Supporting Information for: Capture of Carbon Dioxide from Air and ...
Supporting Information for:
Capture of Carbon Dioxide from Air and Flue Gas in the Alkylamine-Appended Metal-Organic Framework mmen-Mg2(dobpdc)
Thomas M. McDonald, Woo Ram Lee, Jarad A. Mason, Brian M. Wiers, Chang Seop Hong,* and Jeffrey R. Long*
Department of Chemistry, University of California, Berkeley, California, 94720, U. S. A. Department of Chemistry, Korea University, Seoul, 136-713, Republic of Korea
E-mail: [email protected] and [email protected]
J. Am. Chem. Soc.
DOI:
Powder X-ray Diffraction
Figure S1. Experimental X-ray powder diffraction patterns of DEF-Zn2(dobpdc) (red) and DEF-Mg2(dobpdc) (blue) with the calculated diffraction pattern of DEFZn2(dobpdc) (black)
S2
Figure S2. Experimental X-ray powder diffraction pattern of DEF-Mg2(dobpdc) (blue) with calculated diffraction pattern (red) from Le Bail refinement with difference (green).
Figure S3. Experimental X-ray powder diffraction pattern of mmen-Mg2(dobpdc) (blue) with calculated diffraction pattern (red) from Le Bail refinement with difference (green).
S3
Nitrogen Adsorption Isotherms at 77 K
Figure S4. Experimental N2 adsorption isotherm of activated Mg2(dobpdc) at 77 K.
Figure S5. Experimental N2 adsorption isotherm in mmen-Mg2(dobpdc) at 77 K.
S4
BET and Pore Size Distribution Calculations Table S1. BET parameters from for the N2 adsorption isotherms of Mg2(dobpdc) and mmen-Mg2(dobpdc). Slope Y-int C Vm (cm3/g) R2 P/P0 Low P/P0 High
2 mmen-2 0.001327173 0.064920639 4.4504E-06 0.0005808414 299.21 112.77 750.9631854 15.26683052 0.9993160863 0.9997889612 0.044774384 0.036439604 0.071232503 0.050146781
Figure S6. DFT pore size distribution for Mg2(dobpdc) calculated from N2 adsorption at 77 K for a metal oxide surface with a cylinder pore geometry.
S5
Figure S7. DFT pore size distribution for mmen-Mg2(dobpdc) calculated from N2 adsorption at 77 K for a metal oxide surface with a cylinder pore geometry.
S6
CO2 Isotherm Fits and Isosteric Heats of Adsorption Table S2. Dual-site Langmuir-Freundlich parameters (Eqn. 1) for the pre-step region of the CO2 adsorption isotherms of mmen-Mg2(dobpdc) at 25, 50, and 75 °C.
qsat,A (mmol/g) bA (bar-α) αA qsat,B (mmol/g) bB (bar-α) αB
25 °C 0.2 20.3 1.4 3.0 2.5 x 1016 21.1
50 °C 0.46 0.60 0.93 0.03 8.7 x 10-3 18.5
75 °C 0.15 1.64 x 10-12 10.8 0.88 8.2 x 10-2 0.59
Table S3. Modified dual-site Langmuir-Freundlich it parameters (Eqn. 2) for the poststep region of the CO2 adsorption isotherms of mmen-Mg2(dobpdc) at 25, 50, and 75 °C.
qsat,A (mmol/g) bA (bar-α) αA qsat,B (mmol/g) bB (bar-α) αB qsat,C (mmol/g) bC (bar-α) αC pstep
25 °C 2.67 11.7 0.92 0.51 1.9 x 10-2 1.1 1.54 1.16 x 10-5 1.6 0.16
50 °C 2.5 1.32 1.23 0.26 2.5 x 10-3 1.9 2.23 1.1 x 10-3 0.86 1.31
75 °C 2.72 0.19 1 1.76 4.2 x 10-4 1 0 − − 0.59
Table S4. Dual-site Langmuir-Freundlich parameters (Eqn. 1) for the CO2 adsorption isotherms of Mg2(dobpdc) at 25, 35 and 45 °C.
qsat,A (mmol/g) bA (bar-1) αA qsat,B (mmol/g) bB (bar-1) αB
25 °C 5.52 29.6 0.89 11.4 0.105 1
35 °C 5.43 19.2 0.90 6.13 0.180 1
S7
45 °C 5.31 13.0 0.92 4.20 0.247 1
Figure S8. Experimental data and corresponding dual-site Langmuir-Freundlich (DSLF) isotherm fit for CO2 adsorption in Mg2(dobpdc) at 25, 35, and 45 °C.
