SC I ENCE ADVANCES | R E S EARCH ART I C L E

PHYS I CAL SC I ENCE

1Department of Chemical and Biological Engineering, Northwestern University, Evan-ston, IL 60208, USA. 2Department of Chemical and Biological Engineering, ColoradoSchool of Mines, Golden, CO 80401, USA. 3Department of Chemistry, NorthwesternUniversity, Evanston, IL 60208, USA. 4Department of Chemistry, Faculty of Science, KingAbdulaziz University, Jeddah 22254, Saudi Arabia.*These authors contributed equally to this work.†Present address: School of Chemical and Biomolecular Engineering, Pusan Na-tional University, Busan, Korea (South).‡Present address: Department of Chemistry and Biochemistry, Southern IllinoisUniversity, Carbondale, IL 62901, USA.§Corresponding author. Email: snurr@northwestern.edu

Chung et al., Sci. Adv. 2016;2 : e1600909 14 October 2016

2016 © The Authors,

some rights reserved;

exclusive licensee

American Association

for the Advancement

of Science. Distributed

under a Creative

Commons Attribution

NonCommercial

License 4.0 (CC BY-NC).

Do

In silico discovery of metal-organic frameworks forprecombustion CO2 capture using a genetic algorithmYongchul G. Chung,1*† Diego A. Gómez-Gualdrón,1,2* Peng Li,3 Karson T. Leperi,1 Pravas Deria,3‡

Hongda Zhang,1 Nicolaas A. Vermeulen,3 J. Fraser Stoddart,3 Fengqi You,1 Joseph T. Hupp,3

Omar K. Farha,3,4 Randall Q. Snurr1§

Discovery of new adsorbent materials with a high CO2 working capacity could help reduce CO2 emissions fromnewly commissioned power plants using precombustion carbon capture. High-throughput computationalscreening efforts can accelerate the discovery of new adsorbents but sometimes require significant computa-tional resources to explore the large space of possible materials. We report the in silico discovery of high-performing adsorbents for precombustion CO2 capture by applying a genetic algorithm to efficiently searcha large database of metal-organic frameworks (MOFs) for top candidates. High-performing MOFs identifiedfrom the in silico search were synthesized and activated and show a high CO2 working capacity and a highCO2/H2 selectivity. One of the synthesized MOFs shows a higher CO2 working capacity than any MOF reportedin the literature under the operating conditions investigated here.

wn

on July 9, 2020http://advances.sciencem

ag.org/loaded from

INTRODUCTIONScientists, political leaders, and common citizens around the world areincreasingly alarmed by the rapidly rising levels of CO2 in the atmo-sphere (1–3). In the United States, nearly 40% of CO2 emissions comefrom burning fossil fuels to generate electricity in power plants (4).Because renewable energy sources (such as wind and solar) are stillfar from replacing fossil fuels as the primary energy source to powerthe planet, almost all recent scenarios put forth to reduce CO2 emis-sions include a significant midterm role for carbon capture and storage.Capturing CO2 from existing power plants requires the separationof dilute amounts of CO2 from N2 in a low-pressure stream via aso-called “postcombustion” strategy. An easier strategy for newlycommissioned power plants is to use “precombustion” CO2 capturetechnology. An example of precombustion CO2 capture is shown inFig. 1 where natural gas is reformed to produce syngas (a mixture ofCO and H2), which is run through a water-gas shift reaction (WGSR)to produce a high-pressure stream of CO2 and H2. The CO2 is thenremoved from this stream, and the resulting H2 is burned for energyproduction, with water as a by-product (5). Currently available pre-combustion CO2 capture technology using solvents, such as Selexol,methanol, or methyldiethanolamine, is estimated to cost around $60per metric ton of captured CO2 (6), which is 50% higher than the U.S.Department of Energy target. On the other hand, if pressure-swingadsorption (PSA) were used to capture the CO2 from the high-pressuregas mixture obtained from the WGSR (Fig. 1), the cost of CO2

capture could be reduced if a selective adsorbent material with a highCO2 working capacity were available. Increasing the CO2 working ca-pacity means that, for instance, less adsorbent material is needed forthe PSA unit, which in turn reduces the cost of CO2 capture.

Metal-organic frameworks (MOFs) are a class of nanoporousmaterials that could potentially provide higher CO2 working capacitiesfor precombustion CO2 capture than traditional sorbents, such as zeo-lites and activated carbons, because of their high pore volumes andsurface areas. MOFs are synthesized by the self-assembly of organicand inorganic building blocks, and different combinations of thebuilding blocks can produce MOFs with different physical and chem-ical properties, making this class of materials incredibly versatile andtunable for a wide range of applications (7–9). For instance, MOFshave been investigated for a wide range of gas storage and separationapplications (10–13), but only a limited number of MOFs [for exam-ple, Cu-BTTri (14, 15), Mg-MOF-74 (16), and Ni-4PyC (17)] have beentested for precombustion CO2 capture. Given the large number ofMOFs synthesized to date (18), experimental evaluation of all MOFsfor this application would be impractical at best, and other approachesmust be used to identify promising materials.

High-throughput computational screening has emerged in the pastfew years as a powerful approach to accelerating the evaluation of ad-sorbent materials for postcombustion CO2 capture (19, 20), methanestorage (21–24), Xe/Kr separation (25, 26), H2 storage (27), and biofueland hydrocarbon separations (28). In this approach, molecular simu-lations are carried out to evaluate the performance of existing or not-yet-synthesized adsorbent materials to find top-performing candidatesand to reveal key structure-property relationships between materialperformance and physical characteristics (for example, pore volumeand enthalpy of adsorption). In one example, these computationalscreening efforts led to the synthesis of a new MOF for methane stor-age (24). In other cases, computational screening was used to find theperformance limits of MOFs for applications such as methane andhydrogen storage (23, 29). To date, evaluation of the materials hasbeen relatively fast, and hundreds of thousands of hypotheticalmaterials could be screened because the guest species of interestwere small molecules accurately described by classical force fields.However, there are millions of potential MOFs, and for many applica-tions, calculating the performance of each material will be more time-consuming, for example, if quantum mechanical calculations arerequired (30). The current efforts have calculated the properties of allcandidate materials in a given set of materials. However, many of thesecandidates have low performance, and much time is wasted in evaluating

1 of 9

SC I ENCE ADVANCES | R E S EARCH ART I C L E

on July 9, 2020http://advances.sciencem

ag.org/D

ownloaded from

the less-promising materials. Therefore, several groups are investigatingways to reduce the time invested in evaluating large numbers ofmaterials (26, 28, 31–33). For example, Simon et al. (26) trained decisiontrees to identify promising materials for Xe/Kr separation in a databaseof more than 600,000 nanoporous materials. Each material was evalu-ated by the decision trees based on six textural properties and one prop-erty based on the energetics of adsorption sites, and only the mostpromising materials were evaluated using grand canonical Monte Carlo(GCMC) simulations. In another example, Bao et al. (32) implemented agenetic algorithm (GA) to evolve ditopic linkers using well-known reac-tions and commercially available organic compounds. The evolved lin-kers varied greatly in terms of complexity, and provided they fulfilledcertain requirements (for example, being relatively linear), they weresubstituted into their corresponding “parent” MOFs, whose topologyand metal node remained unchanged. The MOF “children” were thenevaluated for methane storage using GCMC simulations. Although theabovementioned study solely focused on linker evolution, a strongly ap-pealing aspect of this work is that not every possible linker was evalu-ated. What is needed is a more efficient way to explore a given databaseor “space” of materials to find top performers without exhaustively eval-uating every material.

