scan range from 5° to 80°, with a scan rate of 5° per minute and a 8 ACS Paragon Plus Environment
Page 8 of 44
Page 9 of 44 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
207
step size of 0.02. The generator voltage and current were set to 40 kV and 40 mA, respectively.
208
Identification of the crystalline phases were done using the reference data from International
209
Center for Diffraction Data (ICDD) and literature.
210
TPD measurements were done using a CHEMBET-3000 analyser with a TCD Detector
211
from Quantachrome Instruments. The catalyst sample was initially degassed by being exposed to
212
100% helium gas accompanied by heating the furnace gradually to a temperature of 250oC at
213
10oC/min ramping. The system was kept at this temperature for 60 minutes after which the
214
temperature was reduced to 30oC. A 3% CO2 gas (balance nitrogen gas) was introduced for 60
215
minutes at a flowrate of 30 ml/min for adsorption to take place. The temperature was then increased
216
to 700oC in a constant flow of helium gas.
217
The surface morphology of the prepared catalyst samples was also investigated by scanning
218
electron microscopy (SEM) using a JEOL 5600 132-10 electron probe micro analyser with an
219
active area of 10 mm2available at CETRI, University of Regina. The sample was first crushed to
220
obtain polished flat surfaces and was then loaded into the specimen chamber. Beams were
221
generated based on the accelerating voltage of 25 kV. The positioning of the beam was controlled
222
by the computer software and micrographs were finally acquired.
223
ICP analysis was conducted to obtain the actual metal content of catalysts. A mass of 0.2g
224
of the catalyst sample was digested in 5ml concentrated HNO3. It was then mixed and left standing
225
for 2 hours. Digestion was done by using a Mars 6 (CEM Corporation) extraction system. The
226
sample was then microwave digested for half an hour by heating at 1200 W and pressure
227
increments from 80 to 150 psi. Dilution of the sample digests was done to 50 ml volume and
228
digests and blanks were then analysed by Agilent 8800 ICP-MS.
229 230
Catalyst Screening
231
Absorption experiments for the catalyst screening were carried out at a temperature of 45±1oC
232
at atmospheric pressure of 1 atm. For a fair basis of comparison, a catalyst weight of 50g for each
233
catalyst was used. This is about 10% of the mixture. It is known that a higher catalyst percentage
234
may tend to overwhelm the reaction itself (i.e. reflects inactivity of catalyst) and causes it to reach
235
maximum conversion. This would mean that the addition of more catalyst would not change the
236
conversion. However, as will be later shown in the results section, considerable and varying 9 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
237
performance (increase in conversion) among the various catalysts was evident. A 4M BEA-AMP
238
solvent concentration, solvent volume of 500 ml and a constant stirring speed of 600 rpm were
239
maintained throughout all runs. The concentration was confirmed by titration with a 1N HCl
240
solution. The apparatus consisted of a 1000 ml three-necked round bottom flask immersed in a
241
preheated oil bath. The catalyst (3.5-4 mm particle size) was carefully placed in a stainless-steel
242
basket and fully immersed into the solution by being suspended with the aid of stainless steel wires
243
to ensure no contact with the magnetic stirrer or bottom of the flask as shown in Figure 1. A non-
244
catalytic (blank) run was also performed and used as the baseline for comparing the performance
245
of the various catalysts. Table 1 shows the experimental conditions used in catalyst screening
246
experiment.
247
Each experiment started with a known volume of solvent (500 ml) introduced into the flask
248
and the filled flask was immersed into an oil bath which was heated to the desired absorption
249
temperature. Via the dispersion tube, the solvent was then bubbled through with a pre-mixed gas
250
of 15%CO2 (balance N2) at a flowrate of 650±5 ml/min. After reaching the desired temperature,
251
samples were taken at regular intervals of 5 minutes during the first hour and subsequently at
252
intervals of 30 minutes. This was done to measure the CO2 loadings at those time intervals with
253
the aid of the Chittick apparatus as described by Ji et al.33 Sampling continued until the solution
254
attained equilibrium at 10 hours (600 minutes) of run time. Solutions were filtered prior to
255
measuring their CO2 loadings in order to eliminate catalyst particle interference with loading
256
measurements. CO2 loading (mol/mol) versus time (minutes) curves were generated and slopes of
257
the linear portion of these curves gave the initial rate of absorption. Following the selection of the
258
catalyst, γ-Al2O3 and colloidal silica (40 wt. %) were employed as binders for the selected catalyst
259
in order for it to be used in the absorption unit of the bench-scale pilot plant. Their effect on the
260
overall performance of the selected catalyst was also tested under the same experimental
261
conditions of the absorption experiments. This was to ensure that there was no added contribution
262
or adverse effect whatsoever from the binders.
10 ACS Paragon Plus Environment
Page 10 of 44
Page 11 of 44 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
263
Figure 1 Experimental set-up of semi-batch run for catalyst screening
264 265 266
Table 1. Operating conditions of semi-batch catalyst screening experiments Parameter
Value
Gas flowrate
650 ml/min
Liquid volume
500 ml
Absorption temperature
45oC
CO2 in feed gas
15%
Catalyst weight
50g
267 268
3.4 Pilot plant
269
The system consisted of two lagged stainless-steel columns each with dimensions of 1.067
270
m in height and an internal diameter of 0.0508m. The absorption column was designed with 4
271
ports, being the gas inlet, off-gas outlet, lean-solvent inlet and rich-solvent outlet. However, the
272
desorber column had 3 ports namely; rich-solvent inlet, lean-solvent outlet and CO2 product gas
273
outlet. Both columns were equipped with 5.08 cm LDX sulzer structured packing arrangement
274
with the solid base catalyst beds interspersed between them for that of the absorption column. The
275
absorber column had the desired solid base catalyst weight evenly distributed between any two 11 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
276
structured packing. Also, the desorber column packing arrangement enclosed a solid acid catalyst
277
bed (HZSM-5) mixed with 3 mm inert marbles. Between the structured packing arrangement and
278
the catalyst bed-3mm inert marbles section, were 6 mm inert marbles which acted as support for
279
the catalyst bed-3mm inert marbles section. The bench-scale pilot plant and absorber and desorber
280
bed arrangements are shown in Figures 2 and 3 respectively. Table 2 shows the operating
281
conditions employed in the bench-scale pilot plant experiment.
282
The key feature of the experimental set-up is the replacement of the reboiler in conventional
283
systems with a rich amine heater. A typical experimental run began with the lean amine solvent,
284
with the desired concentration and flowrate, fed from the storage tank to the top of the absorber
285
column via a variable-speed gear pump. At the same time the heating bath was set to the desired
286
temperature to heat up the rich amine prior to entering the desorber. Once amine solvent circulation
287
was set, a mixture of CO2 and N2 gas at the appropriate CO2 concentration of 15% was introduced
288
to the bottom of the absorber column via a gas flow meter where it was met by a counter-current
289
flow of lean amine solvent from the top of the column. Here, a three-phase system was set-up
290
comprising the amine solvent, CO2-N2 gas and solid base catalyst. Treated gas exited the top of
291
the absorber column while the rich amine solvent exited at the absorber bottom and exchanged
292
heat with the hot lean amine stream coming from the bottom of the desorber. The rich amine stream
293
was further heated, through a heat-exchanger network, by the heating medium. The heated rich
294
amine stream was fed to the top of the desorber column. Upon contacting the catalytic desorber
295
bed, further desorption was enhanced by the catalyst bed and the lean amine leaves the bottom of
296
the desorber, is cooled, and fed into the absorber column making a complete cycle. A condenser
297
is employed to cool the CO2 product gas leaving the top of the desorber column to remove any
298
entrained water and amine, and the product gas is dried prior to being measured by the rotameter.
