ma_jacs_2009

7
“Molecular Basket” Sorbents for Separation of CO 2 and H 2 S from Various Gas Streams Xiaoliang Ma,* Xiaoxing Wang, and Chunshan Song* Clean Fuels and Catalysis Program, EMS Energy Institute and Department of Energy and Mineral Engineering, The PennsylVania State UniVersity 209 Academic Projects Building, UniVersity Park, PennsylVania 16802 Received September 18, 2008; E-mail: [email protected] (X.M.); [email protected] (C.S.) Abstract: A new generation of “molecular basket” sorbents (MBS) has been developed by the optimum combination of the nanoporous material and CO 2 /H 2 S-philic polymer sorbent to increase the accessible sorption sites for CO 2 capture from flue gas (Postdecarbonization), and for CO 2 and H 2 S separation from the reduced gases, such as synthesis gas, reformate (Predecarbonization), natural gas, coal/biomass gasification gas, and biogas. The sorption capacity of 140 mg of CO 2 /g of sorb was achieved at 15 kPa CO 2 partial pressure, which shows superior performance in comparison with other known sorbents. In addition, an exceptional dependence of MBS sorption performance on temperature for CO 2 and H 2 S was found and discussed at a molecular level via the computational chemistry approach. On the basis of the fundamental understanding of MBS sorption characteristics, an innovative sorption process was proposed and demonstrated at the laboratory scale for removing and recovering CO 2 and H 2 S, respectively, from a model gas. The present study provides a new approach for development of the novel CO 2 /H 2 S sorbents and may have a major impact on the advance of science and technology for CO 2 /H 2 S capture and separation from various gases. 1. Introduction The rapidly increasing concentration of greenhouse gas CO 2 in the atmosphere has caused serious concern for the global climate change. Carbon capture and sequestration (CCS) is considered one of the key options for mitigating the greenhouse gas emissions. 1-4 CO 2 can be captured from flue gas (Postde- carbonization), 5 separated from synthesis gas, coal/biomass gasification gas, and reformate (Predecarbonization), 6 or even captured from atmospheric air (Air-decarbonization). 7 In the production of hydrogen, 8-11 green synfuel, 6,12 city gas, and biomethane, one of the major processes is to separate and remove CO 2 and H 2 S from the reduced gases, such as synthesis gas, reformate, natural gas, biogas, coal/biomass gasification gas, and others. In addition to the greenhouse effect of CO 2 , the presence of CO 2 in the product hydrogen and fuel gas reduces significantly the energy content of the gas and lowers the efficiency in the transportation, storage, and application of the product hydrogen and fuel gas. H 2 S is corrosive to equipment and pipelines as well as poisonous to the downstream catalysts and electrode catalysts in the solid oxide fuel cell (SOFC) and proton-exchange membrane fuel cell (PEMFC). 8,13 In both CCS and the hydrogen/green-synfuel production, one of the great challenges is to separate CO 2 from flue gas or separate CO 2 and H 2 S from various process gases more economically and energy efficiently. Amine scrubbing is a dominant technology currently used in industry for removing CO 2 and H 2 S from various gas streams, as the amine solution has a higher capacity and selectivity for removing acidic gases. However, there are some major problems in this process: (1) high energy consumption, (2) low absorption/ desorption rate, resulting in a larger size scrubber for increasing the gas-liquid interface, 14 (3) the solvent loss due to the degradation and evaporation in the process, (4) material cor- rosion due to the liquid amine solution, and (5) difficulty in removing sulfur to the level required for the fuel cell applica- tions. On the basis of conventional technologies, the cost for CO 2 capture and separation from flue gas (Postdecarbonization) is estimated to represent three-fourths of the total cost of a carbon capture, storage, transport, and sequestration system. Consequently, it is highly desired to develop a novel sorbent material and a process with a high capacity, high selectivity, high regenerability, and high energy efficiency for separation of CO 2 and H 2 S from various gas streams for CO 2 capture and for hydrogen, synfuel, and biomethane production. 15 (1) Song, C. S. Catal. Today 2006, 115,2. (2) White, C. M.; Strazisar, B. R.; Granite, E. J.; Hoffman, J. S.; Pennline, H. W. J. Air Waste Manage. Assoc. 2003, 53, 645. (3) Aaron, D.; Tsouris, C. Sep. Sci. Technol. 2005, 40, 321. (4) Riahi, K.; Rubin, E. S.; Taylor, M. R.; Schrattenholzer, L; Hounshell, D. Energiagazdalkodas 2004, 26, 539. (5) Service, R. F. Science 2004, 305, 962. (6) Kintisch, E. Science 2008, 320, 306. (7) Stolaroff, J. K.; Keith, D. W.; Lowry, G. V. EnViron. Sci. Technol. 2008, 42, 2728. (8) Song, C. S. Catal. Today 2002, 77, 17. (9) Ritter, J. A.; Ebner, A. D. Sep. Sci. Technol. 2007, 42, 1123. (10) Ni, M.; Leung, D. Y. C.; Leung, M. K. H.; Sumathy, K. Fuel Process. Technol. 2006, 87, 461. (11) Haryanto, A.; Fernando, S; Murali, N.; Adhikari, S. Energy Fuels 2005, 19, 2098. (12) Tilman, D.; Hill, J.; Lehman, C. Science 2006, 314, 1598. (13) Farrauto, R.; Hwang, S.; Shore, L.; Ruettinger, W.; Lampert, J.; Giroux, T.; Liu, Y.; Ilinich, O. Annu. ReV. Mater. Res. 2003, 33,1. (14) Seader, J. D.; Henley, E. J. Separation Process Principles; John Wiley & Sons: 1998. Published on Web 04/06/2009 10.1021/ja8074105 CCC: $40.75 2009 American Chemical Society J. AM. CHEM. SOC. 2009, 131, 5777–5783 9 5777

Upload: felix-samuel

Post on 26-Oct-2014

136 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Ma_JACS_2009

“Molecular Basket” Sorbents for Separation of CO2 and H2Sfrom Various Gas Streams

Xiaoliang Ma,* Xiaoxing Wang, and Chunshan Song*

Clean Fuels and Catalysis Program, EMS Energy Institute and Department of Energy andMineral Engineering, The PennsylVania State UniVersity 209 Academic Projects Building,

UniVersity Park, PennsylVania 16802

Received September 18, 2008; E-mail: [email protected] (X.M.); [email protected] (C.S.)

Abstract: A new generation of “molecular basket” sorbents (MBS) has been developed by the optimumcombination of the nanoporous material and CO2/H2S-philic polymer sorbent to increase the accessiblesorption sites for CO2 capture from flue gas (Postdecarbonization), and for CO2 and H2S separation fromthe reduced gases, such as synthesis gas, reformate (Predecarbonization), natural gas, coal/biomassgasification gas, and biogas. The sorption capacity of 140 mg of CO2/g of sorb was achieved at 15 kPaCO2 partial pressure, which shows superior performance in comparison with other known sorbents. Inaddition, an exceptional dependence of MBS sorption performance on temperature for CO2 and H2S wasfound and discussed at a molecular level via the computational chemistry approach. On the basis of thefundamental understanding of MBS sorption characteristics, an innovative sorption process was proposedand demonstrated at the laboratory scale for removing and recovering CO2 and H2S, respectively, from amodel gas. The present study provides a new approach for development of the novel CO2/H2S sorbentsand may have a major impact on the advance of science and technology for CO2/H2S capture and separationfrom various gases.