Here, a GA was developed to find top-performing MOFs for pre-combustion CO2 capture, and the method was applied to a databaseof 55,163 hypotheticalMOFs (hMOFs) (21). One of the top-performingMOFs that emerged from the GA-guided search was synthesized, acti-vated, and tested. Experimental pure-component CO2 and H2 iso-therms on the activated MOF showed good agreement with thesimulation predictions. Applying the ideal adsorbed solution theory(IAST) to obtainmixture isotherms from the experimental data, we findthat the synthesizedMOFhas aCO2working capacity of 3.8mol/kg anda CO2/H2 selectivity of 60. The selectivity is high enough to obtain 99%H2 purity according to our PSA process simulations, and the workingcapacity is the highest reported to date for the operating conditionsconsidered here. Using the structure-property relationships obtainedfrom the calculations, we also identified 531 promising MOFs in adatabase of 5109 existing (already synthesized)MOFs.Molecular simu-lations were carried out on these structures, and one of the top-performing materials was identified, synthesized, activated, and tested.

Chung et al., Sci. Adv. 2016;2 : e1600909 14 October 2016

From the measured isotherms and IAST, the MOF shows a CO2

working capacity of 3.1 mol/kg and a CO2/H2 selectivity of 48.

RESULTSValidation of the GAWe implemented a GA search strategy as described in Materials andMethods and Fig. 2 (A and B) and applied it to a large database of51,163 hMOFs (21). Before applying the GA to find high-performinghMOFs for precombustion CO2 capture, we tested the efficiency androbustness of our GA implementation by trying to find the hMOFswith the highest gravimetric and volumetric surface areas and meth-ane working capacity. Note that these properties were previouslycalculated for all hMOFs; thus, we already knew the identity of thebest materials for these test cases. Figure 2 (C to E) shows the histo-grams of methane working capacities and gravimetric and volumetricsurface areas, respectively, for all hMOFs and for the initial populationof 100 hMOFs used in all GA runs. These properties were used asdifferent measures of hMOF “fitness” that the GA should attemptto improve. For each fitness measure, 100 independent GA runs werecarried out for 100 generations each. The histograms of theperformance of the best hMOF at the end of each GA run are shownin Fig. 2 (F to H). For each of the three performance measures, the GAalways found a structure within the top 4% by the 10th generation.

Application of the GAStarting with the same initial hMOF population that was used forthe validation of the GA, the algorithm was applied to search fortop hMOFs for precombustion CO2 capture. We note that, inher-ent to the GA formalism, it is not possible to determine whetherthe best hMOF in the database is identified through the GA search.However, as an objective measure of the success of our GA approach,we sought to identify MOFs with better performance metrics thanthose reported for MOFs to date. Three independent GA runs were per-formed to separately optimize three different fitness measures, namely,the CO2 working capacity (DN1), the CO2/H2 selectivity (aads12 ), and anadsorbent performance score (APS), which is the product of the formertwo quantities, as defined in Materials and Methods. Each GA was runfor 10 generations, and GCMC simulations were carried out for eachnew hMOF considered. To improve the computational efficiency, theGCMC results were saved, and if an hMOF in the nth generationwas already evaluated in a previous generation (from any of the threeruns), no new GCMC simulations were carried out for that structure.Details of the GA runs are provided in section S3. Figure 3 summarizesthe progress of the GA as it searched for top-performing hMOFs foreach fitness measure. For each generation, the average fitness of thepopulation and the fitness of the best-performing hMOF from thepopulation (the elite) are plotted. As the GA evolved, for the top-performing MOF from each generation, the CO2 working capacityimproved from ca. 7 mol/kg (1st generation) to ca. 8 mol/kg (10thgeneration), the CO2/H2 selectivity improved from ca. 700 to ca. 2600,and the APS improved from ca. 1000 mol/kg to ca. 1200 mol/kg.

The progression of genes during the GA run to optimize theAPS is shown in Fig. 4. This figure shows that the building blocksthat consistently result in top-performing hMOFs become dominantas the GA progresses. As the new generations of hMOFs are evolved,zinc paddlewheel nodes (gene “1” from Fig. 4A) and [1,1′:4′,1′′]terphenyl-3,3′,5,5′′-tetracarboxylic (TPTC) acid linkers (gene “38” from Fig. 4, Band C) become dominant. On the other hand, no specific functional

Fig. 1. Simplified schematic ofprecombustionCO2 capture.Schematicwas adaptedfromWilcox’s study (1). Natural gas, which is mainly methane, is reformed to produce amixture of CO and H2, which then goes through a WGSR to produce a mixture of CO2

andH2. The stream from theWGSRgoes throughaCO2 separation unit to producehigh-purity hydrogen, which is combusted to generate electricity.

2 of 9

SC I ENCE ADVANCES | R E S EARCH ART I C L E

on July 9, 2020http://advances.sciencem

ag.org/D

ownloaded from

group becomes dominant, which suggests that the choice of organiclinkers and inorganic building blocks plays a more important role inoptimizing the APS of MOFs for this application.

Table 1 shows that the number of GCMC simulations carried outduring each GA run is significantly smaller than what would be re-quired for an exhaustive search of all hMOFs. Each GA run requiredless than 1% of the computational time compared to a brute forcesearch. Only 730 of 51,163 hMOFs were evaluated even if the threeGA runs, together, were considered. Note that the GA is not guar-anteed to find the very best solution (here, the best hMOF from theentire database). However, the GA validation tests illustrated in Fig.2 (C to H) suggest that the top-performing hMOFs that were iden-

Chung et al., Sci. Adv. 2016;2 : e1600909 14 October 2016

tified should be within the top 4% for each fitness measure (seesection S3.)

Identification and synthesis of the top-performing MOFsThe three GA runs produced data for the CO2 working capacity,the CO2/H2 selectivity, and the APS of 730 genetically uniquehMOFs. Figure 5 summarizes the data obtained from all GA runs,where CO2/H2 selectivity is plotted as a function of CO2 workingcapacity, with the color of the data points indicating the APS value.The top-performing hMOFs based on the CO2 working capacity havevoid fractions between 0.6 and 0.8 and pore diameters between 8 and10 Å. The MOFs with the highest CO2/H2 selectivity have lower void

Fig. 3. Performance of the GA. (A to C) Results for three independent GA runs dedicated to optimize (A) CO2 working capacity, (B) CO2/H2 selectivity, and (C) APS.