Figure S9. Isosteric heats of adsorption (−Qst) as a function of loading for CO2 in Mg2(dobpdc).
S8
Figure S10. Experimental data and corresponding isotherm fits for CO2 adsorption in mmen-Mg2(dobpdc) at 25, 50, and 75 °C.
Figure S11. Experimental data and corresponding isotherm fit for CO2 adsorption in mmen-Mg2(dobpdc) at 25 °C plotted on a logarithmic scale.
S9
Figure S12. Experimental data and corresponding isotherm fit for CO2 adsorption in mmen-Mg2(dobpdc) at 50 °C plotted on a logarithmic scale.
Figure S13. Experimental data and corresponding isotherm fit for CO2 adsorption in mmen-Mg2(dobpdc) at 75 °C plotted on a logarithmic scale.
S10
Figure S14. Residual sum of squares (R2) for the best fit line of ln p versus 1/T as a function of the constant CO2 loading used in the Clausius-Clapeyron equation.
S11
Additional Adsorption and Desorption Isotherms in mmen-Mg2(dobpdc)
Figure S15. Experimental isotherms for CO2 adsorption (empty shapes) and desorption (filled shapes) in mmen-Mg2(dobpdc) at 25 (blue squares), 50 (green triangles), and 75 (red circles) °C.
Figure S16. Four experimental CO2 adsorption isotherms at 75 °C for the batch of mmen-Mg2(dobpdc) described in this work.
S12
Figure S17. Experimental CO2 adsorption isotherm at 75 °C for a second batch of mmen-Mg2(dobpdc) with calculated composition Mg2(dobpdc)(mmen)1.75(H2O)0.25.
Figure S18. Experimental adsorption isotherms for O2 (blue circles) and N2 (red triangles) in mmen-Mg2(dobpdc) at 25 °C.
S13
Table S5. Calculated pressures for model CO2 isotherms from Eqn. 5 for mmen-2. Pressures@348 were measured; other pressures were calculated for equal loadings (q). P@348 (mbar) 0.05742 0.09963 0.21091 0.31617 0.43002 0.48455 0.63701 0.84691 1.00789 2.09581 2.97815 3.83228 5.28017 5.86216 7.8705 9.49118 10.10264 11.19873 11.93207 12.56071 13.6957 14.72252 19.15949 24.66098 33.79363 52.06572 82.35123 100.14469 126.57272 151.60942 176.93202 202.44541 299.69861 402.9908 499.98593 600.63751 700.74869 800.56749 900.77761 1000.21127 1100.60392
P@298 (mbar) 9.35E-04 0.00162 0.00344 0.00515 0.007 0.00789 0.01037 0.01379 0.01642 0.03413 0.0485 0.06242 0.086 0.09548 0.12819 0.15458 0.16454 0.18239 0.19434 0.20457 0.22306 0.23978 0.31205 0.40165 0.55039 0.84799 1.34124 1.63104 2.06147 2.46924 2.88166 3.2972 4.88114 6.56345 8.14319 9.78249 11.41298 13.03872 14.67082 16.29028 17.92536
P@323 (mbar) 0.00859 0.01491 0.03157 0.04732 0.06436 0.07252 0.09534 0.12675 0.15085 0.31367 0.44572 0.57356 0.79026 0.87736 1.17794 1.4205 1.51201 1.67606 1.78581 1.8799 2.04976 2.20344 2.8675 3.69088 5.05772 7.79241 12.32508 14.98814 18.94349 22.6906 26.4805 30.29896 44.85435 60.31355 74.83032 89.89432 104.87744 119.81681 134.81474 149.69646 164.72171