1. Introduction

The rapidly increasing concentration of greenhouse gas CO2

in the atmosphere has caused serious concern for the globalclimate change. Carbon capture and sequestration (CCS) isconsidered one of the key options for mitigating the greenhousegas emissions.1-4 CO2 can be captured from flue gas (Postde-carbonization),5 separated from synthesis gas, coal/biomassgasification gas, and reformate (Predecarbonization),6 or evencaptured from atmospheric air (Air-decarbonization).7 In theproduction of hydrogen,8-11 green synfuel,6,12 city gas, andbiomethane, one of the major processes is to separate andremove CO2 and H2S from the reduced gases, such as synthesisgas, reformate, natural gas, biogas, coal/biomass gasificationgas, and others. In addition to the greenhouse effect of CO2,the presence of CO2 in the product hydrogen and fuel gasreduces significantly the energy content of the gas and lowersthe efficiency in the transportation, storage, and application of

the product hydrogen and fuel gas. H2S is corrosive to equipmentand pipelines as well as poisonous to the downstream catalystsand electrode catalysts in the solid oxide fuel cell (SOFC) andproton-exchange membrane fuel cell (PEMFC).8,13 In both CCSand the hydrogen/green-synfuel production, one of the greatchallenges is to separate CO2 from flue gas or separate CO2

and H2S from various process gases more economically andenergy efficiently.

Amine scrubbing is a dominant technology currently used inindustry for removing CO2 and H2S from various gas streams,as the amine solution has a higher capacity and selectivity forremoving acidic gases. However, there are some major problemsin this process: (1) high energy consumption, (2) low absorption/desorption rate, resulting in a larger size scrubber for increasingthe gas-liquid interface,14 (3) the solvent loss due to thedegradation and evaporation in the process, (4) material cor-rosion due to the liquid amine solution, and (5) difficulty inremoving sulfur to the level required for the fuel cell applica-tions. On the basis of conventional technologies, the cost forCO2 capture and separation from flue gas (Postdecarbonization)is estimated to represent three-fourths of the total cost of acarbon capture, storage, transport, and sequestration system.Consequently, it is highly desired to develop a novel sorbentmaterial and a process with a high capacity, high selectivity,high regenerability, and high energy efficiency for separationof CO2 and H2S from various gas streams for CO2 capture andfor hydrogen, synfuel, and biomethane production.15

(1) Song, C. S. Catal. Today 2006, 115, 2.(2) White, C. M.; Strazisar, B. R.; Granite, E. J.; Hoffman, J. S.; Pennline,

H. W. J. Air Waste Manage. Assoc. 2003, 53, 645.(3) Aaron, D.; Tsouris, C. Sep. Sci. Technol. 2005, 40, 321.(4) Riahi, K.; Rubin, E. S.; Taylor, M. R.; Schrattenholzer, L; Hounshell,

D. Energiagazdalkodas 2004, 26, 539.(5) Service, R. F. Science 2004, 305, 962.(6) Kintisch, E. Science 2008, 320, 306.(7) Stolaroff, J. K.; Keith, D. W.; Lowry, G. V. EnViron. Sci. Technol.

2008, 42, 2728.(8) Song, C. S. Catal. Today 2002, 77, 17.(9) Ritter, J. A.; Ebner, A. D. Sep. Sci. Technol. 2007, 42, 1123.

(10) Ni, M.; Leung, D. Y. C.; Leung, M. K. H.; Sumathy, K. Fuel Process.Technol. 2006, 87, 461.

(11) Haryanto, A.; Fernando, S; Murali, N.; Adhikari, S. Energy Fuels 2005,19, 2098.

(12) Tilman, D.; Hill, J.; Lehman, C. Science 2006, 314, 1598.

(13) Farrauto, R.; Hwang, S.; Shore, L.; Ruettinger, W.; Lampert, J.; Giroux,T.; Liu, Y.; Ilinich, O. Annu. ReV. Mater. Res. 2003, 33, 1.

(14) Seader, J. D.; Henley, E. J. Separation Process Principles; John Wiley& Sons: 1998.

Published on Web 04/06/2009

10.1021/ja8074105 CCC: $40.75 2009 American Chemical Society J. AM. CHEM. SOC. 2009, 131, 5777–5783 9 5777

LjRR_Studio1
Sticky Note
Intriguing... How different from the 'original' MB term... Original papers cited?
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
Blank experiment carried out (in the absence of the "nanoporous material")?
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
Interesting... Conformed (independently) in some of the 63 citing studies?
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
Interesting... So what is the take-home message? And why don't you summarize here (or in the Conclusions) this "fundamental understanding"?
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
So you want to remove CO2 and recover H2S? Really? (Why?)
LjRR_Studio1
Sticky Note
Big words... Justified?
LjRR_Studio1
Highlight
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
OK... but what are the typical concentrations of the two gases in the most important gas streams (e.g., products of coal gasification)? Providing lots of irrelevant information and not focusing on the essentials?
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
Higher than what???
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
If the material is mesoporous, then IT IS NOT nanoporous... Why such sloppiness? (Or just want to impress the reader with popular words?)
Page 2: Ma_JACS_2009

In our previous study, we have developed a novel nanoporousmaterial-supported polymer sorbent, called a “molecular basket”sorbent (MBS), for CO2 capture from flue gas16-19 and for H2Sremoval from fuel gas.20,21 As shown in Figure 1, the idea is toload a CO2/H2S-philic polymer sorbent, polyethylenimine (PEI),on the nanoporous material, such as MCM-41, to increase theaccessible sorption sites per weight/volume unit of sorbent, andto improve the mass transfer rate in the sorption/desorptionprocess by increasing the gas-PEI interface. Our previous studyhas indicated that MBS has some potential advantages, includinga higher capacity (90 mg-CO2/g-sorb at a CO2 partial pressureof 15 kPa),16 higher selectivity (CO2/N2 > 1000),17 no corrosionas MBS is a solid, easy regeneration (at 100 °C),16,18,20 positiveeffect of moisture on sorption capacity for both CO2 andH2S,18,21 and high sorption/desorption rate.17,20 It is becauseMBS combines the merits of both the solid nanoporous materialand the polymer sorbent as well as both the adsorbent and theabsorbent, which increases greatly the accessible sorption siteson/in the sorbent and improves the mass transfer in the sorption/desorption process.

As a part of our continuous effort in the development of MBSfor CO2 capture from flue gas and CO2/H2S removal fromvarious fuel gases, we have made some significant progress inthe present work in increasing the sorption capacity, finding anexceptional dependence of MBS on temperature for CO2 andH2S competitive sorption which has not been reported in theavailable literature to the best of our knowledge, and developingan innovative sorption process for removing CO2 and H2S,respectively, from the gas streams on the basis of the funda-mental understanding of the sorption mechanism.

2. Results and Discussion

2.1. Properties of New Generation of MBS. A new generationof MBS has been developed in our laboratory by loading 50wt% of polyethylenimine (PEI) on nanoporous SBA-15 (PEI/SBA-15), denoted as MBS-2. MBS-2 is different from the firstgeneration of MBS (PEI/MCM-41), denoted as MBS-1, whichwas developed in our previous study by loading 50 wt% of PEIon MCM-41.18,19 Some physical properties of MBS-2 and itssupport material SBA-15 are listed in Table 1 in comparisonwith those of MBS-1 and its support material MCM-41. MBS-1

had a BET surface area of 11 m2/g with 97 v % of the porevolume filled by PEI, while MBS-2 had a BET surface area of80 m2/g, higher than that of MBS-1 by a factor of 7.3, with85 v % of the pore volume filled by PEI, though both MBS-2and MBS-1 had the same weight percent of the PEI loading.SEM images of MBS-2 presented in Figure 2 show that thereare many large stacking (external) pores between the particles,21

which facilitates the diffusion of the gas from the bulk of thegas phase to the surface of the sorbent. Both SEM and N2

physisorption results indicate that the PEI was loaded insidethe pore channels of the nanoporous material. In comparisonwith the typical absorber using the amine solution, a significantadvantage of MBS-2 is the high surface area (80 m2/g of sorb),which provides a gas-sorbent interface area of 4 × 107 squaremeter per cubic meter of the sorbent bed (m2/m3). This specificsurface area is higher than that in the typical absorber in industry((2-3) × 102 m2/m3) by ∼5 orders of magnitude and higherthan that of MBS-1 by more than 6 times. The high specificarea can result in the high sorption-desorption rate with MBS-2, as the sorption-desorption rate per unit volume of sorber isdirectly proportional to the specific surface area.