Fig. 2. Overview and validation of the GA. (A) An example chromosome and the corresponding hMOF structure. Colors help illustrate the correspondence between thegenes and the hMOF structural features. (B) Workflow of GA. (C to E) Histograms for all hMOFs (gray) and for the initial population used in the GA runs (green). (C) Methaneworking capacity. (D) Gravimetric surface area. (E) Volumetric surface area. (F to H) Histograms collected from 100 GA runs show the fitness of the top-performing MOF at theend of each run. (F) Methane working capacity. (G) Gravimetric surface area. (H) Volumetric surface area. The vertical lines in (F) to (H) correspond to the fitness of the topperformer from the initial population (black) and from the whole database (red).

3 of 9

SC I ENCE ADVANCES | R E S EARCH ART I C L E

on July 9, 2020http://advances.sciencem

ag.org/D

ownloaded from

fractions (<0.5) and smaller pore diameters (<5 Å), and consequently,they have a low CO2 working capacity because the pore space is filledbefore the pressure reaches 20 bar. Figure 5 shows that there is a cleartrade-off between CO2 working capacity and CO2/H2 selectivity.

The APS aims to account for the effect of both the selectivityand the working capacity of the adsorbent on the purity and recov-ery of the PSA process. hMOFs with high APS values were consideredfor possible synthesis. In particular, we focused on hMOFs that wereboth located near the Pareto front in Fig. 5 and had better workingcapacities than the experimentally tested MOFs also shown in Fig. 5.

Chung et al., Sci. Adv. 2016;2 : e1600909 14 October 2016

In this way, we obtained a “preliminary” list of nearly 50 high-performing hMOFs. From this list, we identified 12 MOFs based onthe nbo topology, which combines metal paddlewheels and planar tet-racarboxylate organic linkers, as materials that we anticipated wecould successfully synthesize, on the basis of our previous experiencein MOF synthesis. These 12 MOFs correspond to 6 zinc-based MOFsand their 6 copper-based counterparts. Before synthesis, the adsorp-tion properties of these 12 hMOFs were recomputed using densityfunctional theory (DFT)–derived partial atomic charges (34) for thehMOF atoms [approximate charges (35) were used for the GAscreening]. We also arranged the functional groups in each of thesehMOFs so that they were uniformly placed on each constituent organ-ic linker in the material (section S5). From these simulations, we iden-tified the TPTC acid linker [one of the linkers (gene 38) that wasdominant in Fig. 4] functionalized with two ethoxy groups to optimizethe APS value for both the copper- and zinc-based hMOFs (table S3).For both the copper- and zinc-based hMOFs having this linker, wepredicted high working capacities (5.7 mol/kg) and high selectivities(132 and 188 for copper and zinc cases, respectively). Because of theanticipated higher stability of copper paddlewheels upon activation, weselected the copper hMOF as the synthesis target. The selected hMOFis a functionalized version of a previously synthesized MOF, NOTT-101 (36), and our simulations predict it to have a CO2 working capac-ity that is higher than that reported for the few materials previouslyexperimentally tested for precombustion CO2 capture (see Fig. 5).

The ethoxy-functionalized version of NOTT-101 (NOTT-101/OEt)was synthesized and activated. It should be noted that, unbeknown

Fig. 4. Gene evolution during GA optimization of APS. (A to D) Genes corresponding to (A) inorganic building blocks, (B) primary organic linkers, (C) secondaryorganic linkers, and (D) functional groups.

Table 1. Comparison of computational effort for brute force searchversus GA. DN1 is the CO2 working capacity, aads12 is the CO2/H2 selectivity,and APS is the adsorbent performance score, as defined in Eqs. 1 to 3. Thenumber of GCMC simulations for the GA search corresponds to thenumber of simulations carried out up to 10 generations.

Method

Fitnessmeasure

Number of GCMCsimulations

Relative computationaltime (%)

Bruteforce

—

51,163 100

GA

DN1

340 0.66

aads12

322 0.63

APS

268 0.52

4 of 9

SC I ENCE ADVANCES | R E S EARCH ART I C L E

on July 9, 2020http://advances.sciencem

ag.org/D

ownloaded from

to us at the time of synthesis, NOTT-101/OEt had already beensynthesized in 2013 (37). However, to the best of our knowledge, theproperties of NOTT-101/OEt for precombustion CO2 capture had notbeen examined until now. Important for performance, the synthesisand activation protocol we used in this study improves the exper-imental apparent Brunauer-Emmett-Teller (BET) area of the MOFfrom the previously reported 1293 to 1900 m2/g, which is in muchcloser agreement with the simulated BET area for the perfectcrystal (2008 m2/g) and thus indicates a high-quality sample. Theseexperimental and simulated BET areas were obtained from exper-imental and simulated N2 isotherms, applying the four BET consist-ency criteria (38). The saturation loadings from the N2 isotherms wereused to determine the experimental (0.743 cm3/g) and simulated(0.797 cm3/g) MOF pore volumes, indicating a 92% activation of theMOF pores.

The CO2 and H2 adsorption isotherms of NOTT-101/OEt weremeasured experimentally up to 16 bar at 313 K. Figure 6C showsthe comparison between the experimental and simulated absoluteadsorption isotherms for NOTT-101/OEt. In the figure, the experi-mental isotherms were multiplied by a factor of 1.09 to account forthe 92% pore activation. There is good agreement between measuredand simulated H2 isotherms, with only a slight (in absolute terms)underestimation by the simulation. There is fair agreement betweenmeasured and simulated CO2 isotherms, with simulations somewhatoverpredicting the CO2 uptake, especially at the intermediate pressurerange. Note that actual measured data instead of scaled-up data arepresented and discussed in subsequent sections.

Chung et al., Sci. Adv. 2016;2 : e1600909 14 October 2016

Application of structure-property relationships to findcandidates in a separate databaseHigh-throughput computational screening produces large volumesof data that can be used to find underlying structure-property relation-ships, such as how the performance for a given application dependson the MOF surface area, void fraction, etc. From the GA runs, structure-property relationships also emerged but with significantly fewercomputations than with a brute force approach (Table 1). In princi-ple, the structure-property relationships emerging from a GAscreening could be used to find additional high-performing MOFswithout the need for further simulations. To test the applicabilityof this approach, we searched for high-performing MOFs in aseparate database: the computation-ready, experimental (CoRE)MOF database (18). Two key advantages of the 5109 structures inthe CoRE MOF database are that all of them have already beensynthesized and that their synthesis protocols are available in the lit-erature, which can facilitate the synthesis and testing of any candi-dates identified from computational screening or other methods.