P@373 (mbar) 0.29742 0.51606 1.09246 1.63767 2.22738 2.50983 3.29953 4.38676 5.22059 10.85571 15.42599 19.85015 27.34981 30.36436 40.767 49.16167 52.32887 58.00631 61.80481 65.06099 70.93992 76.25856 99.24083 127.73702 175.04161 269.68596 426.55648 518.72165 655.6115 785.29465 916.45868 1048.61094 1552.35549 2087.38031 2589.78812 3111.13533 3629.68341 4146.71704 4665.77761 5180.81632 5700.82234
S14
P@393 (mbar) 0.95363 1.65466 3.50281 5.25097 7.1418 8.04744 10.5795 14.06553 16.7391 34.80734 49.4613 63.64674 87.69339 97.35912 130.71375 157.63011 167.78527 185.9892 198.16856 208.60905 227.45904 244.51253 318.202 409.57109 561.24671 864.71071 1367.69435 1663.20923 2102.1276 2517.93867 2938.49798 3362.22595 4977.41314 6692.89625 8303.79739 9975.42509 11638.07779 13295.87462 14960.17052 16611.57092 18278.89829
q (mmol/g) 0.00448 0.01553 0.0297 0.03758 0.04422 0.04726 0.05466 0.06252 0.06818 0.09686 0.11425 0.13473 0.15989 0.16931 0.1901 0.21732 0.22845 0.26239 0.47472 0.65018 1.01337 1.22148 1.71096 2.01439 2.23896 2.4366 2.58102 2.63613 2.68712 2.73043 2.76682 2.79527 2.87445 2.9366 2.99042 3.04084 3.08927 3.13637 3.1783 3.22269 3.26559
Temperature Swing Desorption
Figure S19. Model CO2 isotherms illustrating temperature swing adsorption; blue line represents a low temperature isotherm and the red line a high temperature isotherm. The working capacity is the difference in capacity between Point C, the amount adsorbed at low pressure (0.15 bar) and low temperature (40 °C), and Point A, the amount adsorbed at high pressure (1 bar) and high temperature (150 °C). Points A and C in Figure 7 correspond to the Points A and C in the isotherms above.
S15
Fourier Transform Infrared (FTIR) Spectroscopy
Figure S20. FTIR spectrum of DEF-Zn2(dobpdc) in KBr.
Figure S21. FTIR spectrum of DEF-Mg2(dobpdc) (blue, bottom) in KBr and neat mmen-Mg2(dobpdc) (green, top).
S16
Thermogravimetric Analysis (TGA)
Figure S22. Thermogravimetric analysis trace of DEF-Zn2(dobpdc) confirms solvent content. Based upon the calculated composition, 16% weight loss is expected for DEF and H2O encapsulated within the pores (25-150 °C), and 28% weight loss for bound DEF (200-375°C).
Figure S23. Thermogravimetric analysis trace of DEF-Mg2(dobpdc) confirms solvent content. Based upon the calculated composition, 25% weight loss is expected for H2O encapsulated within the pores (25-200 °C), and 29% weight loss for bound DEF (250400°C).
S17
Figure S24. Thermogravimetric analysis trace of mmen-Mg2(dobpdc) confirms solvent content. Based upon the calculated composition of the fully evacuated framework, 32% weight loss is expected for bound mmen and H2O molecules (250-400°C). Atmospheric water and CO2, adsorbed onto the pore surfaces when the evacuated framework was exposed to air, desorb between 0-100 °C; these guest species differentiate this sample from the air-free sample studied via elemental analysis.