299
The absorber column temperature profile was consistently monitored to check for
300
attainment of equilibrium. At equilibrium, the CO2 concentrations in the gas phase along the
301
absorber and temperature profiles in both columns were measured using an Infra-Red (IR) gas
302
analyzer from Nova Analytical Systems Inc., Canada and J-type thermocouples from Cole Parmer
303
Inc., Canada respectively. Also, lean and rich amine samples were taken from the bottom of both
304
columns to determine the rich and lean CO2 loadings (liquid phase CO2 concentrations). The CO2
305
loadings were determined using a Chittick apparatus as described by Ji et al.33 Loading 12 ACS Paragon Plus Environment
Page 12 of 44
Page 13 of 44 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
306
measurements were done thrice to ensure repeatability and accuracy. The average deviation was
307
less than 1%. Gas phase CO2 concentrations at the inlet and outlet of the absorber were taken twice
308
also to ensure accuracy in measurements. Prior to this, the IR gas analyzer was calibrated with a
309
15% CO2 premixed gas. A mass balance error calculation was done to determine the validity of
310
each run. The calculation compared the quantity of CO2 removed from the gas phase to the CO2
311
quantity absorbed by the liquid phase. A value of ≤10% was considered a valid run. Table S1
312
(Supporting information) shows the details.
313 314
Table 2. Operating conditions of bench-scale pilot plant Parameter
Value
Feed Gas flowrate
15 standard litres per minute (slpm)
CO2 concentration in feed gas
15%
Solvent flowrate
60 ml/min
Absorber and Desorber Catalyst weight
150 g
Average desorber temperature
85 oC
Pressure in both columns
1 atm
315 316
13 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
317 318
Figure 2. Schematic representation of bench-scale pilot plant experimental set-up34
14 ACS Paragon Plus Environment
Page 14 of 44
Page 15 of 44 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
319
Figure 3. Absorber and desorber columns packing and catalyst bed arrangement
320 321
3.5 Calculations
322
For the screening experiments, the CO2 loading was taken at various time intervals for each
323 324
sample. This was used to determine the initial absorption rate according to equation 7: 𝐶𝑂2𝑙𝑜𝑎𝑑𝑖𝑛𝑔
325
𝐼𝑛𝑖𝑡𝑖𝑎𝑙 𝑎𝑏𝑠𝑜𝑟𝑝𝑡𝑖𝑜𝑛 𝑟𝑎𝑡𝑒 =
(
𝑚𝑜𝑙 𝐶𝑂2 𝑚𝑜𝑙 𝑎𝑚𝑖𝑛𝑒
) × 𝑎𝑚𝑖𝑛𝑒 𝐶𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛 (
𝑡𝑖𝑚𝑒 (𝑚𝑖𝑛𝑢𝑡𝑒𝑠)
𝑚𝑜𝑙 𝑎𝑚𝑖𝑛𝑒 𝐿
)
(7)
326
The Kinetic analysis using the bench-scale CO2 capture pilot plant was evaluated in terms
327
of CO2 conversion and rate at which this conversion occurs. The kinetic data were obtained based
328
on the assumption of an integral type (plug-flow) reactor. The rate of reaction was determined by
329
using the differential method of analysis. CO2 conversion and rate of reaction are expressed in
330
equations 8 and 9 respectively, as:
331
𝑋
𝐶𝑂2 =
(8)
𝐶𝑂2, 𝑖𝑛 ― 𝐶𝑂2, 𝑜𝑢𝑡 𝐶𝑂2, 𝑖𝑛
15 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
332
𝑑𝑋𝐴
Page 16 of 44
𝑑𝑋𝐶𝑂2
(9)
― 𝑟𝐴 = 𝑑(𝑉/𝐹𝐴0) = 𝑑(𝑉/𝐹𝐶𝑂 0) 2
333
The absorption rates were obtained by taking the slopes of 𝑋𝐶𝑂2versus 𝑉/𝐹𝐶𝑂20 curves and
334
evaluating them at different points on the reactor. The average rate of absorption was determined
335
by taking the logarithmic mean of the rates at specific points along the column. For the absorber,
336
gas phase CO2 concentrations were obtained along the column. The desorber is not equipped with
337
sampling points along the column, hence CO2 loading (liquid phase CO2 concentration) at the top
338
and bottom of the column were used in determining CO2 conversion and finally the rate of
339
desorption. Thus, for the rate of desorption, the derivative in the above equation becomes a finite
340
difference, ∆. Other performance parameters evaluated were cyclic capacity, which represents the
341
quantity of CO2 absorbed in the liquid phase, and CO2 removal efficiency (absorber efficiency)
342
which represents the quantity of CO2 removed from the gas phase. They are represented in the
343
form:
344
Cyclic capacity (kg/hr) = 60 × 106 × FAm × MWCO2 × (αrichCAm,rich ― αleanCAm,lean)
345
Removal efficiency (%) =
VinXCO2,in ― VoutXCO2,out VinXCO2,in
×
60 × 103MWCO2 Vm,CO2
× 100%
(10) (11)
346 347
4
Results and Discussion
348
4.1 Characterization
349
The X-ray diffraction spectra for all catalysts studied are shown in S2 (Supporting
350
information). The XRD peaks of crystallized powders of BaCO3 agree with the reflection of a pure
351
orthorhombic structure and single phase of BaCO3 (witherite). This was also evident in the works of
352
Zelati et al.35 and Salehizadeh et al.36 That of CaCO3 revealed the presence of a single
353
rombohedrical crystal structure corresponding to a single calcite phase which is the most
354
thermodynamically stable form of CaCO3 predominant at room temperature. Won et al.37, Harris
355
et al.38 and Render et al.39 also reported similar phase appearance corresponding to calcite. The
356
XRD pattern of the commercially obtained Ca(OH)2 revealed a predominant portlandite phase and
357
a peak corresponding to CaCO3 which corroborates findings in the literature.40,41 The sharp distinct 16 ACS Paragon Plus Environment
Page 17 of 44 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
358
peaks present in the hydrotalcite sample reflect the presence of a layered double hydroxide which
359
is characteristic of magnesium-aluminum hydrotalcites.42-45 Other smaller peaks obtained revealed
360
the presence of gibbsite, brucite and MgO periclase phases.42 The diffraction patterns obtained for
361
Cs2O/Al2O3 and Cs2O/α-Al2O3 are also shown in S2. They both showed phases corresponding to
362
their constituents.46,47 XRD pattern of K/MgO revealed distinct phases of MgO periclase, brucite
363
(Mg(OH)2), and very subtle KOH phases. The results are in close agreement with those in
364
literature.48,49
365
SEM images of the catalysts studied are shown in S3 (Supporting information), where their
366
surface morphology and distribution of active species can be seen. CaCO3 particles show a very
367
large degree of agglomeration while that of BaCO3 displays a rather smooth appearance. It can
368
also be observed that both hydrotalcite and Cs2O/Al2O3 show a somewhat flaky appearance. The
369
surface morphology of K/MgO and Cs2O/α-Al2O3 catalysts are seen to be very porous with the
370
distribution of the active site K clearly evident for the K/MgO catalyst. The hydrothermal treatment
371
of the Cs2O/α-Al2O3 may have resulted in improving the surface characteristics of the catalyst50,51.
372
The TPD profiles of the catalysts are shown in S4 (Supporting information) except for
373
BaCO3 and CaCO3 since they do not desorb CO2 but rather undergo decomposition at very high
374
temperatures52,53. ICP analysis on K/MgO, Cs2O/Al2O3 and Cs2O/α-Al2O3 and calculated weight
375
compositions of the other catalysts are summarised in S5 (Supporting information).
376 377
4.2 Catalyst Screening Results - Initial CO2 absorption rates
378
The CO2 absorption profiles for all catalysts studied are presented in Figures 6(a) and 6(b).
379
As stated earlier, the slopes of the linear portion of these profiles were extracted to represent the
380
initial CO2 absorption rates into the solvent in the presence of the various catalysts studied.
381
Detailed plots showing the linear portions used for the initial absorption rate calculations are
382
shown in S6 (Supplementary information) and the results are summarised in Table 3.