It should be highlighted that PEI is not simply loaded onMCM-41 and SBA-15 through only a physical interaction.Figure 3 shows the Diffused Reflectance Infrared FourierTransform (DRIFT) spectra of SBA-15 and MBS-2 at roomtemperature in flowing N2 with KBr as the background. Bothof them were in situ pretreated in the DRIFT cell with flowingUHP N2 at 75 °C for 2 h to ensure that it was “clean” prior tothe IR study. Two sharp bands at 3747 and 1634 cm-1 and abroad band at ∼3500 cm-1 were observed over SBA-15, whichcan be assigned to hydrogen bonding in molecular H2O andthe H-O-H bend on SBA-15. After PEI loading, these bandseither disappeared completely or were significantly reduced. The

(15) Service, R. F. Science 2004, 305, 963.(16) Xu, X. C.; Song, C. S.; Andresen, J. M.; Miller, B. G.; Scaroni, A. W.

Energy Fuels 2002, 16, 1463.(17) Xu, X. C.; Song, C. S.; Andresen, J. M.; Miller, B. G.; Scaroni, A. W.

Microporous Mesoporous Mater. 2003, 62, 29.(18) Xu, X. C.; Song, C. S.; Miller, B. G.; Scaroni, A. W. Ind. Eng. Chem.

Res. 2005, 44, 8113.(19) Xu, X. C.; Song, C. S.; Miller, B. G.; Scaroni, A. W. Fuel Process.

Technol. 2005, 86, 1457.(20) Wang, X. X.; Ma, X. L.; Xu, X. C.; Sun, L.; Song, C. S. Top. Catal.

2008, 49, 108.(21) Wang, X. X.; Ma, X. L.; Song, C. S. Green Chem. 2007, 9, 695.

Figure 1. Principle for preparation of “molecular basket” sorbent (MBS).

Table 1. Physical Properties and Sorption Capacities of MBS-1,MBS-2, and Support Materials for CO2 and H2S, Respectively

sampleBET surface

area(m2 g-1)

porevolume

(cm3 g-1)

porediameter

(nm)

CO2 cap.a

mg/g ofsorb

H2S cap.b

mg/g ofsorb

MCM-41 1229 1.15 2.7 6.3 -PEI(50)/MCM-41

(MBS-1)11 0.03 0 89.2 62.6

SBA-15 950 1.31 6.6 5.0 0.034PEI(50)/SBA-15

(MBS-2)80 0.20 6.1 140 70

a Sorption at 75 °C and atmospheric pressure for a gas with 14.9 v %of CO2 and 4.3% O2 in N2. b Sorption at 22 °C and atmosphericpressure for a gas with 4000 ppmv of H2S and 20 v % of H2 in N2

Figure 2. SEM images of MBS-2.

5778 J. AM. CHEM. SOC. 9 VOL. 131, NO. 16, 2009

A R T I C L E S Ma et al.

LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
But Figure 1 suggests that (if indeed you have such 'nanotubes') you are COATING only a very small fraction of the surface, at cylinder ends... And what is even more puzzling, your Table 1 shows a collapse of the apparent (BET) surface area, which is NOT AT ALL suggested by the cartoon in Figure 1...
LjRR_Studio1
Sticky Note
Helpful to have such a cartoon... but is it not misleading??? (See below!)
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
Does this make any sense, based on the results in Table 1?? How can you accomplish this if the surface area decreases so precipitously and if the average pore size does NOT increase?
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
Why repeat, when an explanation is what the reader expects... Do you have results for the absorbent alone?
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
So much repetition... And explanations??? Hopefully forthcoming... but already detecting symptoms of avoidance of discussion of critical issues. What is exceptional about this dependence, and what exactly has the literature reported on this issue? After all, so much is known about how adsorption uptakes depend on temperature... Why not show here that you understand that literature...
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
What is 'respective' here??? Why do you (unnecessarily) use terms whose use you do not understand? Sloppy writing means sloppy thinking?
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
Why "new generation"? Again using big words without providing convincing key details...
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
How do you know that you end up with 50%, even if you start with 50% (in solution?)? And what happens if you vary this fraction?
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
These are not pores... And what do you mean by "stacking (external)"? Displaying unacceptable ignorance of basic aspects of porosity. And how is this image consistent with your Figure 1? Where is the SBA-15 part? And where is the PEI part?
LjRR_Studio1
Sticky Note
This is nonsense... CO2 (and H2S?) does not have diffusion problems through >2 nm-diameter pores at room temperature (unless you demonstrate otherwise)... What you should discuss is the diffusion of N2 at 77 K and the collapse of apparent (BET) surface area upon impregnation!
LjRR_Studio1
Rectangle
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
Excuse me??? How so??? This is absurd? The results that you show do not even SUGGEST what you say they 'indicate'... Why don't you show here the PSD results (which you mentioned in the experimental section)? You need to provide an argument that justifies this statement... Again, at critical moments, when the reader expects a coherent scientific argument, you do not provide it... Why? There is so much literature on how the pore structure changes upon introduction of a filler into a binder... Why don't you discuss it?
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
Is this high??? And isn't the surface area relevant ONLY if you have an aDsorbent, whereas for EQUILIBRIUM aBsorption the interfacial area is important onl for the KINETICS of the process? Are you not mixing apples and oranges here, just for your convenience?
LjRR_Studio1
Sticky Note
Interesting... Can verify this analysis?
LjRR_Studio1
Sticky Note
Even if this application-oriented analysis were correct, is it not premature and misplaced... Here you want to take care of the more elementary issues, which you consistently seem to avoid (so far)...
LjRR_Studio1
Pencil
LjRR_Studio1
Rectangle
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
Explained this 'collapse'? Is it apparent or real? Do you have the CO2 surface areas?
LjRR_Studio1
Sticky Note
6.3 mg/g corresponds to some 16 m^2/g, or about 1% surface coverage.
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
How determined? And PSD results? How can you maintain the pore diameter and reduce the specific surface area by an order of magnitude? Why not discuss these basic issues???
Page 3: Ma_JACS_2009

results suggest that there is a chemical interaction between thesilanol groups on the surface of SBA-15 and the amine groupsin PEI, which may form Si-O-N+H3R and/or Si-O-N+H2R.Such chemical interaction works as an anchor to the PEImolecules on the surface of SBA-15 and keeps PEI in the porechannel, resulting in an increase of the thermal stability of thesorbent and decrease of the fluidity of PEI on the surface.