We identified ranges of optimal physical properties (pore-limitingdiameter, largest cavity diameter, gravimetric surface area, and heliumvoid fraction) for each performance measure on the basis of the prop-erties of the hMOFs within the top 1% of the 730 hMOFs evaluatedduring the GA runs (table S4). These properties were then used toidentify 75, 99, and 357 candidate CoRE MOFs for high CO2 workingcapacity, CO2/H2 selectivity, and APS, respectively. For each group ofCoRE MOFs, GCMC simulations were carried out to evaluate theiradsorption properties, and the results showed that 5 (of 75), 14 (of99), and 13 (of 357) of these CoRE MOFs have a high working capac-ity, a high selectivity, and a high APS, respectively (see section S5).Note that although the “hit rate” was low (for example, only 5 of75 candidates had a high working capacity), these properties were stilluseful in identifying high-performing CoRE MOFs without havingto evaluate the full CoRE MOF database. One of the identified high-performing CoRE MOFs [Cambridge Structural Database REFCODE:VEXTUO (39)] with a predicted CO2 working capacity of 6.0 mol/kgand a CO2/H2 selectivity of 83 was selected for synthesis, activation,and testing. Note that although all CoRE MOFs have synthesisprotocols available, successful activation is not guaranteed for all ofthese MOFs (18). Therefore, the comparison of reported BET areasand the geometrically calculated surface areas was also a factor indeciding which MOF to synthesize, because marked differences be-tween the two values could indicate a tendency of the MOF to collapseupon activation or difficulty in removing trapped solvents or otherimpurities. VEXTUO was synthesized following the protocol in theliterature (39). Experimental and simulated BET areas were 1977 and2031 m2/g, respectively, and the measured and simulated pore volumeswere 0.75 and 0.78 cm3/g, respectively. Single-component isotherms ofCO2 and H2 were measured at 303 K. The simulated and experimentalH2 isotherms were in good agreement, but the simulated CO2 iso-therms were ~35% higher (at 15 bar) than the measured isothermsfor VEXTUO (see section S9).

Comparison with other adsorbentsTable 2 summarizes a comparison among the two MOFs from thiswork and three high-performing MOFs (Mg-MOF-74, Cu-BTTri,and Ni-4PyC) known from the literature (all MOFs in Table 2are illustrated in Fig. 6D). We used IAST to compute the mixture iso-therms from high-pressure (up to 16 to 20 bar) experimental single-component CO2 and H2 isotherms of the MOFs in Table 2 (see

Fig. 5. Aggregated data from the GA search (circles) for precombustion CO2

capture. Each point corresponds to an hMOF and is colored according to the valueof the APS. The data point for the synthesis target identified from the GA search(NOTT-101/OEt) obtained from GCMC simulations is shown in gray. Data points forMOFs experimentally tested in the literature for the operating conditions studiedhere are shown in black-outlined yellow squares. The properties of Cu-BTTri andMOF-74 were computed on the basis of the mixture isotherms obtained from IASTreported by Herm et al. (14) and Herm et al. (16), respectively (see section S10). Theproperties of Ni-4PyC were approximated from simulated and experimental data re-ported by Nandi et al. (17). Exp., experimental; sim., simulated.

5 of 9

SC I ENCE ADVANCES | R E S EARCH ART I C L E

on July 9, 2020http://advances.sciencem

ag.org/D

ownloaded from

section S10). For NOTT-101/OEt and VEXTUO, experimental datawere obtained in this work, whereas the experimental isotherms forMg-MOF-74 and Cu-BTTri were obtained from the literature. Pro-cess simulations of an example PSA unit (see section S4) show thatCO2/H2 selectivities higher than 30 are enough to achieve 99% H2

purity, and higher working capacities reduce the amount of adsorb-ent required for the separation.

NOTT-101/OEt has the highest CO2 working capacity amongthe five MOFs and a relatively high CO2/H2 selectivity of 60. Pre-viously, Mg-MOF-74 has been noted for its high CO2 working ca-pacity (14). However, there is some question about how to definethe working capacity for PSA processes. Here, we introduce a def-inition of the CO2 working capacity (see Materials and Methods)based on the gas composition profiles from a PSA process simulation(40) under realistic operating conditions (see section S4). On the basisof this definition, the CO2 working capacity of Mg-MOF-74 is thelowest among the listed MOFs, whereas its CO2/H2 selectivity is thehighest. However, because our process modeling shows that selectiv-ities higher than 30 are enough to reach >99% purity, MOFs such asNOTT-101/OEt (with an approximately 62% higher working capacitywith respect to Mg-MOF-74) could be better suited for precombustioncarbon capture. Similarly, both VEXTUO (which was identified from

Chung et al., Sci. Adv. 2016;2 : e1600909 14 October 2016

the CoRE MOF database) and Ni-4PyC (which was reported duringthe preparation of this manuscript) also have working capacities high-er than that of Mg-MOF-74 (ca. 19% and ca. 31%, respectively), whilehaving selectivities higher than 30. Note that our estimation of the Ni-4PyC working capacity is approximate, on the basis of the availableexperimental and simulation data reported by Nandi et al. (17). Onthe other hand, whereas Cu-BTTri has a high CO2 working capacity(only slightly lower than NOTT-101/OEt), its selectivity is lower than 30.

DISCUSSIONHere, we successfully demonstrated that a GA could be used to ef-ficiently identify top adsorbent materials for precombustion CO2

capture among thousands of hMOFs. The GA reduced the compu-tational time by at least two orders of magnitude relative to a bruteforce search. One of the top-performing MOFs, NOTT-101/OEt, wassynthesized and tested, and the experimental pure-component CO2

and H2 isotherms agree well with the simulation predictions. IAST-predicted mixture isotherms show that the CO2 working capacity ofNOTT-101/OEt is 3.8 mol/kg, with a CO2/H2 selectivity of 60. Wealso showed that the structure-property relationships obtained fromthe GA-guided search could be used to discover top-performing

Fig. 6. MOF studied for precombustion CO2 capture. (A) Inorganic node and organic ligand used to synthesize NOTT-101/OEt. (B) Atomistic representation of NOTT-101/OEt. Copper, carbon, and oxygen atoms are shown in orange, black, and red, respectively. Hydrogen atoms are omitted for clarity. Purple spheres represent thecavities of NOTT-101/OEt. (C) Experimental and simulated absolute single-component CO2 and H2 isotherms for NOTT-101/OEt at 313 K. (D) Crystal structures of otherMOFs listed in Table 2. Mg-MOF-74, Cu-BTTri, Ni-4PyC, and VEXTUO are based on Mg2, Cu4Cl, Ni2O, and Ni2O inorganic nodes, respectively, connected by the linkersillustrated below each MOF. MOF pore cages are illustrated with colored spheres. (The MOFs are not all drawn to the same scale.)