383
17 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
0.60
Loading (mol CO2/mol amine)
0.50 blank BaCO3
0.40
K/Mgo Hydrotalcite
0.30
Ca(OH)2 Cs2O/g-Al2O3
0.20
CaCO3 Cs2O/a-alumina
0.10
K/MgO+g-alumina K/MgO + CS
0.00 0
100
200
300
400
500
600
700
800
time (minutes)
384
Fig. 6 (a) CO2 absorption profiles of various catalysts understudied
385 386
0.45 0.40
loading (mol/mol)
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 18 of 44
blank
0.35
BaCO3
0.30
K/MgO
0.25
Hydrotalcite Ca(OH)2
0.20
Cs2O/g-alumina
0.15
CaCO3
0.10
Cs2O/a-alumina
0.05
K/MgO+g-alumina K/MgO + CS
0.00 -0.05
0
50
100
150
200
250
time (min)
387 388
Fig. 6 (b) linear portion of CO2 absorption profiles
389 18 ACS Paragon Plus Environment
300
Page 19 of 44 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
390
Industrial & Engineering Chemistry Research
Table. 3 Summary of Initial rates of absorption of various catalysts studied Catalyst
Initial rate of absorption (mol/L.min) × 10
Blank (non-catalytic)
5.2
BaCO
5.6
CaCO
4.8
Hydrotalcite
5.2
Ca(OH)
6.0
3 3
2
Cs O/Al O 2
2
2
391
5.6
3
Cs O/α-Al O 2
3
3
6.0
K/MgO
6.0
K/MgO + *CS binder
6.0
K/MgO + γ-alumina binder
5.6
*CS = Colloidal Silica
392
As is well known, the catalysts behavior (Lewis or Bronsted) is dependent on the type of
393
reaction they take part in. According to the well accepted Zwitterion mechanism, the first step is
394
rate determining and involves an electron transfer26. Therefore, the behavior of interest to promote
395
this reaction is the catalyst that exhibit Lewis behavior in this first reaction step. Hence, in this
396
work, the classification of catalysts as being either Lewis or Bronsted is in relation to this reaction
397
step in the Zwitterion mechanism. Based on the above, Lewis base catalysts included K/MgO,
398
Ca(OH)2, Cs2O/Al2O3, BaCO3, Cs2O/α-Al2O3 and CaCO3 whereas Hydrotalcite was classified as
399
a Bronsted base catalyst.
400
From Figure 6 and Table 3, it was observed that K/MgO, Ca(OH)2 and Cs2O/α-Al2O3
401
recorded the fastest rates of absorption, followed by Cs2O/Al2O3, BaCO3, hydrotalcite, blank and
402
CaCO3 in decreasing order of absorption rates. The trend observed can be summarised as: K/MgO
403
~ Ca(OH)2 ~ Cs2O/α-Al2O3 > Cs2O/Al2O3 ~ BaCO3 > Hydrotalcite ~ blank > CaCO3. The reason
404
for the observed trend can be explained based on the mechanism for the CO2 reactions with amines.
405
According to Caplow26, the zwitterion formation step (first step) in the zwitterion
406
mechanism happens to be the rate determining step. As already stated, this step involves the 19 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
407
transfer of electrons. Therefore, it is imperative that an electron transfer process will be enhanced
408
in the presence of electrons. The presence of O2- anions of unsaturated co-ordination accounts for
409
the basic sites in Lewis solid base catalysts thereby possessing the ability to release their electrons
410
ahead of the amine solvents to enhance the reactivity of CO2 with amines, and later recover them
411
from these amines. This results in faster kinetics as can be seen when the initial rate of absorption
412
of the blank (no catalyst) is compared to that of the catalysts.
413
The poor performance in initial absorption rate for CaCO3 catalyst when compared with
414
the blank is due to its very low surface area which was less than 0.1 m2/g. This is supported by the
415
SEM results (S3) which revealed CaCO3 surface had a large degree of agglomeration which may
416
also have resulted in its rather poor performance. In a long run, this may have adversely affected
417
the rate of CO2 absorption thereby resulting in lower rates when compared to the blank. The
418
physical and chemical properties of the catalysts studied are shown in Table 4.
419
K/MgO recorded one of the fastest rates of absorption. Its relatively higher surface area
420
and pore volume in comparison with the other catalysts is a possible reason for its good
421
performance. Also, TPD results (S4 and table 4) showed K/MgO possessed the largest number of
422
basic sites. According to Chen et al.54, the generation of super basic sites was related to the
423
existence of O2- anion vacancies in MgO and its accompanying electrical induction effect. Jimenez
424
et al.48 reported on the interaction between K and Mg in MgO resulting in the weakening of the
425
Mg-O bonds and therefore aiding in the easy migration of the O2- anion species. In the presence
426
of the amine, these electron-rich anion species easily attack dissolved CO2, and this interaction ties
427
the CO2 molecules to the surface of the catalyst, making them readily available for the nitrogen
428
(N) atom of the amine. In this way, a greater contact time is available between the amine solvent
429
and CO2 thereby enhancing the rate of reaction. It is important to note that immediately the catalyst
430
comes into contact with aqueous amine, a small fraction of MgO phase of the calcined catalyst
431
changes to Mg(OH)2. Thus, both MgO and Mg(OH)2 are present during reaction as shown in the
432
XRD pattern of the used catalyst (Figure 7b). This is supported completely by the catalyst XRD
433
results of both the before use and after use catalysts. Another important information from the XRD
434
results of the used catalyst was the absence of carbonates (CO32-) which shows that the catalyst is
435
indeed not a reactant. This type of result has been reported by other authors48, where the absence
436
of MgCO3 was attributed to the presence of the K-promoter which easily decomposes any formed 20 ACS Paragon Plus Environment
Page 20 of 44
Page 21 of 44 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
437
carbonates on the catalyst surface. The very porous nature of the K/MgO catalyst may have also
438
contributed to its superior performance (S3).
439
Apart from O2- anion species, OH- ions also account for basicity. This was the case for
440
Ca(OH)2 which had comparable results with K/MgO. Its good performance could be due to the
441
two OH- ions which might have provided twice the effect of one OH- ion. Also, its somewhat large
442
number of basic sites and porous surface is a possible reason for its good performance as well, as
443
compared to CaCO3, BaCO3, Cs2O/Al2O3 and Hydrotalcite. There is the possibility that Ca(OH)2
444
can form carbonates. However, this was not tested and the catalyst was not selected for further use
445
in the pilot plant. BaCO3 and Cs2O/Al2O3 exhibited moderate activities which may be due to their
446
very low surface area and pore volume. BaCO3, being a carbonate, had a higher initial rate of
447
absorption than the blank (non-catalytic case). This observation nullifies any view that the
448
formation of a carbonate could stifle the reaction. The above results confirm the catalytic
449
contribution of the materials used.
450
Cs2O/α-Al2O3 performance was at par with K/MgO and Ca(OH)2. Its surface area was very
451
close to that of Ca(OH)2 (Table 4). Also, SEM image of Cs2O/α-Al2O3 indicates the possession of
452
a porous structure, which may have resulted in its good performance. The reason for this can be
453
due to the hydrothermal treatment of the α-Al2O3 which improved upon the physical properties of
454
the catalyst. Several studies have shown an improvement in the physical and chemical properties
455
of catalysts when they undergo hydrothermal treatments. This is evident in the work of Kovanda
456
et al.50 where an increase in pore size, crystallite size as well as an improvement in thermal stability
457
was observed when Ni-Al layered double hydroxides and other mixed oxides were hydrothermally
458
treated. Jung et al.51 also reported on an enhancement in the chemical stability of CuO-CeO2 where
459
cuprous ion was shown to have migrated to the surface of catalyst leading to an increase in surface
460
concentration of copper and the subsequent formation of cupric oxide on the surface of catalyst.
461
It should be noted that the catalysts are in the solid form (heterogeneous) and are insoluble
462
in the amine solvent. Therefore, being insoluble, the OH- ions are not free to react. A typical
463
example is the K/MgO-catalyzed reaction, where it is evident that the K/MgO is not reacting
464
chemically, and hence carbonates are not formed. Ca(OH)2 and Al2O3 can form carbonates if they
465
are able to dissolve. However, these catalysts were used for screening and were not selected for
466
further use in the pilot plant. 21 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
Page 22 of 44
467
Hydrotalcite barely showed any activity despite its relatively large number of basic sites.