S18
Estimation of Void Space Effects In addition to the experimentally measured excess adsorption, real solid adsorbents contain potentially significant amounts of gas both in the spaces between crystallites and in the “bulk” of the pore space. When the adsorbent bed is regenerated, these gases will contaminate the product stream since the gas phase N2 concentration is usually significantly larger than the gas phase CO2 concentration in most CO2 capture applications. Thus to approximate the effect of these voids, a simple model which accounts for both absolute adsorption and bed void space is described. The magnitude of the void space in a real bed will depend primarily on how closely crystallites within the bed are packed together. Table S6 with approximate selectivity and purity values for different fractions of void space follows the example calculation below. Assume an adsorbent bed is packed with 2 g of mmen-2 such that the intercrystalline void space is 1/3 (33%). Based upon the calculated crystal density, 1 g of mmen-2 occupies 1.16 cm3. As a result, the total volume of this bed is 3.48 cm3 (2.32 cm3 from mmen-2 and 1.16 cm3 from the void space). For the purpose of the following calculations, a pore volume of 0.85 cm3/g is assumed. For mmen-2, DFT pore size calculations from the N2 adsorption isotherm at 77 K indicated that the pore volume was approximately 0.03 cm3/g. We do not believe this to be a reasonable value for pore size at 298 K, because the amine appended framework adsorbed significantly more CO2 at 298 K than N2 at 77 K. However, the non-functionalized framework 2 adsorbed N2 normally at 77 K, with a pore volume of 1.25 cm3/g calculated via DFT. For mmen-2, the gravimetric pore volume can only be smaller than the pore volume calculated for 2 for two reasons: the amines occupy pore space and the amines makes the framework significantly denser. Because we cannot accurately estimate how much pore space is occupied by the amines, we will only correct for the increased mass of the framework. Thus, assuming the amines occupy no pore space but increase framework mass by 48%, the maximum pore volume of mmen-2 is 0.85 cm3/g. The total void volume for this 2 g adsorbent bed is thus 2.86 cm3 (1.43 cm3/g): 2 * 0.85 cm3 + 1.16 cm3 = 2.86 cm3 Assuming ideal gas behavior, the total void space could be filled with a maximum of 0.058 mmol/g of gas at 1 bar of total pressure: (1.43 cm3/g) / (24.8 cm3/mmol) = 0.058 mmol/g Assuming the void space is filled with 100% N2, we add this value to the gravimetric excess N2 adsorption capacity at 0.75 bar to obtain a maximum total N2 adsorption of 0.137 mmol/g:
S19
0.079 mmol/g + 0.058 mmol/g = 0.137 mmol/g These quantities would result in a selectivity of 114 and a purity of 95.8% for mmen-2 for flue gas capture. Table S6. Selectivity and purity values calculated for various intercrystalline void fractions are shown for the partial pressures relevant to flue gas capture. Void Fraction 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Selectivity 132 125 117 108 98 85 70 52 29
S20
Purity 96.4 96.2 95.9 95.6 95.1 94.5 93.4 91.2 85.4
Approximate Regeneration Energy Calculations From the heat capacity of mmen-2 and the calculated isosteric heat of CO2 adsorption onto an amine, it is possible to compute approximate adsorbent regeneration energies. To verify the methodology, the regeneration energy of mmen-2 for conditions identical to those utilized during the DSC experiment were calculated. The sensible heat required for regeneration (104 J/g) is the specific heat (1.3 J/g•K) multiplied by the temperature change (80 K). Each gram of mmen-2 adsorbed 2.52 mmol of CO2 with a heat of approximately -71 J/mmol. To remove this CO2, approximately 179 J/g are required. To adsorb 1 kg (22.72 mmol) of CO2, 9.02 g of mmen-2 are necessary. Thus, the regeneration energy is 9.02 g * (179 + 104 J/g) = 2.55 kJ/kg CO2 or 2.55 MJ/tonne CO2 This value is close to the measured value of 2.34 MJ/tonne CO2. Desorption of mmen-2 with 100% CO2 reduced the working capacity of the adsorbent and necessitated higher regeneration temperatures, increasing the regeneration energy requirement. The approximate regeneration energy for mmen-2 using CO2 is 3.25 MJ/tonne CO2. Sensible heat: 1.3 J/g•K * 110 K = 130 J/g Regeneration heat: 1.8 mmol/g * 71 J/mmol = 128 J/g 12.6 g * (130 +128 J/g) = 3.25 kJ/kg CO2 Lastly, the regeneration energy for air capture was calculated to be 5.14 MJ/tonne CO2. Sensible heat: 1.3 J/g•K * 125 K = 163 J/g Regeneration heat: 1.05 mmol/g * 71 J/mmol = 75 J/g 21.6 g * (163 +75 J/g) = 5.14 kJ/kg CO2
S21
Capture of Carbon Dioxide from Air and Flue Gas in the Alkylamine-Appended Metal-Organic Framework mmen-Mg2(dobpdc)
Thomas M. McDonald, Woo Ram Lee, Jarad A. Mason, Brian M. Wiers, Chang Seop Hong,* and Jeffrey R. Long*
Department of Chemistry, University of California, Berkeley, California, 94720, U. S. A. Department of Chemistry, Korea University, Seoul, 136-713, Republic of Korea
E-mail: [email protected] and [email protected]
