Design and Synthesis of Nitrogen-Doped Porous Carbon Materials for CO2 Capture and Investigation of CO2 Sorption Kinetics  

Investigators Jennifer Wilcox, Assistant Professor, Energy Resources Engineering; T. Daniel P. Stack, Associate Professor, Chemistry; Zhenan Bao, Associate Professor, Chemical Engineering; Jiajun He, Graduate Researcher, Energy Resources Engineering; Reza Haghpanah, Post-Doctoral Researcher, Energy Resources Engineering; John To, Graduate Researcher, Chemical Engineering; J. Brannon Gary, Post-Doctoral Researcher, Chemistry; Chris Lyons, Graduate Researcher, Chemistry, Stanford University.  

Abstract Solid-state post-combustion CO2 sorbents have certain advantages over traditional aqueous amine systems, including reduced regeneration energy since vaporization of liquid water is avoided, tunable pore morphology, and greater chemical variability. This report includes two parts, i.e., the design and synthesis of nitrogen-doped porous carbons as CO2 sorbents and the development of a robust technique for the measurement of CO2 sorption kinetics.

Mesoporous carbons are promising for CO2 capture due to their chemical inertness, low cost, high surface area, and tunable pore structures. The porous structure and high surface area allow addition of chemical functionality by grafting or impregnation. Nitrogen functionalization plays an important role in surface chemistry to achieve high CO2 adsorption capacity. We report here an ordered mesoporous nitrogen-doped carbon made using the co-assembly of modified-pyrrole and triblock copolymer through a soft-templating method, which is facile, economical, and fast compared to the hard-template approach. In the synthesis, the pyrrole precursor acts as both the carbon and nitrogen source. Carbonization of the resultant polymeric assembly generates a graphitic carbon structure and porous network through the removal of the block copolymer template. A high surface area mesoporous carbon was achieved that is comparable to silica counterparts. The resulting material shows promising CO2 capture performance, reaching equilibrium adsorption of 1.0 mmol CO2 g-1 of material at 25 °C and a pCO2 of 0.1 bar. Furthermore, the hierarchical macro-meso-microporous structure of this N-doped carbon allows for fast diffusion of CO2 gas into the adsorption sites and controlled condensation within the microporous structure. Another potential benefit is that the thermal conductivity of the mesoporous carbon is higher than its silica counterpart, resulting in faster regeneration with enhanced stability using a temperature swing adsorption process.

In addition, we also report a class of hierarchically microporous graphitic carbons with 3D interconnected pore architecture. Our strategy is based on an intrinsic 3D hierarchical nanostructured polymer hydrogel without any sacrificing templates. This sorbent showed promising CO2 capacity compared to commercially available activated carbon and possesses potential for application in dynamic column adsorption processes. Furthermore, the precursor that is used in the manufacturing process is cost effective.

Mesoporous silicas have been used as CO2 sorbent supports because of their well defined pore structures, high surface areas, and established synthetic and covalent modification procedures. By covalent integration of amines into the silica pore network, we aim to probe the relationship between structural properties of the active sorbents, adsorption and desorption conditions, and dynamics of CO2 adsorption/desorption cycling using thermogravimetric analysis (TGA). A robust technique for measuring the kinetics of CO2 adsorption and desorption under a variety of conditions has been developed and calibrated using a commercially available sorbent. Application to novel materials is currently underway.

Introduction The spontaneous reaction of primary amines with CO2 to rapidly form carbamate-ammonium ion pairs in solution makes aqueous amine scrubbing the preferred CO2 capture strategy at scale (Scheme 1). Wide-spread implementation of this technology at large point sources, such as coal-fired power plants, is precluded by the high energy demands of sorbent regeneration, chiefly due to vaporization of the water solvent. Translating amine-based sorbents to solid supports is a popular strategy for mitigating this cost, and extensive research has been geared toward synthesizing materials to maximize CO2 uptake capacity. The importance of CO2 adsorption and desorption kinetics has been considered far less.

Scheme 1. Carbamate-ammonium ion pair formation

Background Nitrogen-Doped Porous Carbon Materials Ordered N-doped mesoporous carbons have attracted considerable attention in the application of CO2 capture owing to their high surface area, tunable pore structure, narrow pore size distribution, and mechanical, thermal, and electric properties. N-doped mesoporous carbons are often prepared by a nanocasting method using a sacrificial template, typically porous silica. Nitrogen functionality can be incorporated by impregnation with nitrogen-containing organic molecules, followed by carbonization and removal of the silica template, or through post-synthesis treatment of mesoporous carbon using acetonitrile or ammonia chemical vapor deposition. These multi-step processes are costly and time consuming. Other methods include the co-assembly of a nitrogen-containing monomer, melamine resin, urea-phenol-formaldehyde resin, or dicyandiamide with a structural directing agent; however, the resulting porous polymers usually exhibit poor thermal stability. The decomposition process is further promoted by the high oxygen content within the triblock polymer. Development of a reliable and facile strategy to synthesize N-doped mesoporous carbon without the use of a hard template is highly desirable. Carbonization of polypyrrole has proven to yield highly graphitic and thermally conductive materials, which is desirable for heat transfer during adsorption and/or thermal regeneration. In this work, we report the successful synthesis of nitrogen-doped porous carbon using a modified-pyrrole monomer as both the carbon and nitrogen

source, which showed high CO2 capacity, high CO2/N2 selectivity, and facile regeneration.