468
It was expected that hydrotalcite should have improved upon the reaction since it is a basic catalyst
469
and considering that the rate determining step (RDS) requires an electron transfer. However, its
470
rather unappreciable performance for this reaction can be attributed to the ineffectiveness of the
471
surface oxygen anions, O2- (Lewis sites) as compared to the interlayer hydroxyl ions, OH-
472
(Bronsted sites) which are predominant in the hydrotalcite structure. 55 Also, the presence of Lewis
473
acid sites (Al3+) present may have hindered the performance of the Hydrotalcite catalyst.56
474
Despite their high activities, Ca(OH)2 and Cs2O/α-Al2O3 exhibited very poor mechanical
475
stability (i.e. ability to withstand crushing upon agitation) and were found to disintegrate easily in
476
the amine solvent even after pelletizing. On the other hand, K/MgO was found to possess good
477
mechanical stability. Consequently, K/MgO was selected and its application was transferred to a
478
bench-scale CO2 capture pilot plant. Prior to its application on the bench-scale pilot plant (which
479
is subject to agitations from process equipment), it was imperative to improve upon the mechanical
480
stability of the K/MgO catalyst without altering its activity or performance. Thus, binders such as
481
40wt% colloidal silica (CS) solution and γ-alumina were added. No change in activity was seen
482
with the CS binder but a drop in activity was observed with the γ-alumina binder (Table 3). Hence
483
the CS binder was selected for use with K/MgO.
484 485
Table 4. Physical and Chemical properties of catalysts studied Catalyst
BET surface Pore area (m2/g)
volume Pore
(cm3/g)
size Number
(nm)
basic
of Basic sites strength (oC)
*(a.u) K/MgO
63.33
0.270
17.08
58785
237
Ca(OH)2
8.43
0.042
19.89368
15565
862
Cs2O/α-Al2O3
7.61
0.041
21.44
8619
132
BaCO3
4.66
0.008
6.46
-
-
Cs2O/Al2O3
3.33
0.009
10.33
818
555
Hydrotalcite
0.46
0.0001
0.86
28258
179
CaCO3
-
-
-
-
-
22 ACS Paragon Plus Environment
Page 23 of 44 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
486
Industrial & Engineering Chemistry Research
*arbitrary units
a
b
487 488
Figure 7. XRD pattern of K/MgO (a) before use (b) after use
489 490
4.3 Statistical Analysis
491
A correlation to relate the catalyst physical properties with initial absorption rate was
492
developed. The variables employed were BET surface area, pore volume and Pore size. These
493
were all normalized to values between 0 and 1. The statistical analysis was conducted using a Non-
494
linear regression tool, NLREG® version 6.3. The R2 and R2-adjusted of the correlation were
495
0.9582 and 0.9163, respectively. The relationship between Initial absorption rate, R, and the
496
physical properties is represented as shown in equation 12:
497
Initial absorption rate, R = (4.941×10-3) + (5.929×10-3 × Surface area) - (5.853×10-3 × Pore
498
volume) + (1.236×10-3 × Pore size)
(12)
499
Another correlation to relate both physical and chemical properties of the catalysts,
500
including surface area, pore size, number of basic sites and basic strength, with the initial
501
absorption rate was also developed. Again, the independent parameters were normalized to values
502
between 0 and 1. The correlation obtained is shown in equation 13:
503
Initial absorption rate, R = (5.243×10-3) + (1.289×10-4 × Surface area × Basic Strength × Number
504
of basic sites) + (7.896×10-4 × Pore size)
(13)
23 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
505
The R2 and R2-adjusted of the correlation were 0.9594 and 0.9323 respectively. The
506
coefficients of each parameter were used to determine their impact on the initial absorption rate.
507
From the results, the most critical parameter was shown to be the Pore size followed by the
508
combined parameter involving number of basic sites, basic strength and surface area. It is observed
509
that the catalysts with the larger pore sizes are K/MgO, Ca(OH)2 and Cs2O/α-Al2O3. The results
510
indicate that, despite the combination of large number of sites, high basic strength and large surface
511
area, the catalytic performance is hinged on how well the reacting molecules have access to these
512
basic sites. This means the greater the access to the basic sites, which is potentially provided by
513
the pore size, the better the catalytic absorption performance. This accounts for the poor catalytic
514
performance of hydrotalcite, BaCO3 and Cs2O/Al2O3.
515 516
4.4 Solvent Performance and effect of solid acid catalyst in Pilot plant
517
The novel solvent, BEA-AMP was tested based on comparative runs with the conventional
518
MEA and MEA-MDEA blend at the bench-scale pilot plant level. Akachuku19, Osei et al.28,
519
Srisang et al.34 and Decardi-Nelson57 studied the use of solid acid catalyst (HZSM-5) on the latter
520
two solvents. Therefore, in this work we compared the novel solvent with these solvents on the
521
same basis of employing their solid acid catalyst in the desorber. Figures 8 and 9 show the
522
absorption and desorption rates of the 3 solvents respectively with and without desorber catalyst
523
(HZSM-5) while figures 10 and 11 show their cyclic capacities and absorption efficiencies in the
524
absence and presence of the solid acid catalyst (HZSM-5). From these figures, it can be observed
525
that the novel solvent, BEA-AMP outperformed the other two, both in the absence and presence
526
of the catalyst. For the blank case (non-catalytic run), BEA-AMP recorded a percentage increase
527
of 73.3% over MEA in absorption rate and 67.4% over blended MEA-MDEA. With the addition
528
of the desorber catalyst (HZSM-5) to all three solvent systems, values of 92.6% and 85.7% increase
529
in absorption rate for BEA-AMP over MEA and blended MEA-MDEA were respectively
530
observed.
531
The temperature profile in the absorber is shown in Figure 12. The trend is proven by these
532
charts where it is observed that the largest bulge in the absorber temperature profile is evident for
533
BEA-AMP signifying higher reactivity of this solvent over the other two in both cases of blank
24 ACS Paragon Plus Environment
Page 24 of 44
Page 25 of 44
534
and the inclusion of HZSM-5 in the desorber. Comparatively, with HZSM-5 addition, larger
535
temperature bulges in the absorber were observed over that of the blank case for all three solvents.
536
reaction rate (mol/L.min)
0.02
0.0181
0.018 0.016
0.0149
0.014
0.0118
0.012 0.01
0.0128 0.0089
0.0086
0.008 0.006 0.004 0.002 0 MEA
MEA-MDEA
BEA-AMP
solvent blank
537 538 539
HZSM-5
Figure 8. CO2 absorption rates of MEA, MEA- MDEA and BEA-AMP with and without desorber catalyst HZSM-5
540 0.045 reaction rate (mol/L.min)
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
0.041
0.04 0.034
0.035
0.035
0.03 0.025 0.02
0.023
0.024
0.0167
0.015 0.01 0.005 0 MEA
MEA-MDEA solvent
541
blank
HZSM-5
25 ACS Paragon Plus Environment
BEA-AMP
Industrial & Engineering Chemistry Research
542 543
Figure 9. CO2 desorption rates of MEA, MEA- MDEA and BEA-AMP with and without desorber catalyst HZSM-5
544 0.14 0.12 cyclic capacity (kg/hr)
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 26 of 44
0.1 0.08 0.06 0.04 0.02 0 MEA
MEA-MDEA
BEA-AMP
Solvent
545
blank
HZSM-5
546
Fig. 10. CO2 cyclic capacity of MEA, MEA- MDEA and BEA-AMP with and without
547
HZSM-5 in desorber
26 ACS Paragon Plus Environment
Page 27 of 44
60 50 40
20
}
30
HZSM-5 blank
}
absorption effiecieny (%)
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
10 0 MEA
MEA-MDEA
BEA-AMP
SOLVENT
548
by CO2 mass flow
by CO2%
by loading
549
Fig. 11. CO2 absorption efficiency of MEA, MEA- MDEA and BEA-AMP with and without
550
HZSM-5 in desorber
551
4.4.1 Absorption Performance
552
For the blank run, the 4M BEA-AMP blend had the highest CO2 absorption rate, followed
553
by 7M MEA-MDEA, with 5M MEA being the slowest. This could be explained on the basis of
554
their structural properties and lean loadings, which were developed into Absorption Parameter and
555
Desorption Parameter by Narku-Tetteh et al.8 in which BEA and AMP exhibited higher Absorption
556
Parameters and Desorption Parameter than MEA and MDEA. That explains why the BEA-AMP
557
blend outperforms MEA in the rate of CO2 absorption. The BEA-AMP bi-blend similarly
558
outperformed MEA-MDEA blend.
559
Similarly, with the incorporation of the solid acid catalyst, HZSM-5 in the desorber, an
560
identical trend was observed with BEA-AMP being the fastest while MEA was the least reactive.