J. Am. Chem. Soc.
DOI:
Powder X-ray Diffraction
Figure S1. Experimental X-ray powder diffraction patterns of DEF-Zn2(dobpdc) (red) and DEF-Mg2(dobpdc) (blue) with the calculated diffraction pattern of DEFZn2(dobpdc) (black)
S2
Figure S2. Experimental X-ray powder diffraction pattern of DEF-Mg2(dobpdc) (blue) with calculated diffraction pattern (red) from Le Bail refinement with difference (green).
Figure S3. Experimental X-ray powder diffraction pattern of mmen-Mg2(dobpdc) (blue) with calculated diffraction pattern (red) from Le Bail refinement with difference (green).
S3
Nitrogen Adsorption Isotherms at 77 K
Figure S4. Experimental N2 adsorption isotherm of activated Mg2(dobpdc) at 77 K.
Figure S5. Experimental N2 adsorption isotherm in mmen-Mg2(dobpdc) at 77 K.
S4
BET and Pore Size Distribution Calculations Table S1. BET parameters from for the N2 adsorption isotherms of Mg2(dobpdc) and mmen-Mg2(dobpdc). Slope Y-int C Vm (cm3/g) R2 P/P0 Low P/P0 High
2 mmen-2 0.001327173 0.064920639 4.4504E-06 0.0005808414 299.21 112.77 750.9631854 15.26683052 0.9993160863 0.9997889612 0.044774384 0.036439604 0.071232503 0.050146781
Figure S6. DFT pore size distribution for Mg2(dobpdc) calculated from N2 adsorption at 77 K for a metal oxide surface with a cylinder pore geometry.
S5
Figure S7. DFT pore size distribution for mmen-Mg2(dobpdc) calculated from N2 adsorption at 77 K for a metal oxide surface with a cylinder pore geometry.
S6
CO2 Isotherm Fits and Isosteric Heats of Adsorption Table S2. Dual-site Langmuir-Freundlich parameters (Eqn. 1) for the pre-step region of the CO2 adsorption isotherms of mmen-Mg2(dobpdc) at 25, 50, and 75 °C.
qsat,A (mmol/g) bA (bar-α) αA qsat,B (mmol/g) bB (bar-α) αB
25 °C 0.2 20.3 1.4 3.0 2.5 x 1016 21.1
50 °C 0.46 0.60 0.93 0.03 8.7 x 10-3 18.5
75 °C 0.15 1.64 x 10-12 10.8 0.88 8.2 x 10-2 0.59
Table S3. Modified dual-site Langmuir-Freundlich it parameters (Eqn. 2) for the poststep region of the CO2 adsorption isotherms of mmen-Mg2(dobpdc) at 25, 50, and 75 °C.
qsat,A (mmol/g) bA (bar-α) αA qsat,B (mmol/g) bB (bar-α) αB qsat,C (mmol/g) bC (bar-α) αC pstep
25 °C 2.67 11.7 0.92 0.51 1.9 x 10-2 1.1 1.54 1.16 x 10-5 1.6 0.16
50 °C 2.5 1.32 1.23 0.26 2.5 x 10-3 1.9 2.23 1.1 x 10-3 0.86 1.31
75 °C 2.72 0.19 1 1.76 4.2 x 10-4 1 0 − − 0.59
Table S4. Dual-site Langmuir-Freundlich parameters (Eqn. 1) for the CO2 adsorption isotherms of Mg2(dobpdc) at 25, 35 and 45 °C.
qsat,A (mmol/g) bA (bar-1) αA qsat,B (mmol/g) bB (bar-1) αB
25 °C 5.52 29.6 0.89 11.4 0.105 1
35 °C 5.43 19.2 0.90 6.13 0.180 1
S7
45 °C 5.31 13.0 0.92 4.20 0.247 1
Figure S8. Experimental data and corresponding dual-site Langmuir-Freundlich (DSLF) isotherm fit for CO2 adsorption in Mg2(dobpdc) at 25, 35, and 45 °C.