6 of 9

SC I ENCE ADVANCES | R E S EARCH ART I C L E

http://advances.D

ownloaded from

MOFs in different databases without the need for a large number ofadditional simulations. The methods demonstrated in this work (boththe GA-guided and the structure-property–guided search) could beapplied to search for high-performing MOFs for other applicationsand should be especially useful when the performance evaluation re-quires a large amount of computational time, such as simulations in-volving large, complex molecules, or when quantum mechanicalcalculations are required.

on July 9, 2020sciencem

ag.org/

MATERIALS AND METHODSCalculation of adsorption propertiesGCMC simulations (41) were carried out as implemented in the RASPAsimulation code (42, 43) to compute adsorption loadings at 313 K. De-tails of simulations and models are provided in section S1. Simulationswere carried out to compute the CO2 adsorption loadings for pure CO2

at 1 bar and the CO2 and H2 adsorption loadings for a 20:80 CO2/H2

mixture at 20 bar. The following adsorbent evaluation criteria wereused to measure the fitness of each MOF

DN1 ¼ Nads1 � Ndes

1 ð1Þ

aads12 ¼Nads

1

�y1

� �

Nads2

�y2

� � ð2Þ

APS ¼ DN1 � aads12 ð3Þ

Here, DN1 is the CO2 working capacity, Nads1 and Nads

2 are the CO2

and H2 adsorption loadings for the CO2/H2 mixture at 20 bar, andNdes1

is the CO2 adsorption loading for pure CO2 at 1 bar. Section S4 dis-cusses the reasons for using this definition of the CO2 working capac-ity, which is different from what is sometimes used. aads12 is the CO2/H2

selectivity, and y1 and y2 are the mole fractions of CO2 (0.2) and H2

(0.8) in the gas phase, respectively. We also defined an APS in Eq. 3,

Chung et al., Sci. Adv. 2016;2 : e1600909 14 October 2016

similar to the performance measure defined by Bai et al. (28), as a wayto account for the impact of both the CO2 working capacity and theCO2/H2 selectivity on the performance of a PSA unit.

Database of hMOFsThe hMOFs explored in this work were obtained from the WLLFHSdatabase of hMOFs (21). The structure of each MOF in this databasecan be characterized by a sequence of six integers (a chromosome).Genes 1 to 6 encode the interpenetration capacity, the actual inter-penetration level, and the identities of the inorganic node, primary link-er, secondary linker, and functional groups of a given hMOF (Fig. 2A).The 137,193 MOFs in the WLLFHS database can be described by51,163 unique chromosomes due to conformational isomers andstructures that differ only in the positioning of the functional groups(see section S2). The simulation of MOFs with identical genes resultedin very similar performance because they have similar structures. There-fore, the original WLLFHS database was reduced to 51,163 MOFs byselecting 1MOF fromeach unique chromosome. This reduced databasewas subsequently explored using the GA developed in this work.

Genetic algorithmGAs are a class of optimization methods that mimic natural selec-tion. In a typical GA, a population of candidate solutions is evolvedin the solution space toward higher values of some fitness function.Here, the solutions were hMOFs, and the GA evolved the geneticinformation of hMOFs to optimize one of the performance mea-sures defined in Eqs. 1 to 3. We started with an initial populationof 100 hMOFs (that is, the first generation) that was selected manuallyto ensure that each possible gene was carried by at least 1 hMOF. Eachgeneration was evolved to create a subsequent generation. All genera-tions had a population of 100 hMOFs. Elitism was implemented toensure that the hMOF with the highest fitness in the nth generationappears in the (n + 1)th generation. All other hMOFs in the (n + 1)thgeneration were obtained by applying genetic operations on hMOFpairs selected from the nth generation. These hMOF pairs wereselected using the tournament method (44). In a tournament, thehMOF with the higher fitness value between the two randomlyselected hMOFs from the nth generation was selected with a 95%probability. Each hMOF pair was obtained from two independenttournament selections, and single-point crossover was subsequentlycarried out on the selected pair of hMOFs with a 65% probability.Each gene in the new chromosome had a 5% probability to undergoa mutation. Each hMOF pair in the nth generation produced onehMOF for the (n + 1)th generation. Figure 2B summarizes the work-flow of the GA, and full details can be found in section S3.

SUPPLEMENTARY MATERIALSSupplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/2/10/e1600909/DC1section S1. Computational methods.section S2. Genetic information for the WLLFHS hMOF database.section S3. Genetic algorithm.section S4. Discussion about a definition of CO2 working capacity.section S5. Identification of top-performing hMOFs for synthesis.section S6. Synthesis of NOTT-101/OEt and VEXTUO.section S7. Powder x-ray diffraction data.section S8. N2 sorption data.section S9. CO2 and H2 simulated and measured isotherms for VEXTUO.section S10. IAST calculations.fig. S1. Nitrogen model.

Table 2. CO2 working capacity and CO2/H2 selectivity for severalMOFs.

MOF C

O2 working capacity (mol/kg)

CO2/H2

selectivityR

eference

NOTT-101/OEt

3.8

60 T his work

Cu-BTTri

3.7 20 (15)

Ni-4PyC*

3.4 279 (17)

VEXTUO†

3.1 48 T his work

Mg-MOF-74

2.6 365 (14)

*Results are approximate and are obtained on the basis of experimental andsimulation data reported by Nandi et al. (17). Specifically, selectivity was ob-tained from mixture simulation data at 313 K/20 bar, and working capacitywas obtained from mixture simulation data at 313 K/20 bar and pure-component experimental data at 303 K/1 bar. †Measured at 303 K.

7 of 9

SC I ENCE ADVANCES | R E S EARCH ART I C L E

http://advances.sciencemag.o

Dow

nloaded from

fig. S2. Carbon dioxide model.fig. S3. Hydrogen model.fig. S4. Correspondence between genes and inorganic nodes.fig. S5. Correspondence between genes and organic linkers.fig. S6. Correspondence between genes and functional groups.fig. S7. Examples of duplicate MOFs.fig. S8. Duplicity of structures in the WLLFHS hMOF database.fig. S9. Performance similarity among duplicate hMOFs.fig. S10. Textural properties in the original and reduced WLLFHS hMOF database.fig. S11. Gene distribution in the reduced WLLFSH hMOF database.fig. S12. Suitable combinations of building blocks.fig. S13. A workflow for the genetic operations of crossover and mutation.fig. S14. Determination of the optimal length of GA runs.fig. S15. Schematic for a Skarstrom cycle of a CO2/H2 separation unit.fig. S16. Gas concentration profiles during the Skarstrom cycle.fig. S17. Impact of CO2/H2 selectivity on H2 purity.fig. S18. Rearrangement of functional groups in potential synthesis targets.fig. S19. MOF representative cluster for DFT calculations.fig. S20. Performance of investigated hMOFs and CoRE MOFs.fig. S21. Ligand for NOTT-101/OEt.fig. S22. Ligand for VEXTUO.fig. S23. Nitrogen adsorption for NOTT-101/OEt.fig. S24. Nitrogen adsorption for VEXTUO.fig. S25. Powder x-ray diffraction data patterns for NOTT-101/OEt.fig. S26. Powder x-ray diffraction data patterns for VEXTUO.fig. S27. CO2 and H2 single-component adsorption for VEXTUO.fig. S28. Dual-site Langmuir fit for CO2 and H2 isotherms of NOTT-101/OEt.fig. S29. Dual-site Langmuir fit for CO2 and H2 isotherms of VEXTUO.fig. S30. Dual-site Langmuir fit for CO2 and H2 isotherms of Mg-MOF-74.fig. S31. Dual-site Langmuir fit for CO2 and H2 isotherms of Cu-BTTri.fig. S32. IAST accuracy test on NOTT-101/OEt.fig. S33. CO2/H2 selectivity versus pressure for NOTT-101/OEt.fig. S34. CO2 working capacity as a function of pressure.table S1. Number of hMOFs in different subsets with the gene-based identification criteria.table S2. Data from GA testing.table S3. CO2/H2 adsorption properties for the Zn- and Cu-based nbo hMOFs.table S4. Textural properties of the top 1% of evaluated hMOFs for three performancemeasures.table S5. List of top 30 CoRE MOFs.table S6. IAST parameters for the 20:80 mixture of CO2/H2.References (45–66)

on July 9, 2020rg/

REFERENCES AND NOTES1. J. Wilcox, Carbon Capture (Springer, 2012).2. B. Smit, J. A. Reimer, C. M. Oldenburg, I. C. Bourg, Introduction to Carbon Capture and