561
This is because the inclusion of the catalyst in the desorber provided an alternative pathway where
562
there was a greater weakening of the N-C bond in carbamate, leading to a faster release of CO2
563
from the solvents. Hence, a considerable drop in solvent lean loadings led to higher reactivity in
564
the absorber as compared to the non-catalytic run in both columns. Lean loadings of the three
565
solvent systems are shown in table 5. The loading values reported are the sum total of all the CO2 27 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
566
groups present in each system. The speciation of the aqueous CO2-amine system for all the amines
567
studied – MEA, MDEA, BEA and AMP – are existent in literature.58-61 From these works, it is
568
clear the CO2 groups present are carbonates, bicarbonates, carbamates and free CO2. Of a certainty,
569
quantifying reaction products is important. In this work, reaction products were represented in the
570
form of the solvents’ CO2 loading, and was enough to perform calculations. An increase in
571
absorption rate of 26% was observed for BEA-AMP with the inclusion of the solid acid catalyst,
572
HZSM-5, while MEA-MDEA blend and MEA respectively had a 14% and 12% increase in the
573
absorption rate.
574 575
4.4.2 Desorption Performance
576
From Figure 9, it can be observed that for the non-catalytic case (blank), BEA-AMP blend
577
had the highest desorption rate, followed by MEA-MDEA blend, with MEA being the slowest.
578
This can be explained from the results obtained in previous studies by Narku-Tetteh et al.8, where
579
the absorption and desorption parameters were compared for several single amine solvents (Figure
580
13). It is observed that the single solvents comprising the novel bi-blend, BEA and AMP had
581
considerably higher absorption and desorption parameter values than MEA. The BEA-AMP bi-
582
blend having the highest desorption rate can be attributed to the steric hindrance effect of AMP as
583
it forms a highly unstable carbamate which easily breaks down to form bicarbonate ions and
584
enhances desorption of CO2. Also, the longer alkyl group (butyl) in the BEA structure forms a
585
highly unstable carbamate hence increasing CO2 desorption. As stated earlier, MEA has an H in
586
place of the alkyl group hence it forms a stable carbamate making it difficult to release CO2.
587
MDEA, being a tertiary amine, forms bicarbonate ions which accepts protons to form carbonic
588
acid and finally releases CO2. Hence, MDEA blended with MEA increased the desorption rate of
589
CO2 as compared to single MEA solvent. Also, BEA-AMP had a higher desorption rate than MEA-
590
MDEA.
591
Aside the effect of the sterically-hindered AMP, the secondary amine, BEA has an electron
592
donating group (butyl) attached to N, whereas MDEA has two electron-withdrawing groups (–OH
593
groups) attached to N as explained earlier. Due to this, a higher electron density is generated around
594
the N in BEA-AMP making the amine more reactive than the lower electron density-N in MEA-
595
MDEA. With the addition of the solid acid HZSM-5 catalyst, a similar trend resulted but at faster
596
desorption rates. The effect of the catalyst is seen with an increase in desorption rates for all the 28 ACS Paragon Plus Environment
Page 28 of 44
Page 29 of 44
597
three solvents. BEA-AMP recorded an increase of 17% in the CO2 desorption rate, while MEA-
598
MDEA blend and MEA had an increase of 41% and 35%, respectively.
599
The presence of the catalyst lowers the activation energy by providing an alternative
600
catalytic pathway. According to Akachuku19, HZSM-5 has both Lewis and Bronsted acid sites
601
which play an important role in increasing the rate of CO2 desorption. For a Bronsted acid site, a
602
proton is donated to the carbamate ion converting it to carbamic acid. Chemisorption on the Al
603
site weakens the N-C bond causing CO2 to break away. Also, a proton can be transferred to
604
bicarbonate ion which eventually leads to the release of CO2. For the Lewis acid site (which lacks
605
electrons), an attack is made on the high electron density Nitrogen (N), again weakening the N-C
606
bond. Consequently, CO2 breaks away. The novel solvent thus exhibited better performance than
607
the conventional solvents for the full cycle operation of the pilot plant in both cases of non-catalytic
608
and catalyst inclusion in the desorber.
609 610 611 1
Height from bottom (m)
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
0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 6
612 613
11
16
21
26
31
Temperature
(oC)
36
41
46
MEA blank
MEA-MDEA blank
BEA-AMP blank
MEA HZSM-5
MEA-MDEA HZSM-5
BEA-AMP HZSM-5
Figure 12. Temperature profile along absorber
614 29 ACS Paragon Plus Environment
51
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 parameter,*10-2 (mol CO2 absorbed)2/(mol amine.min.Lsoltn)
Industrial & Engineering Chemistry Research
0.900 0.800 MEA
0.700
AMP
0.600
BEA
0.500
tBEA
0.400
BDEA
0.300
tBDEA 4-A-1B
0.200
MDEA
0.100 0.000 0.0000
0.0200
0.0400
0.0600
0.0800
Desorption Parameter, *10-2 (mol CO2 desorbed)3/ (kJ.(Lsoltn)2.min
615
Figure 13. Amine Selection Chart8
616 617
Page 30 of 44
Table 5. Solvent lean loading of the three solvents studied Lean loading (mol
5M MEA
5/2M MEA-MDEA
2/2M BEA-AMP
0.42
0.35
0.33
0.41
0.32
0.30
CO2/mol amine) no catalyst HZSM-5
catalyst
(Si/Al = 19) 618 619
4.5 Effect of Solid base catalyst addition in Pilot plant
620
The selected solid base catalyst (K/MgO) from the screening results was transferred to the
621
absorption column of the bench-scale pilot plant and was tested on the novel solvent 4M BEA-
622
AMP. A catalyst weight of 150g and average desorber bed temperature of 85oC were used. This
623
selected catalyst weight was based on previous studies on this same set-up where 150g of a solid
624
acid catalyst (HZSM-5) for the desorption column was the optimum weight after performing a
625
sensitivity analysis.19 Absorption and desorption rates were determined for three configurations
626
shown in Table 6. The absorber temperature profiles obtained from the experiments are shown in 30 ACS Paragon Plus Environment
Page 31 of 44 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
627
Figure 14. The highest reactivity in the absorber is seen with configuration 3 (catalysts in both
628
columns) where its temperature bulge is largest. Configuration 2, the case of only HZSM-5 in the
629
desorber and no absorber catalyst, comes next, and finally, configuration 1 (solvent only) exhibits
630
the smallest bulge in temperature along the absorber.
631
From Figure 15, it can be observed that the addition of HZSM-5 to the desorber system
632
resulted in an increase in the absorption rate from about 0.015 to 0.018 mol/L.min which
633
corresponds to a 22% increment. Such observations can be linked to the lean loading of the solvent.
634
It was observed that the lean loading dropped when HZSM-5 was incorporated into the desorber
635
system. HZSM-5 contributes to better desorption performance, as explained earlier. This translates
636
into enhancing CO2 absorption since the process is cyclic. The leaner solvent means more active
637
free amines were available to react thereby resulting in an increase in the reaction rate. Upon the
638
addition of the solid base-catalyst (K/MgO) into the absorber (configuration 3), a huge
639
improvement is seen in the rate of CO2 absorption. Faster kinetics occurs resulting in a higher rich
640
loading of the amine. When compared to the case of only HZSM-5 in desorber (configuration 2),
641
an increase of 61% is made when K/MgO is placed in the absorber. A synergistic increase in
642
absorption rate of about 99% is observed with the addition of K/MgO and HZSM-5 (configuration
643
3) using configuration 1 as basis of comparison. The explanation for the observed trend is based
644
on established proof that CO2 reactions with amines proceeds through the Zwitterion mechanism
645
and that the rate determining step is the Zwitterion formation.26 As highlighted earlier, this step
646
involves the transfer of electrons. Therefore, any enhancement in electron transfer will speed up
647
the Zwitterion formation. Since K/MgO is a Lewis base catalyst (electron donor), it facilitates the
648
easy transfer of electrons which accelerates the rate limiting step, hence improving upon the overall
649
rate of reaction. Hence, a faster rate of CO2 absorption is observed. With no solid base catalyst in
650
the absorber, the electron transfer process is highly hinged on the inherent solvent characteristics
651
which results in relatively slower reaction rates as compared to when an easier electron transfer
652
facilitator (solid base catalyst) is present in the process.