CO2 Sorption Kinetics Comparison of amine-functionalized sorbent performance is difficult, as there is no adherence to a standard set of CO2 adsorption or desorption conditions in the literature, and those reported often do not faithfully mimic flue gas conditions (i.e., pCO2 0.1 – 0.15 bar; pH2O 0.1 – 0.2 bar; pO2 0.05 bar; pN2 0.6 bar; 40 – 60 oC). An equilibrium CO2 capacity of 3.3 mmol g-1 under humid conditions (pCO2 0.05 bar; pH2O 0.02 bar; 25 oC) was reported by Sayari et al. using a pore-expanded MCM-41 silica synthesized with grafted primary amines at 4.3 mmol g-1 loading.[1] While this equilibrium CO2 capacity is impressive for a relatively low pCO2, adsorption and desorption rates were not provided. The formation of bicarbonate is invoked during CO2 capture; however, this is not a viable reaction at scale as the uncatalyzed reaction is far too slow for adsorption/desorption cycles on the time scale of minutes.[2] The modified SBA-15 mesoporous silica by Jones et al. with covalently attached highly branched amines, created by in situ aziridine polymerization (nitrogen loading of 9.8 mmol g-1),[2] exhibits the highest reported equilibrium capacity of 5.6 mmol g-1 after a 3-hr exposure (pCO2 0.1 bar; pH2O 0.02 bar; 25 oC), which decreases to 2.8 mmol g-1 with a 0.1 hr exposure. Desorption kinetics were not measured; however, in a previous report a similar material required a 3-hr desorption cycle at 130 °C under N2 flow.[3]

Both thermal- and pressure-swing desorption processes are possible sorbent regeneration strategies. Thermal-swing desorption is generally performed under active flow of an inert purge gas such as N2 or Ar, resulting in counterproductive dilution of the captured CO2, which may be avoided using a pressure-swing approach. Steam-assisted desorption is also possible, but with subsequent costs of condensation or pressurization to remove liquid water.

Using thermogravimetric analysis (TGA), we have developed a method for measuring the CO2 adsorption and desorption capacities and rates of amine-functionalized silica materials under a variety of conditions. Correlating sorbent kinetic performance to the electronic and geometric structure of amines under realistic conditions will provide insights for further sorbent refinement to achieve cycle times on the time scale of minutes.

Results Nitrogen-Doped Porous Carbon We have synthesized a variety of carbon-based materials using polymer templates with pyrrole precursors possessing moderately high surface areas (800 m2 g-1) and interesting hierarchical porous structures (Figure 2). Structural characterization using XRD, SAXS, SEM, and TEM was performed. Elemental analysis and XPS showed considerable nitrogen functionality incorporated into the carbon materials. CO2 sorption capacities of 1.0 mmol g-1 and 3.1 mmol g-1 were measured at 25 oC and a pCO2 of 0.1 and 1.0 bar, respectively (Figure 3).

Figure 2: Left) Scanning electron microscopy (SEM) and Right) transmission electron microscopy (TEM) images showing the sponge like macroscopic structure of SNC-1 and the ordered mesopores feature.

Figure 3: CO2 isotherms of the SNC-1 material at 273, 298 and 323 K and N2 isotherm at 298 K. CO2 Sorption Kinetics TGA is ideally suited for monitoring adsorption/desorption kinetics, as mass change is monitored with high accuracy in time. To calibrate our TGA methods, we have investigated a commercially-available mesoporous silica sorbent, functionalized with a monolayer coverage of ethylenediamine units (ED, loading of 1.6 mmol g-1, Scheme 2).

Scheme 2. Sili-ED CO2 adsorption. Bare silica specs: surface area = 480 m2 g-1, pore diameter = 6 nm, pore volume = 0.7 mL g-1.

Figure 4 shows the CO2 adsorption of Sili-ED at 40 oC and a pCO2 of 0.15 bar (N2 balance). A final equilibrium CO2 capacity, q, of 0.83 mmol g-1 is obtained, with ca. 90 % of the final capacity achieved after 8 minutes. A CO2:amine ratio of 1:1 is expected for intramolecular carbamate-ammonium ion pair formation with ED (i.e. CO2 capacity of 1.6 mmol g-1). The relatively low CO2 uptake could be the result of a densely packed monolayer and inaccessible ED on the surface (Scheme 2).

