Deciphering the Relations between Pore Structure and Adsorption

Settings. My Account · Help; Full Site. Hide Menu Back. Please wait while the data is being loaded.. Hide Menu Back Journals ...
0 downloads 0 Views 2MB Size
Download PDF
Subscriber access provided by University of Rochester | River Campus & Miner Libraries

Communication

Deciphering the Relations between Pore Structure and Adsorption Behavior in MOFs: Unexpected Lessons from Argon Adsorption on Cu-BTC Silvio Dantas, Lev Sarkisov, and Alexander V. Neimark J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.9b00906 • Publication Date (Web): 15 May 2019 Downloaded from http://pubs.acs.org on May 16, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Deciphering the Relations between Pore Structure and Adsorption Behavior in MOFs: Unexpected Lessons from Argon Adsorption on Cu-BTC Silvio Dantas1, Lev Sarkisov2, and Alexander V. Neimark1,* 1- Department of Chemical and Biochemical Engineering, Rutgers, The State University of New Jersey, 98 Brett Road, Piscataway, NJ 08854, USA 2- Institute for Materials and Processes, School of Engineering, The University of Edinburgh, Edinburgh EH9 3JL, United Kingdom

  Supporting  Information  Placeholder ABSTRACT: Consistent adsorption characterization of metalorganic frameworks (MOFs) is imperative for their wider adoption in industry and practical applications. Current approaches are based on the conventional intuitive representation of MOF pore space as a regular network of pore compartments (cages and channels), adsorption in which occurs independently according to their geometric dimensions. Here, we demonstrate that this conventional approach is unable to describe even qualitatively the shape of Ar adsorption isotherms on hydrated and dehydrated Cu-BTC structures, one of the most well-known MOF materials. A combination of geometric characterization of MOF crystallographic structure, molecular simulation and virtual visualization of the adsorption process reveals that the filling of the adjacent pore compartments proceeds in parallel in a complex cooperative fashion. The proposed synergistic approach helps understanding the relations between pore structure geometric and chemical features and adsorption behavior laying down a foundation for improved methods for MOF characterization.

Metal-organic frameworks (MOFs) and related classes of porous crystals have dominated literature on adsorption studies in the last ten years.1-12 The promise these materials hold for various applications stems from their exceptional adsorption characteristics and wide range of textural properties. The important issue that needs to be addressed is consistent structural characterization of MOFs. Despite being crystalline materials in theory, real samples often deviate from the ideal structure due to residual solvent and reactants in the structure, missing linkers, and partially collapsed regions of the framework. The most instructive technique that is routinely used for MOF characterization is physical adsorption of nitrogen or argon at their normal boiling temperatures.13 Measured adsorption isotherms contain important information about the sample porosity, surface area and pore sizes provided a clear link exists between adsorption behavior and pore structure. It is assumed that, as the gas pressure increases, the adsorbate fills the pores in the order of their sizes, so that the pore size distribution (PSD) can be derived using a particular kernel of reference adsorption isotherms in model pores of different sizes.13 However, generic adsorption characterization methods rely on simplified pore structure models (independent cylindrical, spherical, or slit-shaped pores), which neglect real complex pore

morphology of MOF crystals and produce, in some cases, PSDs inconsistent with the crystallographic parameters.14-15 Molecular simulations and computational structure characterization tools can be of immense help to remedy this situation, since in silico, unlike in experiments, one can investigate properties of an ideal, defectfree crystal, generate adsorption isotherms using well-developed forcefields and directly obtain information on the pore morphology. Superficially, one may think that for the case of an idealized, perfect MOF, development of relations between structural characteristics and adsorption isotherms should be a straightforward task. However, application of this approach to a well-known Cu-BTC structure led to several challenges and unexpected lessons which we would like to share and explore in more detail, as they will be instrumental in further efforts to develop reliable characterization approaches. Since its first reported synthesis in 1999 by Chui et al.16, CuBTC (aka HKUST-1) has been extensively investigated using both experimental and theoretical approaches. Noteworthy, the first molecular simulation study of adsorption in MOFs was performed on Cu-BTC.17 The basic building block of Cu-BTC is formed by Cu atoms coordinated by benzene-1,3,5-tricarboxilate (BTC) ligands. Adsorption and catalytic properties are mainly determined by accessibility of unsaturated metal sites that might be easily hydrated or occupied by impurities or ligands. One of the goals of structural characterization is to distinguish between the samples with different levels of hydration. In order to get a better understanding of the effects of hydration on structural and adsorption properties, we consider two reference Cu-BTC structures, one with fully saturated Cu sites (hydrated18) and another with unsaturated sites (dehydrated19). The cubic unit cell of Cu-BTC can be compartmentalized into three types of cages denoted here as S1, L2, and L3.20 The pore structure represents a 3D network of alternating larger spheroidal cages L2 and L3 with tetrahedral-shaped small S1 cages connected to L3 cages. In Figure 1A and C, we demonstrate the sizes of these cages by using probe spheres inscribed inside each cage, without overlapping with any framework atoms, and serving as yardsticks: purple for S1, green for L2, and orange for L3. For dehydrated structure, the defined cage diameters are: S1 – 0.45 nm, L2 – 1.05 nm and L3 – 1.29 nm; as such, S1 falls in the category of ultramicropores (D < 0.7 nm), while L2 and L3 – supramicropores (0.7 nm < D < 2 nm).13