653
The desorption rate had a similar trend with configuration 1 having the slowest desorption
654
rate, followed by configuration 2, and the fastest being configuration 3. The performance in both
655
columns are linked in that whatever transpires in the absorber is translated to the desorber. Since
656
the fastest rate of CO2 absorption was seen for configuration 3, it implies more CO2 was absorbed 31 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
Page 32 of 44
657
into the amine solvent, hence any little application of heat coupled with the presence of solid acid
658
catalyst (HZSM-5) in the desorber will lead to faster release of CO2. An increase of about 37% is
659
seen in desorption rate for configuration 3 when compared to configuration 2 (only HZSM-5 in
660
desorber). Using configuration 1 (non-catalytic) as basis, a synergistic improvement in desorption
661
rate of 63% is realized with the addition of both K/MgO and HZSM-5 in the absorber and desorber
662
respectively. Desorption rates are displayed in Figure 16. Table 7 shows the lean and rich loadings
663
for the three configurations studied.
664 665 666 667 668 669 670
Table 6. Absorber and desorber configurations System Configuration
Absorber
Desorber
1
Solvent
Solvent
2
Solvent
Solvent + solid acid catalyst (HZSM-5)
3
Solvent + solid alkaline catalyst Solvent + solid acid catalyst (K/MgO)
(HZSM-5)
671 672 673
32 ACS Paragon Plus Environment
Page 33 of 44
1
Height from bottom (m)
0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 6
11
16
21
26
31
36
41
46
Temperature (oC) 1
674
2
3
Figure 14. Temperature profile along absorber
675 0.035
Absorption rate (mol/L.min)
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
0.03 0.025 0.02 0.015 0.01 0.005 0 1
2
3
System Configuration
676 677
Figure 15. CO2 absorption rates of different system configurations
678
33 ACS Paragon Plus Environment
51
Industrial & Engineering Chemistry Research
0.07
Desorption rate (mol/L.min)
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 34 of 44
0.06 0.05 0.04 0.03 0.02 0.01 0 1
2
3
System Configuration
679 680
Figure 16. CO2 desorption rates of different system configurations
681 682
Table 7. Lean and Rich loadings of the different system configurations studied. Configuration
1
2
3
Lean loading (mol
0.33
0.3
0.32
0.49
0.49
0.58
CO2/mol amine) Rich loading (mol CO2/mol amine) 683 684
4.6 Effect of solid base catalyst weight
685
To determine the optimum solid base catalyst weight for the process, the catalyst weight
686
was varied while keeping the CO2 flowrate constant. The pilot plant represents a much larger scale
687
than the semi-batch absorption experiments, therefore a greater catalyst weight was required for
688
the testing. In view of that, the catalyst weights were varied from 50g to 170g to clearly observe
689
any changes in performance. Accordingly, it was evident that there was an increase in conversion
690
as the catalyst weight was increased, signifying catalytic activity until it tapered off at 170g.
691
Figures 17 and 18 display the effect of catalyst weight on the CO2 conversion in the absorber as 34 ACS Paragon Plus Environment
Page 35 of 44
692
well as the rate of CO2 absorption. It can be observed that as the catalyst weight was increased,
693
CO2 conversion also increased. More catalyst means higher availability of active surface area and
694
the physical presence of greater number of porous surfaces, thereby allowing for a greater number
695
of reacting species to have access to these additional sites, hence resulting in an increase in
696
conversion. This resulted in an increased rate of absorption, as expected, introduced by increasing
697
the weight of the catalyst. A percentage increase in conversion of about 8.5% and 5.9% was seen
698
with catalyst increments from 50 to 100g and 100 to 150g, respectively. However, generally, after
699
a weight of 150g, the conversion in the absorber is seen to be fairly constant. This might be the
700
result of the reaction having attained its thermodynamic limit and as such no increase is seen with
701
further addition of catalyst.
702 1
fractional CO2 Conversion
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
0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0g
50g
100g
150g
170g
catalyst weight
703 704
Figure 17. Effect of catalyst weight on CO2 conversion (Desorber bed temperature: 85oC; Amine
705
flowrate: 60 ml/min)
706
35 ACS Paragon Plus Environment
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
rate of absorption (mol/L.min)
Industrial & Engineering Chemistry Research
Page 36 of 44
0.04 0.035 0.03 0.025 0.02 0.015 0.01 0.005 0 0g
50g
100g
150g
170g
catalyst weight
707 708
Figure 18. Effect of catalyst weight on Absorption rate (Desorber bed temperature: 85oC; Amine
709
flowrate: 60 ml/min)
710 711 712 713
5
Conclusions
714
Out of the seven solid basic catalysts screened on a semi-batch level to select the most
715
suitable on the basis of initial rate of absorption and mechanical stability, K/MgO was shown to
716
be the most suitable as absorber catalyst in a bench-scale pilot plant.
717
Comparative solvent studies between the novel 4M BEA-AMP bi-solvent blend and
718
conventional solvents 5M MEA and blended 7M MEA-MDEA revealed better carbon capture
719
characteristics of the former over the two latter solvents. Absorption and desorption kinetics were
720
both highest for BEA-AMP blend followed by MEA-MDEA with MEA being least for both cases
721
with catalyst and without catalyst in the desorber. The solvent structural properties and lowest lean
722
loading of BEA-AMP resulted in faster reaction rate as compared to single MEA and MEA-MDEA
723
blend. For the blank case (non-catalytic run), BEA-AMP recorded a percentage increase of 73.3%
724
over MEA in absorption rate and 67.4% over blended MEA-MDEA.
36 ACS Paragon Plus Environment
Page 37 of 44 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
725
With the addition of the desorber catalyst (HZSM-5) to all three solvent systems, values of
726
92.6% and 85.7% increase in absorption rate for BEA-AMP over MEA and blended MEA-MDEA
727
respectively were observed. The inclusion of HZSM-5 resulted in leaner CO2 concentrations hence
728
resulting in higher rates of absorption. As with absorption rates, the desorption rates for the blank
729
case (non-catalytic run), followed the same trend with BEA-AMP recording a percentage increase
730
of about 110% over MEA and 45.8% over blended MEA-MDEA. With the addition of the desorber
731
catalyst (HZSM-5) to all three solvent systems, there were increments of 78.3% and 20.6% in
732
desorption rates for BEA-AMP over MEA and blended MEA-MDEA respectively. The steric
733
effect of AMP in the blend contributed to the fastest CO2 desorption rate for the BEA-AMP blend.
734
The novel blend thus was proven to be superior over the conventional MEA and MEA-MDEA
735
solvents. HZSM-5 increased desorption rates by providing an alternative pathway where
736
bicarbonate ions were produced hence resulting in the faster release of CO2.
737
Upon the addition of the solid base-catalyst (K/MgO) into the absorber, a huge
738
improvement was seen in the rate of CO2 absorption. When compared to the case of only HZSM-
739
5 in desorber (no absorber catalyst), an increase of 61% is made when K/MgO is added. A
740
synergistic increase in absorption rate of about 99% is observed with the addition of K/MgO and
741
HZSM-5 using the blank case of no catalyst in both columns (non-catalytic case) as basis of
742
comparison. K/MgO exhibited excellent absorption performance due to its superior electron
743
donating ability. Using the non-catalytic case as basis, there was a synergistic improvement in
744
desorption rate of 63% with the addition of both K/MgO and HZSM-5 catalysts in the absorber
745
and desorber, respectively.
746
The configuration of K/MgO in the absorber and HZSM-5 in the desorber greatly enhanced
747
the reactivity of the amine blend with CO2. Consequently, higher absorption and desorption rates
748
with the addition of these catalysts to the post combustion capture process implies shorter columns
749
hence huge reduction in capital costs.
750 751
Acknowledgements
752
The financial support provided by Natural Science and Engineering Research Council of Canada
753
(NSERC), Government of Saskatchewan, Clean Energy Technologies Research Institute (CETRI) 37 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
754
and Faculty of Graduate Studies and Research (FGSR), University of Regina is greatly
755
acknowledged.
756 757 758 759 760 761
References 1. Allam R., Bolland O. IPCC Special Report on Carbon Dioxide Capture and Storage. 2005 IPCC Working Group III. 109 -276. 2. Chakraborty A. K., Astarita G., Bischoff K. B. CO2 absorption in aqueous solutions of hindered amines. Chemical Engineering Science. 1986, Vol. 41. No. 4. pp. 997-1003.