Figure 4. TGA profile for Sili-ED CO2 adsorption (40 oC, pCO2 0.15 bar, pN2 0.85 bar, 20 mL min-1)

Figure 5 Left shows the CO2 adsorption/desorption profile for Sili-ED over 4 cycles. Adsorption at 40 oC with pCO2 0.15 bar is limited to 12 minutes, and ca. 95 % of the equilibrium CO2 capacity is obtained. Full regeneration is achieved by heating from 40 – 80 oC at 5 oC min-1 under N2 flow and holding isothermal for 12 minutes.

Figure 5 Right shows the time derivative of the TGA trace and provides a clearer assessment of rates of CO2 adsorption/desorption. Adsorption features are sharp and reflect fast CO2 uptake, as expected for primary amines. Desorption features are significantly broader and desorption is complete only after 20 minutes, highlighting a limiting characteristic of the material. Furthermore, desorption under N2 is not practical at scale as the resultant gas stream has a low CO2 concentration.

Figure 5. Left) TGA profile for Sili-ED CO2 adsorption/desorption cycles. Adsorption: 40 oC, pCO2 0.15 bar, pN2 0.85 bar, 20 mL min-1. Desorption: 40 o– 80 oC, 5 oC min-1, pCO2 0.0 bar, pN2 1.0 bar, 20 mL min-1. Right) Time derivative of CO2 adsorption/desorption trace.

Progress Promising nitrogen-doped mesoporous carbons with unique hierarchical macro-, meso-, and microporous architectures were synthesized using a soft template approach and characterized using various analytical methods. Nitrogen functionalities within the carbon matrix enhance the CO2 sorption capacity and selectivity relative to other activated carbons.

In addition, a robust and reproducible method for monitoring CO2 adsorption/desorption kinetics has been developed using TGA. The commercially available sorbent Sili-ED displays rapid CO2 uptake kinetics and a good CO2 capacity after 12 minutes. Kinetics for complete CO2 desorption are significantly slower at ca. 20 minutes. The method can be applied to a variety of solid sorbents to simultaneously measure CO2 capacity and sorption kinetics.

Future Plans Efforts to increase the nitrogen content and control nitrogen speciation in the mesoporous carbon materials to enhance CO2 capacity and selectivity are currently underway. Additional nitrogen functional groups can also be incorporated into the framework through the addition of nitrogen-containing additives or post-synthesis impregnation or grafting.

In addition, physical or chemical impregnation of carbonic anhydrase into the mesoporous carbon substrate is under investigation. This material may provide a means to catalytically convert CO2 into bicarbonate in a robust and efficient manner. Furthermore, the CO2 capture performance of several commercial solid sorbents as well as synthetic materials under realistic coal-fired post-combustion flue gas conditions is under investigation. The impacts of related trace acidic species (SOx and NOx) on the CO2 capture performance will be studied.

Investigation of novel silica and carbon materials using TGA is ongoing. Development of efficient desorption methods using vacuum and/or steam is also planned.

Patents 1. Wilcox, J., D.T. Stack, Z. Bao, J. He, J. Tao, G. Brannon, Biomimetic Sorbents for

CO2 Capture, Provisional Patent S12-500, 2013. 2. Bao, Z., J. Wilcox, J. To, J. He, Nitrogen-Doped Micro and Mesoporous Carbon

Sorbents for CO2 Capture, Provisional Patent S13-374, 2013.

References 1. Serna-Guerrero, R., E. Da'na, and A. Sayari, New Insights into the Interactions of

CO2 with Amine-Functionalized Silica. Industrial & Engineering Chemistry Research, 2008. 47(23): p. 9406-9412.

2. Drese, J.H., et al., Synthesis-Structure-Property Relationships for Hyperbranched Aminosilica CO2 Adsorbents. Advanced Functional Materials, 2009. 19(23): p. 3821-3832.

3. Hicks, J.C., et al., Designing adsorbents for CO2 capture from flue gas-hyperbranched aminosilicas capable,of capturing CO2 reversibly. Journal of the American Chemical Society, 2008. 130(10): p. 2902-2903.

Contacts Jennifer Wilcox: wilcoxj@stanford.edu T. Daniel P. Stack: stack@stanford.edu Zhenan Bao: zbao@stanford.edu Jiajun He: jiajunhe@stanford.edu Reza Haghpanah: rezahagh@stanford.edu John To: johnto@stanford.edu J. Brannon Gary: jbgary@stanford.edu

Chris Lyons: ctlyons@stanford.edu