Sequestration (Imperial College Press, 2014).3. 2015 Paris agreement under the United Nations framework convention on climate

change; http://ec.europa.eu/clima/policies/international/negotiations/paris.4. P. Friedlingstein, R. M. Andrew, J. Rogelj, G. P. Peters, J. G. Canadell, R. Knutti, G. Luderer,

M. R. Raupach, M. Schaeffer, D. P. van Vuuren, C. Le Quéré, Persistent growth of CO2

emissions and implications for reaching climate targets. Nat. Geosci. 7, 709–715 (2014).5. S. Sircar, T. C. Golden, Purification of hydrogen by pressure swing adsorption. Sep. Sci.

Technol. 35, 667–687 (2000).6. U.S. Department of Energy Office of Fossil Energy, “Carbon Capture Technology

Program Plan” (Department of Energy Office of Fossil Energy, 2013); https://www.netl.doe.gov/File%20Library/Research/Coal/carbon%20capture/Program-Plan-Carbon-Capture-2013.pdf.

7. H. Furukawa, K. E. Cordova, M. O’Keeffe, O. M. Yaghi, The chemistry and applications ofmetal-organic frameworks. Science 341, 1230444 (2013).

8. A. G. Slater, A. I. Cooper, Function-led design of new porous materials. Science 348,aaa8075 (2015).

9. F.-X. Coudert, A. H. Fuchs, Computational characterization and prediction of metal–organic framework properties. Coord. Chem. Rev. 307, 211–236 (2016).

10. S. Krause, V. Bon, I. Senkovska, U. Stoeck, D. Wallacher, D. M. Tobbens, S. Zander, R. S. Pillai,G. Maurin, F.-X. Coudert, S. Kaskel, A pressure-amplifying framework material with negativegas adsorption transitions. Nature 532, 348–352 (2016).

11. W. R. Lee, S. Y. Hwang, D. W. Ryu, K. S. Lim, S. S. Han, D. Moon, J. Choi, C. S. Hong,Diamine-functionalized metal–organic framework: Exceptionally high CO2 capacitiesfrom ambient air and flue gas, ultrafast CO2 uptake rate, and adsorption mechanism.Energy Environ. Sci. 7, 744–751 (2014).

Chung et al., Sci. Adv. 2016;2 : e1600909 14 October 2016

12. A. M. Fracaroli, H. Furukawa, M. Suzuki, M. Dodd, S. Okajima, F. Gándara, J. A. Reimer,O. M. Yaghi, Metal–organic frameworks with precisely designed interior for carbondioxide capture in the presence of water. J. Am. Chem. Soc. 136, 8863–8866 (2014).

13. J. A. Mason, J. Oktawiec, M. K. Taylor, M. R. Hudson, J. Rodriguez, J. E. Bachman,M. I. Gonzalez, A. Cervellino, A. Guagliardi, C. M. Brown, P. L. Llewellyn, N. Masciocchi,J. R. Long, Methane storage in flexible metal–organic frameworks with intrinsic thermalmanagement. Nature 527, 357–361 (2015).

14. Z. R. Herm, J. A. Swisher, B. Smit, R. Krishna, J. R. Long, Metal–organic frameworks asadsorbents for hydrogen purification and precombustion carbon dioxide capture. J. Am.Chem. Soc. 133, 5664–5667 (2011).

15. T. M. McDonald, D. M. D’Alessandro, R. Krishna, J. R. Long, Enhanced carbon dioxidecapture upon incorporation of N,N′-dimethylethylenediamine in the metal–organicframework CuBTTri. Chem. Sci. 2, 2022–2028 (2011).

16. Z. R. Herm, R. Krishna, J. R. Long, CO2/CH4, CH4/H2 and CO2/CH4/H2 separations at highpressures using Mg2(dobdc). Microporous Mesoporous Mater. 151, 481–487 (2012).

17. S. Nandi, P. De Luna, T. D. Daff, J. Rother, M. Liu, W. Buchanan, A. I. Hawari, T. K. Woo,R. Vaidhyanathan, A single-ligand ultra-microporous MOF for precombustion CO2 captureand hydrogen purification. Sci. Adv. 1, e1500421 (2015).

18. Y. G. Chung, J. Camp, M. Haranczyk, B. J. Sikora, W. Bury, V. Krungleviciute, T. Yildirim,O. K. Farha, D. S. Sholl, R. Q. Snurr, Computation-ready, experimental metal–organicframeworks: A tool to enable high-throughput screening of nanoporous crystals. Chem.Mater. 26, 6185–6192 (2014).

19. L.-C. Lin, A. H. Berger, R. L. Martin, J. Kim, J. A. Swisher, K. Jariwala, C. H. Rycroft,A. S. Bhown, M. W. Deem, M. Haranczyk, B. Smit, In silico screening of carbon-capturematerials. Nat. Mater. 11, 633–641 (2012).

20. C. E. Wilmer, O. K. Farha, Y.-S. Bae, J. T. Hupp, R. Q. Snurr, Structure–property relationshipsof porous materials for carbon dioxide separation and capture. Energy Environ. Sci. 5,9849–9856 (2012).

21. C. E. Wilmer, M. Leaf, C. Y. Lee, O. K. Farha, B. G. Hauser, J. T. Hupp, R. Q. Snurr, Large-scalescreening of hypothetical metal–organic frameworks. Nat. Chem. 4, 83–89 (2011).

22. D. A. Gómez-Gualdrón, C. E. Wilmer, O. K. Farha, J. T. Hupp, R. Q. Snurr, Exploring thelimits of methane storage and delivery in nanoporous materials. J. Phys. Chem. C 118,6941–6951 (2014).

23. C. M. Simon, J. Kim, D. A. Gomez-Gualdron, J. S. Camp, Y. G. Chung, R. L. Martin,R. Mercado, M. W. Deem, D. Gunter, M. Haranczyk, D. S. Sholl, R. Q. Snurr, B. Smit, Thematerials genome in action: Identifying the performance limits for methane storage.Energy Environ. Sci. 8, 1190–1199 (2015).

24. D. A. Gomez-Gualdron, O. V. Gutov, V. Krungleviciute, B. Borah, J. E. Mondloch, J. T. Hupp,T. Yildirim, O. K. Farha, R. Q. Snurr, Computational design of metal–organic frameworksbased on stable zirconium building units for storage and delivery of methane. Chem.Mater. 26, 5632–5639 (2014).

25. B. J. Sikora, C. E. Wilmer, M. L. Greenfield, R. Q. Snurr, Thermodynamic analysis of Xe/Krselectivity in over 137 000 hypothetical metal–organic frameworks. Chem. Sci. 3,2217–2223 (2012).