Figure S9. Isosteric heats of adsorption (−Qst) as a function of loading for CO2 in Mg2(dobpdc).
S8
Figure S10. Experimental data and corresponding isotherm fits for CO2 adsorption in mmen-Mg2(dobpdc) at 25, 50, and 75 °C.
Figure S11. Experimental data and corresponding isotherm fit for CO2 adsorption in mmen-Mg2(dobpdc) at 25 °C plotted on a logarithmic scale.
S9
Figure S12. Experimental data and corresponding isotherm fit for CO2 adsorption in mmen-Mg2(dobpdc) at 50 °C plotted on a logarithmic scale.
Figure S13. Experimental data and corresponding isotherm fit for CO2 adsorption in mmen-Mg2(dobpdc) at 75 °C plotted on a logarithmic scale.
S10
Figure S14. Residual sum of squares (R2) for the best fit line of ln p versus 1/T as a function of the constant CO2 loading used in the Clausius-Clapeyron equation.
S11
Additional Adsorption and Desorption Isotherms in mmen-Mg2(dobpdc)
Figure S15. Experimental isotherms for CO2 adsorption (empty shapes) and desorption (filled shapes) in mmen-Mg2(dobpdc) at 25 (blue squares), 50 (green triangles), and 75 (red circles) °C.
Figure S16. Four experimental CO2 adsorption isotherms at 75 °C for the batch of mmen-Mg2(dobpdc) described in this work.
S12
Figure S17. Experimental CO2 adsorption isotherm at 75 °C for a second batch of mmen-Mg2(dobpdc) with calculated composition Mg2(dobpdc)(mmen)1.75(H2O)0.25.
Figure S18. Experimental adsorption isotherms for O2 (blue circles) and N2 (red triangles) in mmen-Mg2(dobpdc) at 25 °C.
S13
Table S5. Calculated pressures for model CO2 isotherms from Eqn. 5 for mmen-2. Pressures@348 were measured; other pressures were calculated for equal loadings (q). P@348 (mbar) 0.05742 0.09963 0.21091 0.31617 0.43002 0.48455 0.63701 0.84691 1.00789 2.09581 2.97815 3.83228 5.28017 5.86216 7.8705 9.49118 10.10264 11.19873 11.93207 12.56071 13.6957 14.72252 19.15949 24.66098 33.79363 52.06572 82.35123 100.14469 126.57272 151.60942 176.93202 202.44541 299.69861 402.9908 499.98593 600.63751 700.74869 800.56749 900.77761 1000.21127 1100.60392
P@298 (mbar) 9.35E-04 0.00162 0.00344 0.00515 0.007 0.00789 0.01037 0.01379 0.01642 0.03413 0.0485 0.06242 0.086 0.09548 0.12819 0.15458 0.16454 0.18239 0.19434 0.20457 0.22306 0.23978 0.31205 0.40165 0.55039 0.84799 1.34124 1.63104 2.06147 2.46924 2.88166 3.2972 4.88114 6.56345 8.14319 9.78249 11.41298 13.03872 14.67082 16.29028 17.92536
P@323 (mbar) 0.00859 0.01491 0.03157 0.04732 0.06436 0.07252 0.09534 0.12675 0.15085 0.31367 0.44572 0.57356 0.79026 0.87736 1.17794 1.4205 1.51201 1.67606 1.78581 1.8799 2.04976 2.20344 2.8675 3.69088 5.05772 7.79241 12.32508 14.98814 18.94349 22.6906 26.4805 30.29896 44.85435 60.31355 74.83032 89.89432 104.87744 119.81681 134.81474 149.69646 164.72171