762
3. Liao C. H., Li M. H. Kinetics of absorption of carbon dioxide into aqueous solutions of
763
monoethanolamine + N-Methyldiethanolamine Chemical Engineering Science. 2002, 57
764
4569 – 4582.
765
4. Mandal B. P., Biswas A. K., Bandyopadhyay S. S. Absorption of carbon dioxide into aqueous
766
blends of 2-amino-2-methyl-1-propanol and diethanolamine. Chemical Engineering Science.
767
2003, 58 4137 – 4144.
768
5. Sun W. C., Yong C. B., Li M. H. Kinetics of the absorption of carbon dioxide into mixed
769
aqueous solutions of 2-amino-2-methyl-l-propanol and piperazine Chemical Engineering
770
Science. 2005, 60 503 – 516.
771
6. Choi W. J., Seo J. B., Jang S. Y., Jung J. H., Oh K. J. Removal characteristics of CO2 using
772
aqueous MEA/AMP solutions in the absorption and regeneration process Journal of
773
Environmental Sciences. 2009, 21907–913.
774 775
7. Sutar P. N., Jha A., Vaidya P. D., Kenig E. Y. Secondary amines for CO2 capture: A kinetic investigation using N-ethylmonoethanolamine. 2012.
776
8. Narku-Tetteh J., Muchan P., Saiwan C., Supap T., Idem, R. Selection of components for
777
formulation of amine blends for post combustion CO2 capture based on the side chain
778
structure of primary, secondary and tertiary amines. Chem. Eng. Sci. 2017, Volume 170, 542-
779
560.
780
9. Muchan P., Saiwan C., Narku-Tetteh J., Idem R., Supap T., Tontiwachwuthikul P. Screening
781
tests of aqueous alkanolamine solutions based on primary, secondary, and tertiary structure
782
for blended aqueous amine solution selection in post combustion CO2 capture. Chemical
783
Engineering Science. 2017, 170 574–582. 38 ACS Paragon Plus Environment
Page 38 of 44
Page 39 of 44 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
784
10. Sharma, M.M., Danckwerts, P.V. Fast reactions of CO2 in alkaline solutions—(a) carbonate
785
buffers with arsenite, formaldehyde and hypochlorite as catalysts (b) aqueous mono-
786
isopropanol amine (1-amino-2-propanol) solutions. Chem. Eng. Sci. 1963, 18, 729–735.
787
11. Bandyopadhyay, S.S., Tarafdar, R.N., Dutta, B.K., Biswas, A.K. Determination of catalytic
788
rate constant for carbon dioxide absorption in carbonate–bicarbonate solution with arsenite
789
at high ionic strengths. Indian J. Technol. 1980, 18, 475–477.
790 791
12. Ghosh, U.K., Kentish, S., Stevens, G.W. Absorption of carbon dioxide into aqueous potassium carbonate promoted by boric acid. Energy Procedia 2009, 1 1075–1081.
792
13. Guo, D., Thee, H., da Silva, G., Chen, J., Fei, W., Kentish, S., Stevens, G.W. Borate-catalyzed
793
carbon dioxide hydration via the carbonic anhydrase mechanism. Environ. Sci. Technol.
794
2011, 45, 4802–4807.
795
14. Nicholas, N.J., da Silva, G., Kentish, S., Stevens, G.W., (2014). Use of vanadium (V) oxide
796
as a catalyst for CO2hydration in potassium carbonate systems. Ind. Eng. Chem. Res. 53,
797
3029–3039.
798
15. Phan D. T., Maeder M., Burns R. C., Puxty G. Catalysis of CO2 absorption in aqueous
799
solution by vanadate and sulfate and their application to post combustion capture.
800
International Journal of Greenhouse Gas Control. 2015, 36 60–65.
801
16. Sivanesan D., Kim Y. E., Youn M. H., Park K. T., Kim H. J., Grace A. N., Jeong S. K. The
802
salt-based catalytic enhancement of CO2 absorption by a tertiary amine medium. RSC Adv.,
803
2016, 6, 64575.
804 805
17. Idem, R., Shi, H., Gelowitz, D., Tontiwachwuthikul, P. Catalytic Method and Apparatus for Separating a Gas Component from an Incoming Gas Stream, WO Patent 12013821. 2011.
806
18. Shi, H., Naami, A., Idem, R., Tontiwachwuthikul, P. Catalytic and non-catalytic solvent
807
regeneration during absorption-based CO2 capture with single and blended reactive amine
808
solvents. Int. J. Greenhouse Gas Control. 2014, 26, 39–50.
809
19. Akachuku A. Kinetic Study of the Catalytic Desorption of Carbon Dioxide (CO2) from CO2-
810
loaded Monoethanolamine (MEA) and blended Monoethanolamine-Methyldiethanolamine
811
(MEA-MDEA) during Post-Combustion CO2 Capture from industrial flue gases. 2016,
812
MASc. Thesis, University of Regina, Canada.
39 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
813
20. Zhang, X.; Zhang, X.; Liu, H.; Li, W.; Xiao, M.; Gao, H.; Liang, Z. Reduction of energy
814
requirement of CO2 desorption from a rich CO2-loaded MEA solution by using solid acid
815
catalysts. Appl. Energy 2017, 202, 673−684.
816
21. Zhang, X.; Zhang, R.; Liu, H.; Gao, H.; Liang, Z. Evaluating CO2 desorption performance
817
in CO2-loaded aqueous tri-solvent blend amines with and without solid acid catalysts. Appl.
818
Energy 2018, 218, 417−429.
819
22. Liu, H.; Zhang, X.; Gao, H.; Liang, Z.; Idem, R.; Tontiwachwuthikul, P. Investigation of
820
CO2 regeneration in single and blended amine solvents with and without catalyst. Ind. Eng.
821
Chem. Res. 2017, 56, 7656−7664.
822
23. Shi, Huancong & Zhou, Yunlong & Zuo, Yuanhui & Cui, Lifeng & Idem, Raphael &
823
Tontiwachwuthikul, Paitoon. Heterogeneous catalysis of CO2-diethanolamine absorption
824
with MgCO3 and CaCO3 and comparing to non-catalytic CO2-monoethanolamine
825
interactions. Reaction Kinetics, Mechanisms and Catalysis. 2017. 10.1007/s11144-017-
826
1219-6.
827
24. Tanabe K. Solid Acids and Bases. Kodansha/Academic Press, 1970. Chapt. 3, p.35.
828
25. Tanabe K., Misono M., Ono Y., Hattori H. New Solid Acids and Bases. Kodansha/ Elsevier.
829 830 831 832 833
1989, Chapt. 2, p. 5. 26. Caplow, M. Kinetics of carbamate formation and breakdown. J. Am. Chem. Soc. 1968, 90 (24), 6795–6803. 27. Sartori, G., Savage, D.W. Sterically-hindered amines for carbon dioxide removal from gases. Ind. Eng. Chem. Fundam. 1983, 22 (2), 239–249.
834
28. Osei P. A., Akachuku A., Decardi-Nelson B., Srisang W., Pouryousefi F.,
835
Tontiwachwuthikul P., Idem R. Mass transfer studies on catalyst-aided CO2 desorption from
836
CO2-loaded amine solution in a Post-combustion CO2 capture plant. Chemical Engineering
837
Science. 2017, 170 508–517.
838 839 840 841
29. Cavani F., Trifirb F., Vaccari A. Hydrotalcite-type anion clays: Preparation, Properties and Application. Catalysis Today. 1991, 11 173-301. 30. Climent, M.J. Corma A., Iborra S., Velty A. Activated Hydrotalcites as catalysts for the synthesis of chalcones of pharmaceutical interest. Journal of Catalysis. 2004, 221 474–482.
40 ACS Paragon Plus Environment
Page 40 of 44
Page 41 of 44 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
842
31. D´ıez V.K., Apestegu´ıa. C.R., Di Cosimo J.I. Acid–base properties and active site
843
requirements for elimination reactions on alkali-promoted MgO catalysts. Catalysis Today.
844
2000, 63 53–62.
845
32. Gorzawski H., Hoedlerich W. F. Preparation of superbases and their use as catalysts for
846
double-bond isomerization. Journal of Molecular Catalysis A: Chemical. 1999, 144 181–187.