2.2. Sorption Capacity of MBS-2. The sorption of CO2 onMBS-2 was conducted in a fixed-bed flow system at atemperature range from 22 to 100 °C. It was found that MBS-2at 75 °C gave the highest capacity, as also observed for MBS-1.16 The measured sorption capacities of MBS-2 at 75 °C as afunction of CO2 partial pressure are shown in Figure 4, incomparison with our previous results for MBS-1 and the datareported in the literature for some commercial and laboratorysorbents. The sorption capacity of MBS-2 is significantly higherthan those of MBS-116 and the state-of-the-art absorbents andadsorbents. MBS-2 gave a sorption capacity of 140 mg of CO2/gof sorb at 75 °C under a CO2 partial pressure of 15 kPa, whichis a typical value corresponding to the CO2 partial pressure influe gas (∼15 vol% of CO2). This capacity value is ∼50% higher

than that of MBS-1, more than 100% higher than the saturationabsorption capacity of the 15 wt% MEA aqueous solution22 and30 wt% DEA aqueous solution,23 and more than 300% higherthan the saturation absorption capacity of the 47 wt% MDEAaqueous solution24 at the same partial pressure. The capacityof MBS-2 is higher than that of the hyperbranched aminosilicasorbent (SBA-HA), reported recently by Hicks et al.,25 by∼50%, as MBS-2 has an amine group density of ∼12.3 mmol/gof sorb, which is higher than that (7.0 mmol/g of sorb) of SBA-HA by a factor of 1.8. At the CO2 partial pressure of 15 kPa,the weight-based capacity of MBS-2 is higher than that of thestate-of-the-artmaterialzeoliticimidazolateframeworks(ZIF-69)26,27

by a factor of 4 and of the metal-organic framework (MOF)28,29

by a factor of even more than 4, as shown in Figure 4. To thebest of our knowledge, MBS-2 has the highest weight-basedCO2 capacity at the partial pressure range from 10 to 100 kPaand the comparable temperature range in all state-of-the-artsorption materials reported in the literature.

The sorptive removal of H2S by using MBS-1 and MBS-2was also conducted in the fixed-bed flow system at a temperaturerange from 22 to 100 °C. Different from the CO2 sorption, bothMBS-1 and MBS-2 at 22 °C gave the highest H2S sorptioncapacity in the temperature range examined.18 Before thebreakthrough, the H2S concentration at the outlet was less than60 ppbv, which was the H2S detection limit of the instrumentemployed in our laboratory, indicating that both MBS-1 andMBS-2 are capable of removing H2S at least to a level of 60ppbv, which can meet the stringent requirement for the fuelcell applications. The measured saturation sorption capacitiesof MBS-1 and MBS-2 at 22 °C as a function of H2S partialpressure are shown in Figure 5 in comparison with 50 wt %MDEA aqueous solution30 and methanol31 reported in theliterature. MBS-2 gave a sorption capacity of 70 mg of H2S/gof sorb at a H2S partial pressure of 0.4 kPa. This value is ∼12%higher than that of MBS-1, ∼6.7 times higher than that of the50 wt% MDEA aqueous solution, and ∼10 times higher thanthat of methanol at the same H2S partial pressure.

It is interesting to note that the sorption capacity of MBS-2for CO2 is higher than that of MBS-1 by 50%, although theloading amount of PEI in MBS-1 and MBS-2 is the same. Itindicates that the support material plays an important role indetermining the sorption performance. By comparison of thedifference in the physical properties between the two supportmaterials, a much higher capacity of MBS-2 than MBS-1 maybe ascribed to the two factors of the support materials: (1) thepore diameter of SBA-15 is approximately twice that of MCM-41, and (2) the pore volume of SBA-15 (1.31 cm3/g) is higherthan that of MCM-41 (1.15 cm3/g) by ∼14%, which allows theMBS-2 prepared from SBA-15 to have a higher surface area

(22) Austgen, D. M.; Rochelle, G. T. Ind. Eng. Chem. Res. 1991, 30, 543.(23) Rebolledo-Libreros, M. E.; Trejo, A. Fluid Phase Equilib. 2004, 218,

261.(24) Sidi-Boumedine, R.; Horstmann, S.; Fischer, K.; Provost, E.; Furst,

W.; Gmehling, J. Fluid Phase Equilib. 2004, 218, 85.(25) Hicks, J. C.; Drese, J. H.; Fauth, D. J.; Gray, M. L.; Qi, G.; Jones,

C. W. J. Am. Chem. Soc. 2008, 130, 2902.(26) Service, R. F. Science 2008, 319, 893.(27) Banerjee, R.; Phan, A.; Wang, B.; Knobler, C.; Furukawa, H.;

O’Keeffe, M.; Yaghi, O. M. Science 2008, 319, 939.(28) Walton, K. S.; Millward, A. R.; Dubbeldam, D.; Frost, H.; Low, J. J.;

Yaghi, O. M.; Snurr, R. Q. J. Am. Chem. Soc. 2008, 130, 406.(29) Bastin, L.; Barcia, P. S.; Hurtado, E. J.; Silva, J. A. C.; Rodrigues,

A. E.; Chen, B. J. Phys. Chem. C 2008, 112, 1575.(30) Huttenhuis, P. J. G.; Agrawal, N. J.; Hogendoorn, J. A.; Versteeg,

G. F. J. Petrol. Sci. Eng. 2007, 55, 122.(31) Fischer, K.; Chen, J.; Petri, M.; Gmehling, J. AIChE J. 2002, 48, 887.

Figure 3. DRIFT spectra of SBA-15 and MBS-2 under N2 atmosphere at75 °C.

Figure 4. CO2 sorption capacities of MBSs as a function of CO2 partialpressure in comparison with some data reported in the literature for sometypical commercial and developing absorbents and adsorbents. (a) 15 wt% MEA/H2O at 40 °C by Austgen and Rochelle;22 (b) 30 wt % DEA/H2Oat 40 °C by Rebolledo-Libreros and Trejo;23 (c) 47 wt % MDEA/H2O at40 °C by Sidi-Boumedine et al.;24 (d) SBA-HA at 75 °C by Hicks et al.;25

(e) MOF-508b at 30 °C by Bastin et al.;29 (f) ZIF-69 at 0 °C by Banerjeeet al.;27 (g) MBS-1 at 75 °C; (h) MBS-2 at 75 °C.

J. AM. CHEM. SOC. 9 VOL. 131, NO. 16, 2009 5779

Separation of CO2 and H2S from Various Gas Streams A R T I C L E S

LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
Do the porosity results support this?
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
Why would PEI be fluid on the surface??? What is its melting point?? (Why did you give us its boiling point, and not its melting point? Regeneration at 110 oC OK?)
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
When you report such an intriguing result, and before you try to explain it using (inappropriately?) computational chemistry, tell us how the INDIVIDUAL components of this composite, and therefore complex (and ill-characterized?) material behave? Does the uptake of CO2 increase or decrease monotonically with increasing temperature? In one case you have adsorption, in the other you presumably have absorption. What does BASIC theory say about their temperature dependence?
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
How selected? Randomly? Or for convenience? Or using some hopefully objective criterion?
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
Have you told us how you know that these are indeed "state-of-the-art"?
LjRR_Studio1
Sticky Note
Interesting... and nice comparison... Should be verified by application-oriented readers... But JACS readers are (should be) interested in explanations and understanding...
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
Is weight-based capacity of greater practical interest than volume-based?
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
Suggesting what? Physical adsorption? You haven't discussed yet the basic principles of temperature dependence of adsorption vs. absorption? Why?
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
Why are you bringing in this concept here... You are discussing capacity (that is, thermodynamics presumably). What does breakthrough (kinetics) have to do with anything here???
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
This is the initial loading amount (which you can presumably control). How do you know that in the FINAl product (composite material) you have the SAME amounts?
Page 4: Ma_JACS_2009

than that from MCM-41 after the same PEI loading (50 wt %),as shown in Table 1. Both higher surface area and larger porediameter of SBA-15 may significantly increase the total numberof the accessible sorption sites in MBS-2 and thus enhance theCO2 mass transfer in the sorption process. It should be pointedout that using SBA-15 as a support improved the sorptioncapacity of CO2 more significantly than that of H2S, indicatingthat the diffusion barrier of CO2 in the bulk of PEI held in thepores may be higher than that of H2S, which will be furtherdiscussed below via the computational chemistry approach.

The effect of the moisture in the gas on the sorption capacityfor CO2 is another important issue that needs to be clarified.The moisture effect on the sorption capacity of MBS-2 for CO2

was examined by adding 3.0 v% of H2O into a simulated fluegas with 15 v% of CO2 and 4.5 v% O2 in N2. It was found thatthe presence of 3.0 v% of H2O increased the saturation capacityof MBS-2 for CO2 sorption by ∼35%, which is consistent withour previous finding for MBS-1.18 The result further confirmsthat the presence of the moisture in the gas has a promotingeffect on the sorption capacity of MBS for CO2.