P@373 (mbar) 0.29742 0.51606 1.09246 1.63767 2.22738 2.50983 3.29953 4.38676 5.22059 10.85571 15.42599 19.85015 27.34981 30.36436 40.767 49.16167 52.32887 58.00631 61.80481 65.06099 70.93992 76.25856 99.24083 127.73702 175.04161 269.68596 426.55648 518.72165 655.6115 785.29465 916.45868 1048.61094 1552.35549 2087.38031 2589.78812 3111.13533 3629.68341 4146.71704 4665.77761 5180.81632 5700.82234
S14
P@393 (mbar) 0.95363 1.65466 3.50281 5.25097 7.1418 8.04744 10.5795 14.06553 16.7391 34.80734 49.4613 63.64674 87.69339 97.35912 130.71375 157.63011 167.78527 185.9892 198.16856 208.60905 227.45904 244.51253 318.202 409.57109 561.24671 864.71071 1367.69435 1663.20923 2102.1276 2517.93867 2938.49798 3362.22595 4977.41314 6692.89625 8303.79739 9975.42509 11638.07779 13295.87462 14960.17052 16611.57092 18278.89829
q (mmol/g) 0.00448 0.01553 0.0297 0.03758 0.04422 0.04726 0.05466 0.06252 0.06818 0.09686 0.11425 0.13473 0.15989 0.16931 0.1901 0.21732 0.22845 0.26239 0.47472 0.65018 1.01337 1.22148 1.71096 2.01439 2.23896 2.4366 2.58102 2.63613 2.68712 2.73043 2.76682 2.79527 2.87445 2.9366 2.99042 3.04084 3.08927 3.13637 3.1783 3.22269 3.26559
Temperature Swing Desorption
Figure S19. Model CO2 isotherms illustrating temperature swing adsorption; blue line represents a low temperature isotherm and the red line a high temperature isotherm. The working capacity is the difference in capacity between Point C, the amount adsorbed at low pressure (0.15 bar) and low temperature (40 °C), and Point A, the amount adsorbed at high pressure (1 bar) and high temperature (150 °C). Points A and C in Figure 7 correspond to the Points A and C in the isotherms above.
S15
Fourier Transform Infrared (FTIR) Spectroscopy
Figure S20. FTIR spectrum of DEF-Zn2(dobpdc) in KBr.
Figure S21. FTIR spectrum of DEF-Mg2(dobpdc) (blue, bottom) in KBr and neat mmen-Mg2(dobpdc) (green, top).
S16
Thermogravimetric Analysis (TGA)
Figure S22. Thermogravimetric analysis trace of DEF-Zn2(dobpdc) confirms solvent content. Based upon the calculated composition, 16% weight loss is expected for DEF and H2O encapsulated within the pores (25-150 °C), and 28% weight loss for bound DEF (200-375°C).
Figure S23. Thermogravimetric analysis trace of DEF-Mg2(dobpdc) confirms solvent content. Based upon the calculated composition, 25% weight loss is expected for H2O encapsulated within the pores (25-200 °C), and 29% weight loss for bound DEF (250400°C).
S17
Figure S24. Thermogravimetric analysis trace of mmen-Mg2(dobpdc) confirms solvent content. Based upon the calculated composition of the fully evacuated framework, 32% weight loss is expected for bound mmen and H2O molecules (250-400°C). Atmospheric water and CO2, adsorbed onto the pore surfaces when the evacuated framework was exposed to air, desorb between 0-100 °C; these guest species differentiate this sample from the air-free sample studied via elemental analysis.