26. C. M. Simon, R. Mercado, S. K. Schnell, B. Smit, M. Haranczyk, What are the best materialsto separate a xenon/krypton mixture? Chem. Mater. 27, 4459–4475 (2015).

27. Y. J. Colón, D. Fairen-Jimenez, C. E. Wilmer, R. Q. Snurr, High-throughput screening ofporous crystalline materials for hydrogen storage capacity near room temperature. J. Phys.Chem. C 118, 5383–5389 (2014).

28. P. Bai, M. Y. Jeon, L. Ren, C. Knight, M. W. Deem, M. Tsapatsis, J. I. Siepmann, Discovery ofoptimal zeolites for challenging separations and chemical transformations usingpredictive materials modeling. Nat. Commun. 6, 5912 (2015).

29. J. Goldsmith, A. G. Wong-Foy, M. J. Cafarella, D. J. Siegel, Theoretical limits of hydrogenstorage in metal–organic frameworks: Opportunities and trade-offs. Chem. Mater. 25,3373–3382 (2013).

30. S. O. Odoh, C. J. Cramer, D. G. Truhlar, L. Gagliardi, Quantum-chemical characterization ofthe properties and reactivities of metal-organic frameworks. Chem. Rev. 115, 6051–6111(2015).

31. Y. Bao, R. L. Martin, M. Haranczyk, M. W. Deem, In silico prediction of MOFs with highdeliverable capacity or internal surface area. Phys. Chem. Chem. Phys. 17, 11962–11973(2015).

32. Y. Bao, R. L. Martin, C. M. Simon, M. Haranczyk, B. Smit, M. W. Deem, In silico discovery ofhigh deliverable capacity metal organic frameworks. J. Phys. Chem. C 119, 186–195(2015).

33. M. Fernandez, P. G. Boyd, T. D. Daff, M. Z. Aghaji, T. K. Woo, Rapid and accurate machinelearning recognition of high performing metal organic frameworks for CO2 capture.J. Phys. Chem. Lett. 5, 3056–3060 (2014).

34. C. M. Breneman, K. B. Wiberg, Determining atom-centered monopoles from molecularelectrostatic potentials. The need for high sampling density in formamide conformationalanalysis. J. Comput. Chem. 11, 361–373 (1990).

35. C. E. Wilmer, K. C. Kim, R. Q. Snurr, An extended charge equilibration method. J. Phys.Chem. Lett. 3, 2506–2511 (2012).

8 of 9

SC I ENCE ADVANCES | R E S EARCH ART I C L E

on July 9, 2020http://advances.sciencem

ag.org/D

ownloaded from

36. X. Lin, I. Telepeni, A. J. Blake, A. Dailly, C. M. Brown, J. M. Simmons, M. Zoppi, G. S. Walker,K. M. Thomas, T. J. Mays, P. Hubberstey, N. R. Champness, M. Schroder, High capacityhydrogen adsorption in Cu(II) tetracarboxylate framework materials: The role of pore size,ligand functionalization, and exposed metal sites. J. Am. Chem. Soc. 131, 2159–2171(2009).

37. T. A. Makal, X. Wang, H.-C. Zhou, Tuning the moisture and thermal stability of metal–organic frameworks through incorporation of pendant hydrophobic groups. Cryst.Growth Des. 13, 4760–4768 (2013).

38. D. A. Gómez-Gualdrón, P. Z. Moghadam, J. T. Hupp, O. K. Farha, R. Q. Snurr, Application ofconsistency criteria to calculate BET areas of micro- and mesoporous metal–organicframeworks. J. Am. Chem. Soc. 138, 215–224 (2016).

39. J. Jia, X. Lin, C. Wilson, A. J. Blake, N. R. Champness, P. Hubberstey, G. Walker, E. J. Cussen,M. Schröder, Twelve-connected porous metal–organic frameworks with high H2

adsorption. Chem. Commun. 2007, 840–842 (2007).40. K. T. Leperi, R. Q. Snurr, F. You, Optimization of two-stage pressure/vacuum swing

adsorption with variable dehydration level for postcombustion carbon capture. Ind. Eng.Chem. Res. 55, 3338–3350 (2016).

41. G. E. Norman, V. S. Filinov, Investigations of phase transitions by a Monte Carlo method.High Temp. 7, 216–222 (1969).

42. D. Dubbeldam, A. Torres-Knoop, K. S. Walton, On the inner workings of Monte Carlocodes. Mol. Simulat. 39, 1253–1292 (2013).

43. D. Dubbeldam, S. Calero, D. E. Ellis, R. Q. Snurr, RASPA: Molecular simulation software foradsorption and diffusion in flexible nanoporous materials. Mol. Simul. 42, 81–101(2016).

44. B. L. Miller, D. E. Goldberg, Genetic algorithms, selection schemes, and the varying effectsof noise. Evol. Comput. 4, 113–131 (1996).

45. D. Peng, D. B. Robinson, A new two-constant equation of state. Ind. Eng. Chem. Fund. 15,59–64 (1976).

46. A. K. Rappe, C. J. Casewit, K. S. Colwell, W. A. Goddard III, W. M. Skiff, UFF, a full periodic-table force-field for molecular mechanics and molecular-dynamics simulations. J. Am.Chem. Soc. 114, 10024–10035 (1992).

47. J. J. Potoff, J. I. Siepmann, Vapor–liquid equilibria of mixtures containing alkanes, carbondioxide, and nitrogen. AIChE J. 47, 1676–1682 (2001).

48. F. Darkrim, D. Levesque, Monte Carlo simulations of hydrogen adsorption in single-walledcarbon nanotubes. J. Chem. Phys. 109, 4981–4984 (1998).

49. C. E. Wilmer, O. K. Farha, T. Yildirim, I. Eryazici, V. Krungleviciute, A. A. Sarjeant, R. Q. Snurr,J. T. Hupp, Gram-scale, high-yield synthesis of a robust metal–organic framework forstoring methane and other gases. Energy Environ. Sci. 6, 1158–1163 (2013).

50. Y. Peng, G. Srinivas, C. E. Wilmer, I. Eryazici, R. Q. Snurr, J. T. Hupp, T. Yildirim, O. K. Farha,Simultaneously high gravimetric and volumetric methane uptake characteristics of themetal–organic framework NU-111. Chem. Commun. 49, 2992–2994 (2013).

51. Q. Yang, C. Zhong, Molecular simulation of adsorption and diffusion of hydrogen inmetal–organic frameworks. J. Phys. Chem. B 109, 11862–11864 (2005).

52. K. A. Forrest, T. Pham, K. McLaughlin, J. L. Belof, A. C. Stern, M. J. Zaworotko, B. Space,Simulation of the mechanism of gas sorption in a metal–organic framework withopen metal sites: Molecular hydrogen in PCN-61. J. Phys. Chem. C 116, 15538–15549(2012).

53. B. J. Sikora, R. Winnegar, D. M. Proserpio, R. Q. Snurr, Textural properties of a largecollection of computationally constructed MOFs and zeolites. Micropor. Mesopor. Mat.186, 207–213 (2014).