2.3. Regenerability and Stability of MBS-2. For practicalapplication, the sorbent should not only possess a high sorptioncapacity and high selectivity but also have excellent regener-ability and stability in the sorption-desorption cycles. Using aTPD technique, 20 cycles of sorption-desorption experimentwere carried out. Figure 6 shows the measured sorption capacityof MBS-2 for CO2 as a function of the number of thesorption-desorption cycles. During the 20 cycles, the CO2

sorption capacity of MBS-2 was kept at ∼170 mg of CO2/g ofsorbent at the CO2 partial pressure of 100 kPa, and no significantchange in the CO2 sorption capacity was observed. Thedesorption can be conducted by increasing the temperature to110 °C, and the sorption capacity of the spent MBS-2 can berecovered completely after the regeneration, which is consistentwith those observed in our previous studies in the regenerationof MBS-1 for CO2 sorption16,18 and the regeneration of MBS-2for H2S sorption.20 It should be mentioned that in our prelimi-nary experiment we have found that the coexisting SOx and NOx

in the real flue gas caused the degradation of MBS-2 due toformation of the heat stable amine salts with PEI in MBS-2, as

also observed in the sorption-desorption of CO2 over MBS-1using real flue gas.18 It implies that the strong acidic gases, SO2

and NO2, need to be removed when using MBS-2 for CO2

capture from flue gas, which is the same as that in the aminescrubbing process.

2.4. Dependence of MBS Sorption Performance onTemperature. Figure 7 shows the breakthrough curves onMBS-1 for CO2 and H2S, respectively, at 22 and 75 °C undera gas hourly space velocity (GHSV) of 1011 h-1 in a fixed-bedflow system using a model gas containing CO2 or H2S in N2.For CO2 removal, the sorption at 75 °C gave a much highersorption capacity (70.8 mg of CO2/g of sorb) than that at 22 °C(27.7 mg of CO2/g of sorb). In distinct contrast, H2S sorptionat 22 °C gave a significantly higher sorption capacity (87.0 mgof H2S/g of sorb) than that at 75 °C (7.8 mg of H2S/g of sorb).These results clearly revealed the significant differences in thetemperature dependences of MBS-1 for CO2 and H2S sorption,respectively. In many practical cases, CO2 and H2S coexist inthe gas streams, such as fuel gas, syngas, and biogas. To clarifywhether the presence of CO2 in the gas streams inhibits thesorption of H2S on the sorbent, a sorption experiment with a

Figure 5. H2S sorption capacities of MBSs as a function of H2S partialpressure in comparison with data reported in the literature for some typicalcommercial absorbents. (k) MBS-2 at 22 °C; (l) MBS-1 at 22 °C; (m) 50wt % MDEA/H2O at 25 °C by Huttenhuis et al.;30 and (n) MeOH at 25 °Cby Fischer et al.31

Figure6. CO2 sorptioncapacityversus thenumberof thesorption-desorptioncycles over MBS-2 measured by TPD method. The CO2 sorption wasperformed at 75 °C under a pure CO2 flow for ∼30 min. The desorptionwas carried out at 110 °C under a helium flow at a rate of 5 °C/min for 20min.

Figure 7. Sorption breakthrough curves for CO2 and H2S in single-stagesorption process using a model gas with 1.00 v% of CO2 or 1.00 v% ofH2S in N2 over MBS-1.

5780 J. AM. CHEM. SOC. 9 VOL. 131, NO. 16, 2009

A R T I C L E S Ma et al.

LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
How so???? If these were important, then you would NOT need the polymer... This is getting to be so confusing and unconvincing... The reader's 'fault'? Or the writers'?
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
This is a typical example of fuzzy thinking, or at least fuzzy writing... What is in the bulk of PEI? Presumably CO2... Right? So CO2 is presumably absorbed. Now, what is in the pores? Presumably PEI. Right? Now, how can the above mentioned result "indicate" anything about the diffusion barrier for CO2? One can make the ASSUMPTION that the cause of the improved sorption capacity is some property of SBA-15, but if this capacity is an EQUILIBRIUM PROPERTY, as it usually is, the diffusion barrier should be IRRELEVANT.
LjRR_Studio1
Highlight
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
OK... But have you "clarified" anything here? WHY does the presence of moisture have such an effect? Is that typical? Has a similar effect been reported by other investigators? What has it been attributed to? Do you support such an interpretation?
LjRR_Studio1
Sticky Note
Why not cite here some examples where this has been shown to be a problem? When you have weak chemisorption or physisorption, this is NOT typically a big problem.
LjRR_Studio1
Sticky Note
And H2S? (Is this figure really necessary? One sentence of text would be enough to describe these results...)
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
What is "respective" here??? Haven't BOTH gases been analyzed at 22 and 75 oC?
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
Isn't it more important to show the behavior of MBS-2 (if only one of the two is shown)?
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
Comment on this disagreement w/r to Table 1?
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
Why such a large discrepancy w/r to Table 1??? Comment??? You again do not comment on what the reader expects you to discuss, and waste time on unnecessary or obvious statements... WHY??
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
These results "revealed" NOTHING... They simply show the (hopefully reproducible!) behavior of this material: unusual temperature dependence, which now needs to be rationalized and explained on the basis of some elementary thermodynamics (or kinetics?). Only then would you reveal the reasons for such behavior. This needs to be done BEFORE using computational quantum chemistry, which may NOT be the appropriate tool in this case. Another example of your inappropriate use of big words, which upsets an inquisitive reader...
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
What is the relationship between this sentence and the previous discussion? Why is it in the same paragraph as the previous sentence? And does it belong in the section devoted to the temperature dependence?
Page 5: Ma_JACS_2009

model gas containing 0.40 v % of H2S, 2.40 % of CO2, and20 v % of H2 in N2 was performed at 22 °C and 1011 h-1 ofGHSV. It was found that at this condition both CO2 and H2Sbroke through at the beginning, as shown in Figure 8, indicatingthat MBS-1 failed to remove H2S and/or CO2 at this condition.

The questions that arose are why does MBS-1 performanceshow a varying dependence on operating temperature for CO2

and H2S, why does MBS-1 have a lower sorption capacity forCO2 than that for H2S at 22 °C but shows the opposite trend at75 °C, and why does the presence of CO2 strongly inhibit thesorption of H2S at 22 °C. To answer these questions, acomputational study was conducted by using a semiempiricalquantum chemical calculation method. It was hypothesized thatthe sorption of CO2 or H2S on MBS involves two steps: theadsorption of CO2 or H2S on surface of PEI and the diffusionof the adsorbate from the surface into the bulk of PEI in pores.The thermodynamic parameters of the heat of adsorption onthe PEI surface and kinetic barrier of the diffusion from site tosite in the PEI bulk for both CO2 and H2S were estimated,respectively, by semiempirical calculations. The results showthat there are two types of sorption sites (site-I and site-II) forboth CO2 and H2S sorption on PEI. The adsorption conforma-tions of CO2 and H2S on site-I and site-II are shown in Figure9. On site-I, the H2S molecule interacts with a nitrogen atom inthe amine group through the H atoms in H2S, while the CO2

molecule interacts with the nitrogen atom through the C atomin CO2. On site-II, sorption of H2S molecule is through aninteraction of the two H atoms in H2S, simultaneously, withtwo N atoms in two amine groups (in two neighboring PEIchains), while sorption of CO2 is through an interaction of theC atom in CO2 with the two N atoms simultaneously in twoamine groups. The relative sorption heat of CO2 and H2S andthe diffusion barrier in the bulk of PEI are shown in Figure 10.The results indicate that the heat of adsorption for CO2 on theamine group is higher than that for H2S, as CO2 has strongeracidity than H2S, which is consistent with the experimental heatof absorption in the amine solution reported in the literature.32

It implies that the thermodynamics favors the adsorption of CO2

on the PEI surface more than that of H2S. On the other hand, itis of interest to note that the estimated kinetic barrier fordiffusion of the sorbed CO2 from the surface into the bulk ofPEI is higher than that for diffusion of the sorbed H2S by afactor of ∼3, indicating that the diffusion of the sorbed CO2

from the exposed surface of PEI into the bulk of PEI is muchmore difficult than that of the sorbed H2S.