S18
Estimation of Void Space Effects In addition to the experimentally measured excess adsorption, real solid adsorbents contain potentially significant amounts of gas both in the spaces between crystallites and in the “bulk” of the pore space. When the adsorbent bed is regenerated, these gases will contaminate the product stream since the gas phase N2 concentration is usually significantly larger than the gas phase CO2 concentration in most CO2 capture applications. Thus to approximate the effect of these voids, a simple model which accounts for both absolute adsorption and bed void space is described. The magnitude of the void space in a real bed will depend primarily on how closely crystallites within the bed are packed together. Table S6 with approximate selectivity and purity values for different fractions of void space follows the example calculation below. Assume an adsorbent bed is packed with 2 g of mmen-2 such that the intercrystalline void space is 1/3 (33%). Based upon the calculated crystal density, 1 g of mmen-2 occupies 1.16 cm3. As a result, the total volume of this bed is 3.48 cm3 (2.32 cm3 from mmen-2 and 1.16 cm3 from the void space). For the purpose of the following calculations, a pore volume of 0.85 cm3/g is assumed. For mmen-2, DFT pore size calculations from the N2 adsorption isotherm at 77 K indicated that the pore volume was approximately 0.03 cm3/g. We do not believe this to be a reasonable value for pore size at 298 K, because the amine appended framework adsorbed significantly more CO2 at 298 K than N2 at 77 K. However, the non-functionalized framework 2 adsorbed N2 normally at 77 K, with a pore volume of 1.25 cm3/g calculated via DFT. For mmen-2, the gravimetric pore volume can only be smaller than the pore volume calculated for 2 for two reasons: the amines occupy pore space and the amines makes the framework significantly denser. Because we cannot accurately estimate how much pore space is occupied by the amines, we will only correct for the increased mass of the framework. Thus, assuming the amines occupy no pore space but increase framework mass by 48%, the maximum pore volume of mmen-2 is 0.85 cm3/g. The total void volume for this 2 g adsorbent bed is thus 2.86 cm3 (1.43 cm3/g): 2 * 0.85 cm3 + 1.16 cm3 = 2.86 cm3 Assuming ideal gas behavior, the total void space could be filled with a maximum of 0.058 mmol/g of gas at 1 bar of total pressure: (1.43 cm3/g) / (24.8 cm3/mmol) = 0.058 mmol/g Assuming the void space is filled with 100% N2, we add this value to the gravimetric excess N2 adsorption capacity at 0.75 bar to obtain a maximum total N2 adsorption of 0.137 mmol/g:
S19
0.079 mmol/g + 0.058 mmol/g = 0.137 mmol/g These quantities would result in a selectivity of 114 and a purity of 95.8% for mmen-2 for flue gas capture. Table S6. Selectivity and purity values calculated for various intercrystalline void fractions are shown for the partial pressures relevant to flue gas capture. Void Fraction 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Selectivity 132 125 117 108 98 85 70 52 29
S20
Purity 96.4 96.2 95.9 95.6 95.1 94.5 93.4 91.2 85.4
Approximate Regeneration Energy Calculations From the heat capacity of mmen-2 and the calculated isosteric heat of CO2 adsorption onto an amine, it is possible to compute approximate adsorbent regeneration energies. To verify the methodology, the regeneration energy of mmen-2 for conditions identical to those utilized during the DSC experiment were calculated. The sensible heat required for regeneration (104 J/g) is the specific heat (1.3 J/g•K) multiplied by the temperature change (80 K). Each gram of mmen-2 adsorbed 2.52 mmol of CO2 with a heat of approximately -71 J/mmol. To remove this CO2, approximately 179 J/g are required. To adsorb 1 kg (22.72 mmol) of CO2, 9.02 g of mmen-2 are necessary. Thus, the regeneration energy is 9.02 g * (179 + 104 J/g) = 2.55 kJ/kg CO2 or 2.55 MJ/tonne CO2 This value is close to the measured value of 2.34 MJ/tonne CO2. Desorption of mmen-2 with 100% CO2 reduced the working capacity of the adsorbent and necessitated higher regeneration temperatures, increasing the regeneration energy requirement. The approximate regeneration energy for mmen-2 using CO2 is 3.25 MJ/tonne CO2. Sensible heat: 1.3 J/g•K * 110 K = 130 J/g Regeneration heat: 1.8 mmol/g * 71 J/mmol = 128 J/g 12.6 g * (130 +128 J/g) = 3.25 kJ/kg CO2 Lastly, the regeneration energy for air capture was calculated to be 5.14 MJ/tonne CO2. Sensible heat: 1.3 J/g•K * 125 K = 163 J/g Regeneration heat: 1.05 mmol/g * 71 J/mmol = 75 J/g 21.6 g * (163 +75 J/g) = 5.14 kJ/kg CO2
S21