54. Y.-S. Bae, R. Q. Snurr, Development and evaluation of porous materials for carbon dioxideseparation and capture. Angew. Chem. Int. Ed. Engl. 50, 11586–11596 (2011).

55. R. T. Yang, Adsorbents: Fundamentals and Applications (John Wiley & Sons Inc., 2003).56. K. Deb, A. Pratap, S. Argawal, T. Meyarivan, A fast and elitist multi-objective genetic

algorithm: NSGA-II. IEEE Trans. Evol. Comput. 6, 182–197 (2002).57. J. I. Feldblyum, M. Liu, D. W. Gidley, A. J. Matzger, Reconciling the discrepancies between

crystallographic porosity and guest access as exemplified by Zn-HKUST-1. J. Am. Chem.Soc. 133, 18257–18263 (2011).

58. K. Tan, N. Nijem, P. Canepa, Q. Gong, J. Li, T. Thonhauser, Y. J. Chabal, Stability andhydrolyzation of metal organic frameworks with paddle-wheel SBUs upon hydration.Chem. Mater. 24, 3153–3167 (2012).

59. M. K. Bhunia, J. T. Hughes, J. C. Fettinger, A. Navrotsky, Thermochemistry of paddle wheelMOFs: Cu-HKUST-1 and Zn-HKUST-1. Langmuir 29, 8140–8145 (2013).

Chung et al., Sci. Adv. 2016;2 : e1600909 14 October 2016

60. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman,G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li,H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai,T. Vreven, J. A. Montgomery Jr., J. E. Peralta, F. Ogliaro, M. J. Bearpark, J. Heyd,E. N. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari,A. P. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, N. J. Millam, M. Klene,J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma,V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels,Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian 09 (Gaussian Inc.,Wallingford CT, 2009).

61. A. D. Becke, A new mixing of Hartree-Fock and local density-functional theories. J. Chem.Phys. 98, 1372–1377 (1993).

62. C. Lee, W. Yang, R. G. Parr, Development of the Colle-Salvetti correlation-energy formulainto a functional of the electron-density. Phys. Rev. B 37, 785–789 (1988).

63. R. Krishnan, J. S. Binkley, R. Seeger, J. A. Pople, Self‐consistent molecular orbital methods.XX. A basis set for correlated wave functions. J. Chem. Phys. 72, 650–654 (1980).

64. M. J. Frisch, J. A. Pople, J. S. Binkley, Self-consistent molecular-orbital methods 25.Supplementary functions for Gaussian basis sets. J. Chem. Phys. 80, 3265–3269 (1984).

65. T. H. Dunning Jr., Gaussian basis sets for the atoms gallium through krypton. J. Chem.Phys. 66, 1382–1383 (1977).

66. K. S. Walton, D. S. Sholl, Predicting multicomponent adsorption: 50 years of the idealadsorbed solution theory. AIChE J. 61, 2757–2762 (2015).

Acknowledgments: We thank C. E. Wilmer for his assistance in the identification of the MOFgenes in the WLLFHS hMOF database. We also thank T. C. Wang (Northwestern University) andD. P. Broom (Hiden Isochema) for discussions on adsorption measurements. Funding: R.Q.S.acknowledges funding from the Division of Chemical Sciences, Geosciences, and Biosciences,Office of Basic Energy Sciences, U.S. Department of Energy, under award DE-FG02-12ER16362.F.Y. thanks the Global Climate and Energy Project for funding the work in section S4. J.T.H. andO.K.F. acknowledge support from the U.S. Department of Energy under award DE-FG02-08ER15967. J.F.S. thanks King Abdulaziz City for Science and Technology (KACST) andNorthwestern University (NU) for their continued support for this research, which is part of theJoint Center of Excellence in Integrated Nanosystems at KACST and at NU. Simulations werecarried out with computational resources from the high-performance computing system QUEST atNorthwestern University and the National Energy Research Scientific Computing Center high-performance computing systems of the U.S. Department of Energy. Author contributions: Y.G.C.,D.A.G.-G., and R.Q.S. conceived the research, designed and guided the computational work,and drafted the manuscript. Y.G.C. wrote the GA routine, performed GA robustness tests, andcarried out GCMC simulations for hMOFs and CoRE MOFs. D.A.G.-G. obtained the geneticinformation of hMOFs, obtained the reduced WLLHFS database, performed DFT calculations, andcarried out GCMC simulations for hMOFs and potential synthesis targets. K.T.L. carried out theprocess simulations. P.L. synthesized NOTT-101/OEt and VEXTUO and measured N2 isotherms.N.A.V. synthesized the TPTC linker. P.D. measured high-pressure CO2 and H2 isotherms. H.Z. carriedout IAST calculations. O.K.F., J.T.H., and J.F.S. supervised the experimental work. F.Y. supervisedthe macroscopic process simulation work. R.Q.S. supervised all computational work. All authorscontributed to the final version of the manuscript. Competing interests: J.T.H., O.K.F., and R.Q.S.have a financial interest in NuMat Technologies, a start-up company that is seeking tocommercialize MOFs. All other authors declare that they have no competing interests. Data andmaterials availability: All data needed to evaluate the conclusions of the paper are present inthe paper and/or the Supplementary Materials. Additional data related to this paper may berequested from the authors.

Submitted 27 April 2016Accepted 1 September 2016Published 14 October 201610.1126/sciadv.1600909

Citation: Y. G. Chung, D. A. Gómez-Gualdrón, P. Li, K. T. Leperi, P. Deria, H. Zhang,N. A. Vermeulen, J. F. Stoddart, F. You, J. T. Hupp, O. K. Farha, R. Q. Snurr, In silico discoveryof metal-organic frameworks for precombustion CO2 capture using a genetic algorithm. Sci.Adv. 2, e1600909 (2016).

9 of 9

genetic algorithm capture using a2In silico discovery of metal-organic frameworks for precombustion CO

Vermeulen, J. Fraser Stoddart, Fengqi You, Joseph T. Hupp, Omar K. Farha and Randall Q. SnurrYongchul G. Chung, Diego A. Gómez-Gualdrón, Peng Li, Karson T. Leperi, Pravas Deria, Hongda Zhang, Nicolaas A.

DOI: 10.1126/sciadv.1600909 (10), e1600909.2Sci Adv 

ARTICLE TOOLS http://advances.sciencemag.org/content/2/10/e1600909

MATERIALSSUPPLEMENTARY http://advances.sciencemag.org/content/suppl/2016/10/11/2.10.e1600909.DC1

REFERENCES

http://advances.sciencemag.org/content/2/10/e1600909#BIBLThis article cites 60 articles, 3 of which you can access for free

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Terms of ServiceUse of this article is subject to the

is a registered trademark of AAAS.Science AdvancesYork Avenue NW, Washington, DC 20005. The title (ISSN 2375-2548) is published by the American Association for the Advancement of Science, 1200 NewScience Advances

Copyright © 2016, The Authors

on July 9, 2020http://advances.sciencem

ag.org/D

ownloaded from