On the basis of the computational results, the lower CO2

sorption capacity at 22 °C than that at 75 °C can be ascribed tothe higher kinetic barrier for diffusion of the CO2 sorbed fromthe surface into the bulk of PEI, which reduces significantlythe total number of the accessible sorption sites for CO2 at 22°C, although low temperature thermodynamically favors theadsorption of CO2 on the surface of PEI. The increase intemperature facilitates the transfer of the adsorbed CO2 mol-ecules from the surface into the bulk of PEI by overcoming thekinetic barrier. This leads to a significant enhancement of thetotal number of the accessible sorption sites at 75 °C, althoughthe increase in temperature does not favor thermodynamicallythe sorption of CO2 on the surface and in the bulk of PEI. Onthe other hand, the higher heat of sorption for CO2 makes thesorption affinity at 75 °C high enough to capture CO2. As a(32) Posey, M. L.; Rochelle, G, T. Ind. Eng. Chem. Res. 1997, 36, 3944.

Figure 8. Sorption breakthrough curves for CO2 and H2S in single stageand two-stage sorption processes using a model gas containing 0.40 v%H2S, 2.40 v% CO2, and 20 v% H2 in N2 over MBS-1.

Figure 9. Computationally optimized sorption conformation of CO2 andH2S on Site-I and Site-II in PEI.

Figure 10. Potential energy surface for sorption and transfer of CO2 andH2S on Site-I and Site-II in PEI.

J. AM. CHEM. SOC. 9 VOL. 131, NO. 16, 2009 5781

Separation of CO2 and H2S from Various Gas Streams A R T I C L E S

LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
You have this (arguably very bad) habit of confusing cause and effect and using CIRCULAR arguments... When you say 'indicating', what goes before this word ("both CO2 and H2S broke through at the beginning") should be the effect and what goes it after should be the presumable cause. And yet you simply restate the effect ("MBS-1 failed to remove H2S and/or CO2 at this condition"). And why are you saying "and/or" when it's clear that BOTH H2S and CO2 are NOT removed in this case? Now, this is a very intriguing and potentially disturbing result. Why don't you discuss it? Has a similar behavior been reported for other adsorbents? Has it been explained? Again you avoid the ESSENTIAL arguments... Why?
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
Again MBS-1... Why not MBS-2?
LjRR_Studio1
Sticky Note
Are these results consistent with what is shown in Figure 7? Why does CO2 break through LATER when its concentration in this case is HIGHER than in Figure 7? Shouldn't it break through EARLIER? And why don't you discuss the KINETICS of sorption in both figures? Isn't that the MAIN usefulness of breakthrough curves (i.e., their comparison with equilibrium capacity measurements)?
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
Or vice versa? How do you know that CO2 inhibits H2S? Maybe H2S inhibits CO2? Does it matter? And has this been reported for other adsorbents?
LjRR_Studio1
Sticky Note
Wrong tool? Discuss first ELEMENTARY thermodynamics of adsorption/absorption... And of course you should also consider the kinetics of the process, because when the uptake increases with increasing temperature, that's usually attributed to kinetics and/or activated diffusion. Unless you first show that these simple explanations are inadequate, an informed reader will have difficulty believing in any arguments constructed on the basis of (more sophisticated and arguably inappropriate)computational chemistry results.
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
Here again you seem to be confusing hypothesis and conclusion and thus engaging in CIRCULAR argumentation... You seem to want to conclude this and you are now presenting it as a hypothesis. BUT you need to JUSTIFY this hypothesis based on some literature reports, or based on some of your own results. Why is such a hypothesis reasonable? Which results suggest it?
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
Again!? You like this word... If you like it so much, shouldn't you learn how to use it properly?
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
Too strong? Why would there be only two types of sorption sites??? You are analyzing a trimer, and ignoring any interaction with the 'substrate'... And haven't you argued in conjunction with Figure 3 that there is some chemical interaction between the two components of this COMPOSITE material? And haven't you also said that SBA-15 and MSC-41 behave differently? So now you are (for convenience?) ignoring all that? Is that your "scientific method"?
LjRR_Studio1
Sticky Note
Let's assume that this is OK... But shouldn't you at least show that other conceivable interactions are less favorable?
LjRR_Studio1
Sticky Note
So why do you need computational chemistry for this, if it's a simple acidity argument? Isn;t it more important to discuss the energies involved in these processes, and whether their relative values make sense? What are the units in Figure 10?
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
What is ST? The reader should not have to look for this information in the text...
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
How do you know this???? You are again withholding the essential information... Where is this number? What does it represent? In Section 4 you make vague reference to transition state calculations... But HOW EXACTLY are such calculations related to this statement about the diffusion of CO2 from the surface into the bulk???
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
Wow... This is too much even for a casual reader... Again confusing cause and effect and restating the same argument as if one were the presumed effect and the other the presumable cause.
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
Isn't this exactly what you POSTULATED (or assumed)? And now you are "concluding" exactly the same? Isn't this another (and a very disappointing) example of a circular argument, confusing cause and effect? What you need to show, and discuss in much more detail" is how exactly the very cryptic results shown in Figure 10 (usually JACS reviewers require that supporting information for the computational 'experiments' be provided) support the validity of this statement. How is the calculation of these adsorption energies related to the diffusion of CO2 from the surface to the bulk??? Instead you just repeat the same sentences in the Abstract, in the Introduction, here in what is supposed to be a "discussion section", and then in your Concluding Remarks... Why? Deluding yourself (or deceiving the reader) that you know what you are talking about?
Page 6: Ma_JACS_2009

result, MBS-1 exhibited a higher sorption capacity for CO2 at75 °C than at 22 °C. A further increase of temperature above75 °C reduces the CO2 sorption capacity, as the control of thesorption shifts from the kinetic regime to the thermodynamicregime.

For sorption of H2S, due to a lower kinetic barrier for thetransfer of the sorbed H2S in the bulk of PEI, the H2S moleculesadsorbed on the surface are easier to diffuse into the PEI bulkeven at 22 °C. On the other hand, the lower temperaturethermodynamically favors the increase in the equilibriumsorption capacity. Consequently, MBS-1 exhibited a highersorption capacity for H2S at 22 °C than at 75 °C. Due to thesignificantly higher heat of sorption for CO2 than for H2S, thecoexisting CO2 molecules preferentially occupy the sorption siteson the surface and, thus, block the way for sorption of H2S onthe surface, as indicated in Figure 8. Both the thermodynamicand kinetic factors work together to determine the exceptionaldependence of MBS performance on temperature for CO2 andH2S sorption. The higher kinetic diffusion barrier for CO2 thanfor H2S also explains why the sorption performance of MBS-2is much better than that of MBS-1 for CO2, but the sorptionperformance of both are almost the same for H2S. It is probablybecause for CO2 sorption, but not for H2S sorption, the diffusionis a primary factor that affects the total number of the accessiblesorption sites, and the higher surface area of MBS-2 and largerpore diameter of SBA-15 in MBS-2 facilitate the diffusion ofCO2 in MBS-2.