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 5

  Figure 1. Geometrical characterization of the pore structure in the dehydrated and hydrated Cu-BTC crystals. The framework atoms are set in their crystallographic positions. A and C: Compartmentalization of the unit cell into 3 types of cages shown by inserted spheres: S1 (purple), L2 (green) and L3 (orange). B and D: Assignment of the pore diameters in geometric analysis by largest inserted spherical probes shown in different colors; E: geometric PSDs (differential and cumulative) of the dehydrated (top plot) and hydrated (bottom plot) structures; F: schematic representation of the pore space connecting L2 and L3 cages in the hydrated structure. The hydrated structure, Figure 1C, has a more complex shape due to the presence of the extra 12 oxygen atoms protruding into the cage. These extra atoms significantly decrease the size of the sphere that can fit within the L3 cage (0.85 nm). Several methods for geometric analysis were developed to calculate the pore volume, surface area, and PSD in model porous materials, both disordered (activated carbons, polymers) and crystalline (zeolites, MOFs, ZIFs, etc.).21-24 Here, we employ the Poreblazer code,24 described in detail in SI. It is based on the Connolly surface method25 originally described by Gelb and Gubbins in the context of model porous glasses.26 The algorithm assigns a point within the available pore space to a pore of a particular size with respect to the maximum diameter of a spherical probe as shown in Figure 1B and D. The geometric PSD analysis brings to light the striking effect of the hydration on the pore shape and connectivity. From the visual inspection of the pore structure compartmentalization presented in Figure 1A and C, one might expect to see three distinct peaks on the differential PSD corresponding to the three cage types. Indeed, this is what is observed for the PSD of the dehydrated structure (Figure 1E top). Each of these peaks corresponds to the pore spaces represented in Figure 1B by spheres of characteristic size and color (0.45 nm – purple, 1.0 nm - green, 1.29 nm - orange). Similar distributions are found in the literature20, 27 using the geometrical methods applied in RASPA19 and Poreblazer24. The green spherical caps seen on top of the orange sphere represent the regions in the L3 cage not embraced by the largest fitted sphere due to not ideally spherical pore shape. Nevertheless, the volume of each cage type can be unambiguously determined from the geometrical PSD with the

total pore volume distributed between S1 (15%), L2 (42.5%) and L3 (42.5%) cages. In contrast, the PSD of the hydrated version (Figure 1E bottom) has four, rather than three, pronounced peaks that should indicate the existence of four pore types of distinct sizes. This effect is a result of the protrusion of oxygen atoms into L3 cage, causing the geometric algorithm to interpret the shape of L3 cage as being comprised of smaller pores. Figure 1D illustrates (additional schematics are provided in SI) the nature of this effect. First, the maximum size of the fitted sphere in L3 cage reduces from 1.29 nm in dehydrated structure (orange in Figure 1B) to 0.85 nm (yellow in Figure 1D). Secondly, the remainder of L3 cage space is filled by additional blue (0.65 nm), and purple (0.45 nm) spheres. The shape of L2 cage becomes also more complex with the 1.01 nm cage in the center (green) and smaller peripheral compartments. The complex shape of the connected L2 and L3 cages is schematically presented in Figure 1F as a corrugated channel comprised of overlapping purple, yellow, blue and green spheres. Noteworthy, the blue spheres between the L2 and L3 cages correspond to the window spaces that were assigned to L3 cages in the dehydrated structure. Finally, smaller purple spheres represent not only S1 cages, but also the peripheral regions of L2 and L3 cages, making the correspondent peak bigger than the one seen in the PSD for the dehydrated sample. The characteristic differences in the pore geometry should be reflected in the adsorption behavior, specifically by the characteristic steps and inflection points on the adsorption isotherms. Assuming the Cu-BTC structure as consisting of three cage types, one could expect a three-step adsorption isotherm for Ar at 87K, with the first step at low pressures corresponding to the filling of ultramicropores (S1) and two subsequent distinct

ACS Paragon Plus Environment

Page 3 of 5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society steps corresponding to the filling of supramicropores L2 and L3. Such three-step isotherm would correspond to the geometrical