847
33. Ji, L., Miksche, S., Rimpf, L., Farthing, G. CO2 Chemical Solvent Screening. 2009, pp. 4–7.
848
34. Srisang, W., Pouryousefi, F., Osei, P. A., Decardi-Nelson, B., Akachuku, A.,
849
Tontiwachwuthikul, P., Idem, R. Evaluation of the Heat Duty of Catalyst-Aided Amine-
850
Based Post Combustion CO2 Capture. 2017, Chemical Engineering Science 170, 48–57.
851
35. Zelati A., Amirabadizadeh A., Kompany A. Preparation and Characterization of Barium
852
Carbonate Nanoparticles International Journal of Chemical Engineering and Applications.
853
2011, Vol. 2, No. 4.
854
36. Salehizadeh S. A., Chenari H. M., Shabani M., Ahangar H. A., Zamiri R., Rebelo A., Kumar
855
J. S., Graça M. P. F., Ferreir J. M. F. Structural and impedance spectroscopy characteristics
856
of BaCO3/BaSnO3/SnO2 nanocomposite: observation of a non-monotonic relaxation
857
behavior. RSC Adv. 2018, 8, 2100–2108.
858 859
37. Won Y. H., Jang H. S., Chung D. W., Stanciu L. A. Multifunctional calcium carbonate microparticles: Synthesis and biological applications J. Mater. Chem., 2010, 20, 7728–773.
860
38. Harris J., Mey I., Hajir M., Mondeshki M., Wolf S. E. Pseudomorphic transformation of
861
amorphous calcium carbonate films follows spherulitic growth mechanisms and can give rise
862
to crystal lattice tilting. Electronic Supplementary Information.
863
39. Render D., Samuel T., King H., Vig M., Jeelani S., Babu R. J., Rangari V. Biomaterial-
864
Derived
Calcium
Carbonate
Nanoparticles
865
dx.doi.org/10.1155/2016/3170248.
for
Enteric
Drug
Delivery.
2016.
866
40. Khachani M., El-Hamidi A., Halim M., Arsalane S. Non-isothermal kinetic and
867
thermodynamic studies of the dehydroxylation process of synthetic calcium hydroxide
868
Ca(OH)2. J. Mater. Environ. Sci. 2014, 5 (2) (2014) 615-624.
869
41. Saoud K. M., Ibala I., El-Ladki D., Ezzeldeen O., Saeed S. Microwave Assisted Preparation
870
of Calcium Hydroxide and Barium Hydroxide Nanoparticles and Their Application for
871
Conservation of Cultural Heritage. DOI: 10.1007/978-3-319-13695-0_33.
41 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
872
42. Meira1 D. M., Cortez G. G., Monteiro W. R., Rodrigues J. A. J. Vanadium Oxides Supported
873
on Hydrotalcite-Type Precursors: The Effect of Acid-Base Properties on the oxidation of
874
Isopropanol. Brazilian Journal of Chemical Engineering. 2006, Vol. 23, No. 03, pp. 351 –
875
358.
876
43. Macala G. S., Robertson A. W., Johnson C. L., Day Z. B., Lewis R. S., White M. G., Iretskii
877
A. V., Ford P. C. Transesterification Catalysts from Iron Doped Hydrotalcite-like Precursors:
878
Solid Bases for Biodiesel Production. Catal. Lett. 2008, 122:205–209.
879
44. Obadiah A., Kannan R., Ravichandran P., Ramasubbu A., Kumar S. V. Nano Hydrotalcite as
880
a Novel Catalyst for Biodiesel Conversion. Digest Journal of Nanomaterials And
881
Biostructures. 2012, Vol. 7, No. 1, January - March 2012, p. 321 – 327.
882
45. Yanming S., Dongbin L., Shifeng L., Lihui F., Shuai C., Haque M. A. Removal of lead from
883
aqueous solution on glutamate intercalated layered double hydroxide. Arabian Journal of
884
Chemistry. 2013, http://dx.doi.org/10.1016/j.arabjc.2013.08.005.
885
46. Kakooei S., Rouhi J., Mohammadpour E., Alimanesh M., Dehzangi A. Synthesis and
886
Characterization of Cr-Doped Al2O3 nanoparticles prepared via Aqueous Combustion
887
Method. Caspian Journal of Applied Sciences Research. 2012, 1(13), pp. 16-22.
888 889 890 891 892 893
47. Hu F, Wu X., Wang Y., Lai X. Ultrathin g-Al2O3 nanofibers with large specific surface area and their enhanced thermal stability by Si-doping. RSC Adv., 2015, 5, 54053. 48. Jimenez R., Garcia X., Cellier C., Ruiz P., Gordon A. L., (2006) Soot combustion with K/MgO as catalyst Applied Catalysis A: General 297 125–134. 49. Díez V.K., Apesteguía C.R., Di Cosimo J.I. Aldol condensation of citral with acetone on MgO and alkali-promoted MgO catalysts. Journal of Catalysis. 2006, 240 235–244.
894
50. Kovanda F., Rojka T., Bezdicka P., Jiratova K., Obalova L., Pacultova K., Bastl Z., Grygar
895
T. Effect of hydrothermal treatment on properties of Ni–Al layered double hydroxides and
896
related mixed oxides. Journal of Solid State Chemistry. 2009, 182 27-37.
897
51. Jung C., Kundu A., Nam S. W., Lee H. I. Selective oxidation of carbon monoxide over CuO–
898
CeO2 catalyst: Effect of hydrothermal treatment. Applied Catalysis B: Environmental. 2008,
899
84 426–432.
900 901
52. Arvanitidis I., Sichen D., Seetharaman S. A study of the thermal decomposition of BaCO3. Metallurgical and Materials Transactions B. Volume 27B, 1996-409.
42 ACS Paragon Plus Environment
Page 42 of 44
Page 43 of 44 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
902 903 904 905 906 907 908 909
Industrial & Engineering Chemistry Research
53. Galan I., Glasser F. P., Andrade C. Calcium carbonate decomposition. J Therm Anal Calorim (2013) 111:1197–1202. 54. Chen L., Zhao J., Yin S. F., Au C. T. A mini-review on solid superbase catalysts developed in the past two decades. RSC Advances. 2013, 3, 3799. 55. Ono Y. and Hattori H. Solid base catalysis. Springer-Verlag Berlin Heidelberg and Tokyo Institute of Technology Press 2011. 56. Sharma R. K., Stanley D. C., Holm P. L., Olson E. S. Hydrotalcite catalysis of hydrotreating reactions. anl.gov.
910
57. Decardi-Nelson B. A flexible and robust model for low temperature catalytic desorption of
911
CO2 from CO2-loaded amines over solid acid catalysts, 2017. MASc. Thesis, University of
912
Regina, Canada.
913
58. Jakobsen J. P., Krane J., Svendsen H. F. Liquid-Phase Composition Determination in
914
CO2−H2O−Alkanolamine Systems: An NMR Study. Ind. Eng. Chem. Res., 2005, 44 (26),
915
pp 9894–990.
916
59. Ciftja, A. F., Ardi H., Hallvard F. S. 13C NMR as a method species determination in CO2
917
absorbent systems. International Journal of Greenhouse Gas Control 16 (2013): 224-232.
918
60. Guo C., Shaoyun C., Yongchun Z. A 13C NMR study of carbon dioxide absorption and
919
desorption in pure and blended 2-(2-aminoethylamine) ethanol (AEEA) and 2-amino-2-
920
methyl-1-propanol (AMP) solutions. International Journal of Greenhouse Gas Control 28
921
(2014): 88-95.
922
61. Barzagli, F., Giorgi, C., Mani, F., Peruzzini, M. Reversible carbon dioxide capture by
923
aqueous and non-aqueous amine-based absorbents: A comparative analysis carried out by 13
924
C NMR spectroscopy. Applied Energy 220 (2018): 208-219.
925 926 927 928 929 930 931 932 933 934 935 43 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
936
Graphical Abstract - TOC
937 0.035
Absorption rate (mol/L.min)
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
0.03 0.025 0.02 0.015 0.01 0.005 0 no catalyst in both columns
solid acid catalyst in desorber only
solid acid catalyst (desorber)+solid base catalyst (absorber)
System Configuration
938 939
44 ACS Paragon Plus Environment
Page 44 of 44