2.5. Two-Stage Process for Respective Removal of CO2

and H2S. On the basis of the fundamental understanding of thesorption mechanism, a novel two-stage sorption process wasproposed with two sorption beds in series for removing CO2

and H2S, respectively, from gas streams. The two-stage processwas demonstrated in our laboratory by using MBS-1 and amodel gas with 0.40 v % of H2S, 2.40 v % of CO2, and 20 v %of H2 in N2 gas at a flow rate of 60 mL/min. A scheme of theexperimental two-stage process is shown in Figure 11. The firststage with MBS-1 as a sorbent was operated at 75 °C forremoving CO2, and the second stage with the same sorbent atroom temperature (22 °C) for removing H2S. The CO2 concen-tration at the outlet of the first stage and the H2S concentrationat the outlet of the second stage as a function of time are plottedin Figure 8, for comparison. In the first stage, the CO2

breakthrough time was ∼95 min, corresponding to a break-through capacity of 80 mg of CO2/g of sorb, indicating that the

sorption of CO2 at 75 °C is not affected by the coexistence ofH2S, as expected. In the second stage, the H2S breakthroughtime was ∼95 min, corresponding to a capacity of 19 mg ofH2S/g of sorb. This value was lower than the expected one,because the CO2 that had broken through in the first sorptionbed entered the second sorption bed and inhibited the H2Ssorption in the second bed, resulting in the early H2S break-through in the second bed. The results clearly show that thetwo-stage process successfully removed CO2 and H2S, respec-tively, from the simulated fuel gas. It indicates that the developedprocess has a potential and wide application in the cleanup ofthe reduced gases, including hydrogen, reformate, synthesis gas,natural gas, biogas, coal/biomass gasification gas, and others.

3. Concluding Remarks

In summary, a new generation of “molecular basket” sorbent,MBS-2, has been developed in our laboratory for CO2/H2Scapture. MBS-2 gives a high sorption capacity of 140 mg ofCO2/g of sorb at 75 °C under 15 kPa CO2 partial pressure, whichis ∼50% higher than that of the previously developed MBS(MBS-1). MBS-2 shows the highest CO2 capacity under theCO2 partial pressure range from 10 to 100 kPa in the comparabletemperature range among all the commercial and state-of-the-art adsorbents, absorbents, and sorbents reported to date.

The exceptional dependence of MBS performance on tem-perature for CO2 and H2S sorption has been found and explainedat a molecular level via the computational chemistry approach.On the basis of the new findings, an innovative two-stageprocess for removing CO2 and H2S, respectively, from gasstreams was proposed and demonstrated in a laboratory ap-paratus. The sorbent and process developed in this work havemany distinct advantages: (1) high sorption capacity andselectivity for CO2 and H2S, (2) capable of removing H2S toless than 60 ppbv, (3) higher sorption-desorption rate due tohigher gas-sorbent interface area, (4) good regenerability andstability in sorption-desorption cycles, (5) promoting effect ofmoisture in the gas on the sorption capacity, and (6) ability toremove and recover CO2 and H2S, respectively.

The present study provides a new approach for the develop-ment of novel sorbents by the combination of a solid nanoporousmaterial and a polymer sorbent, of an adsorbent and anabsorbent, and of inorganic and organic materials, which mayhave a major impact on the advance of science and technologyfor CO2 capture from flue gas, CO2/H2S separation from variousreduced gases, and other gas separation.

4. Experiment and Calculation Method

4.1. Preparation of MBS. MBS-1 and MBS-2 were preparedby loading polyethylenimine (PEI) on mesoporous molecular sievesMCM-41 and SBA-15, respectively, using a wet impregnationmethod. PEI used in the present study was a linear polymer, whichwas purchased from Aldrich with an average molecular weight of423 g/mol, boiling point of ∼250 °C, and viscosity of 200 cP at25 °C. MCM-41 and SBA-15 were synthesized by a hydrothermalmethod. MCM-41 was synthesized from a mixture with thefollowing composition: 50SiO2/4.32Na2O/2.19(TMA)2O/15.62CTAB/3165H2O, which was established in our laboratory,33,34 based onthe method invented by Mobil researchers.35,36 The SBA-15 was

(33) Reddy, K. M.; Song, C. S. Catal. Lett. 1996, 36, 103.(34) Reddy, K. M.; Song, C. S. Stud. Surf. Sci. Catal. 1998, 117, 291.(35) Kresge, C. T.; Leonowicz, M. E.; Roth, W. J.; Vartuli, J. C.; Beck,

J. S. Nature (London) 1992, 359, 301.

Figure 11. Scheme of the experimental two stage process for removal ofCO2 and H2S from a model fuel gas. Inlet gas: 0.40 v % H2S, 2.40% CO2,and 20% H2 in N2; sorbent in both the first and second stages: MBS-1.

5782 J. AM. CHEM. SOC. 9 VOL. 131, NO. 16, 2009

A R T I C L E S Ma et al.

LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
OK... So why do you call these 'nanoporous' materials?????
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
Isn't the melting point of equal or greater interest?
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
More details on how exactly this method was implemented? Again withholding essential information, while spending too much time using big words and grandiose statements... For example, how do you control and measure the wt% of the components in this composite material? What does 50 in Table 1 mean?
LjRR_Studio1
Sticky Note
So many questions remain unanswered, even the basic ones... Does any of the 63 citing papers clarify some of them? Is it fair to conclude that the authors have some apparently very interesting (or intriguing?) results, but that their discussion, and especially the explanations offered, are rather disappointing (or at least frustrating for an informed and inquisitive reader)? Still, the analysis of this paper may be quite instructive for the methodology of paper writing (which is the objective of FSC 506)...
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
Now this may be the only sensible statement so far... At the lower temperatures you may indeed have a diffusion-limited CO2 uptake, which is common for moelcular-sieving materials. So instead of spending (wasting?) time discussing the highly speculative computational chemistry results, wouldn't it be more convincing to discuss the PSD results and to discuss Figures 7 and 8 in terms of kinetics rather than thermodynamics?
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
WHY would this barrier be lower??? This was your assumption, mostly related to thermodynamics, and based on a simple acidity argument, which presumably your convenient ASSUMPTION about the nature of interaction between POSTULATED types of sites and H2S presumably confirms... But what would be the INDEPENDENT argument that confirms the validity of these ASSUMPTIONS? Is H2S a smaller molecule than CO2? Why does it not have a similar kinetic barrier. All this is discussed in the relevant literature that you do not cite... Why?
LjRR_Studio1
Sticky Note
You keep repeating the same statements and consistently shy away from key arguments: WHY would diffusion be a "primary factor" for CO2 and not for H2S?
LjRR_Studio1
Sticky Note
Now the reader is UTTERLY confused... Didn't you say that the 'reason' is the diffusion kinetic energy, as 'revealed' by your computational chemistry results which have NOTHING to do with the pore diameter of the substrate? Now, you seem to be changing your story by invoking differences in pore diameter... Why? Especially when you really have not explained where these values come from, how they were determined, how wide is the pore size distribution and how difficult it is to "define" a single characteristic pore diameter... All this is puzzling, since you did mention in Section 4 that you have PSD data...
Page 7: Ma_JACS_2009

synthesized according to the procedure reported in the literature.37,38

Typically, a homogeneous mixture, which was composed of triblockcopolymer Pluronic P123 (EO20PO70EO20, MW ) 5800, Aldrich)and tetraethyl orthosilicate (TEOS) in hydrochloric acid, was stirredat 40 °C for 20 h and then further treated at 100 °C for 24 h. Afterthe synthesis, the resultant solid was recovered by filtration, washedand dried at 100 °C overnight, and finally calcined at 550 °C for6 h. The typical preparation methods of MBS-1 and MBS-2 werereported in detail in our previous paper.20

4.2. Characterization of Sorbents. The pore structure andsurface area of the prepared MCM-41, SBA-15, MBS-1, and MBS-2were characterized by adsorption/desorption of nitrogen at -196°C using a Micromeritics ASAP2010 surface area and porosimetryanalyzer. Standard BET and DR models were respectively appliedto derive surface area and pore volume. The pore size distributionwas calculated according to the Barrett-Joyner-Halenda (BJH)model.39 The scanning electron microscopy (SEM) images wereobtained on a Hitachi S-3500N instrument operating at 5 kV.

A Nicolet NEXUS 470 FT-IR spectrometer (Thermo ElectronCorp.) was used to obtain the DRIFT Spectra of the SBA-15 andMBS-2 samples. The powder of each sample (∼20 mg) was placedinto the DRIFT cell and pretreated in flowing UHP N2 at 75 °C for2 h. Then the DRIFT spectra were collected under N2 atmosphereat 75 °C. KBr was used as the background at the same conditions.The IR resolution was 4 cm-1.

4.3. Sorption Measurements. The sorption separation of CO2

and/or H2S from the model gases was performed in a fixed-bedflow sorber (straight glass tube with inner diameter of 9.5 mm)operated at atmospheric pressure. Special tubing and fittings coatedwith a sulfur inert material (purchased from Restek Corp.) wereused for the sorption system to reduce the effect of adsorption ofH2S on the tubing wall. In a typical sorption process, ∼1.5 g ofthe sorbent was packed into the bed (bed length was ∼75 mm),and the empty spaces were filled with inert glass beads. Before thesorption test, the sorbent bed was heated to 100 °C in nitrogen ata flow rate of 100 mL/min and held overnight to desorb allpresorbed species. Then, the sorbent bed was cooled down to roomtemperature, and the whole sorber was sealed and weighed tocalculate the real weight of the used sorbent. After the sorber wasconnected back into the system, the sorbent bed was heated to thedesired sorption temperature, and then, the model gas wasintroduced into the bed. The treated gas out of the sorber wasanalyzed online by using an SRI 8610C gas chromatograph withmolecular sieve 5A and Porapak T columns and with a TCDdetector (GC-TCD), and an ANTEK 9000NS Sulfur Analyzer,respectively, for CO2 and H2S. For the gas samples with H2Sconcentration less than 1 ppmv, a gas detection system, Sensidyne/Gastec, was used. The saturation capacity, which was denoted asCap, milligram of CO2 or H2S per gram of sorbent (mg/g of sorb),was calculated by using the following equation:

Cap)MW × FR ×∫o

t(Co -Ct) dt

Wsorb(1)

where t is the sorption time (min); FR is the flow rate (mmol/min);Wsorb is the weight of sorbent (g); MW is the molecular weight(g/mol) of CO2 or H2S; and Co and Ct are the inlet and outletconcentration of CO2 or H2S.

4.4. Evaluation of Regenerability and Stability of MBS-2. Thetemperature-programmed desorption (TPD) method was used tomeasure the regenerability and stability of MBS-2 for CO2 sorption.CO2-TPD was conducted by using a Micromeritics AutoChem 2910instrument with a TCD. MBS-2 (∼100 mg) was loaded into aU-shape quartz reactor and pretreated at 100 °C under pure heliumfor 30 min. Then the temperature was cooled down to 75 °C, andCO2 sorption was performed at this temperature under a pure CO2

flow for ∼30 min. After that, the temperature was decreased toroom temperature under CO2 flow. The desorption experiments werethen carried out by passing helium through the tube and rampingthe temperature from room temperature to 110 °C at a rate of 5°C/min and holding at 110 °C for 20 min. The sorption capacity ofthe regenerated MBS-2 was measured on the basis of the amountof the desorbed CO2. The same sorption-desorption procedure wasconducted for 20 cycles to evaluate the regenerability and stabilityof MBS-2.

4.5. Computational Method. All quantum chemical calculationsin this study were performed by means of the semiempirical PM5method, using the CAChe program. To reduce the computationalcost, a simple model (a trimer of ethylenimine (TEI)):

NH2-CH2-CH2-NH-CH2-CH2-NH-CH2-CH3

which contains one primary amine group and two secondary amines,was used to mimic PEI. The adsorption conformations of CO2 andH2S on the secondary amine in TEI were optimized by the PM5method. The pseudo heat of sorption was defined and estimatedon the basis of the following equation:

∆Hsorption )∆Hof,TEI-sorbate - (∆Hof,TEI +∆Hof,sorbate) (2)

where ∆H°f,TEI and ∆H°f,sorbate are the heat of formation of TEI andsorbate, respectively. ∆H°f, TEI-sorbate is the heat of formation of TEI-sorbate (heat of formation for whole TEI and sorbate afterinteraction). In the present study, we only examined the interactionbetween CO2 (or H2S) and the secondary amine group in PEI, aswe used a linear PEI for preparing MBS-1 and MBS-2, which hada secondary-amine/primary-amine ratio of ca. 9 according to theaverage molecular weight.

The kinetic barrier (Et) for transfer of the sorbed molecule fromone sorption site to the other was estimated by finding the transitionstate and calculation according to the following equation:

Et )∆Hof,ST -∆Hof,TEI-sorbate (3)

where, ∆H°f, ST is the heat of formation of the transition state fromone site to the other.

Acknowledgment. This study was supported in part by Penn-sylvania Energy Development Authority through PA Departmentof Environmental Protection Grant PG050021, by the U.S. Officeof Naval Research through ONR Grant N00014-08-1-0123, andby the U.S. Department of Energy, National Energy TechnologyLaboratory through DOE Grant DE-FC26-08NT0004396.

JA8074105

(36) Beck, J. S.; Vartuli, J. C.; Roth, W. J.; Leonowicz, M. E.; Kresge,C. T.; Schmitt, K. D.; Chu, C. T. W.; Olson, D. H.; Sheppard, E. W.;McCullen, S. B.; Higgins, J. B.; Schlenker, J. C. J. Am. Chem. Soc.1992, 114, 10834.

(37) Zhao, D.; Feng, J.; Huo, Q.; Melosh, N.; Fredrickson, G. H.; Chmelka,B. F.; Stucky, G. D. Science 1998, 279, 548.

(38) Wang, X.; Zhang, Q.; Yang, S.; Wang, Y. J. Phys. Chem. B 2005,109, 20835.

(39) Choma, J.; Jaroniec, M.; Kloske, M. Sep. Sci. Technol. 2002, 20, 307.

J. AM. CHEM. SOC. 9 VOL. 131, NO. 16, 2009 5783

Separation of CO2 and H2S from Various Gas Streams A R T I C L E S

LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
Really used DR? And where are the PSD results?
LjRR_Studio1
Sticky Note
Hmmm... for H2S the ENTIRE assembly may require such a (Teflon?) coating... Shown the results of a blank experiment to convince the reader that this well known problem has been avoided?
LjRR_Studio1
Pencil
LjRR_Studio1
Rectangle
LjRR_Studio1
Sticky Note
Why 'saturation'? You mean 'sorption'?
LjRR_Studio1
Rectangle
LjRR_Studio1
Pencil
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
Why so many? Has anyone reported this to be a problem?
LjRR_Studio1
Sticky Note
But this is for chemical bonds? How is (physical? weak chemical?) adsorption related to the formation of covalent bonds???
LjRR_Studio1
Sticky Note
But all this is for the polymer only... Have you shown uptake results for the polymer WITHOUT the adsorbent?
LjRR_Studio1
Highlight
LjRR_Studio1
Sticky Note
I am confused here... How is this related to diffusion? Which transition state? Between nonadsorbed and adsorbed? Or really from "one sorption site to the other"? Doesn't the adsorbate first have to desorb from one site in order to adsorb on the other? And how is this related to diffusion??? Again, not discussing critical aspects of this study, while wasting time on providing extraneous information.
LjRR_Studio1
Sticky Note