PSD with

  Figure 2. Molecular simulations of adsorption. A: Adsorption isotherms of Ar at 87.3 K. B and C: Adsorption potential maps with isosurfaces at -11 (red), -8 (light-green) and -4 (purple) kJ/mol. three distinct peaks such as the one for the dehydrated structure. On the other side, the PSD distribution with four peaks found for the hydrated Cu-BTC makes one expect a four-step adsorption isotherm reflecting the subsequent filling of four groups of pores. Surprisingly, the adsorption isotherms obtained using grand canonical Monte Carlo (GCMC) simulations (computational details are given in SI) do not conform to any of these expected scenarios (Figure 2A). Aside from the filling of the utramicropore cages S1 at 10-5 to 10-3 P/P0, the dehydrated sample features essentially one step in the region between 10-3 P/P0 and 10-2 P/P0. This step is initially very sharp followed by a gradual increase before reaching saturation levels. On the other hand, in the hydrated sample in the same pressure range there are two very distinct steps. Additional insight is brought by the adsorption potential maps (Figure 2B and C) with isosurfaces for three characteristic potential energies: -11 (red), -8 (light-green) and -4 (purple) kJ/mol. A side-by-side comparison of the isosurfaces shows the differences of the key adsorption sites in the structures considered. The first isosurface, with energy of -11 kJ/mol (shown in red), is larger for the hydrated structure, which explains the higher adsorption in ultramicropores, which includes S1 cages, at pressures below 10-3 P/P0. The second isosurface (-8 kJ/mol, light-green) is also larger for the hydrated structure, which is, most likely, the cause of slightly earlier Ar condensation in L2 cages. Lastly, the third isosurface (-4 kJ/mol, purple) is smaller for the hydrated structure, which could explain delayed condensation in L3 pores in the hydrated structure in contrast to the dehydrated one. To understand the adsorption mechanisms in these structures, it is instructive to visualize the distributions of molecules along the isotherms. Snapshots in Figure 3 show that although the

geometric PSD analysis interprets L3 cage as a system of smaller pores in the hydrated sample (even smaller than L2), the sequence of complete pore filling in both structures is the same: S1, then L2, then L3 (detailed description of Figure 3 is in SI). The significant difference is the fact that, in the dehydrated sample, the steps associated with condensation of argon in L2 and L3 are not distinct. While the step associated with the filling L2 cages is quite sharp, filling of the L3 cages proceeds gradually due to relatively large windows between these cages. From this perspective, the supramicropore structure of the dehydrated sample can be thought of as a network of corrugated channels rather than a network of distinct cages (L2 and L3) connected by narrower windows. For the hydrated sample, apparently, the adsorption in the windows between L2 and L3 cages is delayed and, despite the geometrical size of L3 cage is smaller due to the protruding oxygen atoms, filling of L3 cage is reflected as a separate step on the isotherm. In conclusion, a synergistic combination of geometrical characterization of the pore structure, Monte Carlo simulation of adsorption isotherms and molecular visualizations is needed to elucidate adsorption mechanisms and to establish non-trivial relationships between geometrical and chemical features of a MOF and specifics of adsorption behavior. Our results give impetus to developing a novel approach to MOF characterization based on the comparison of the experimental isotherms with their in-silico simulated counterparts to reveal the similarity and the difference between the real samples and the ideal crystals. This will allow to identify the fraction of MOF crystals in the sample, the degree of hydration, and possible crystal defects due to residual solvents and other imperfections.

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 5

  Figure 3. Simulated adsorption isotherms of Ar at 87.3 K in dehydrated (A) and hydrated (E) structures of Cu-BTC and respective snapshots at the pressure points 4 x 10-4 (B and F), 3.2 x 10-3 (C and G), 1 x 10-1 (D and H). Framework atoms shown: C – grey, O – red, Cu – tan. To distinguish the distribution of adsorbate in difference cages, Ar molecules are shown in purple within S1, green within L2, and orange within L3. B and F – complete filling of S1; C and G – complete filling of L2 and partial filling of L3; D and H – complete filling.

ASSOCIATED CONTENT Supporting Information Poreblazer details, Monte Carlo simulation details and Details of visualization of adsorption behavior. Computational files: Adsorption Isotherms, Geometrical Pore Size Distributions, Potential Maps. Input-output files and codes are Supplementary Files.

AUTHOR INFORMATION Corresponding Author [email protected]

ACKNOWLEDGEMENT This work is supported in parts by the National Science Foundation (grant No 1834339) and by the CAPES Foundation through the Science Without Borders program.

REFERENCES 1. Forse, A. C.; Milner, P. J.; Lee, J. H.; Redfearn, H. N.; Oktawiec, J.; Siegelman, R. L.; Martell, J. D.; Dinakar, B.; Porter-Zasada, L. B.; Gonzalez, M. I.; Neaton, J. B.; Long, J. R.; Reimer, J. A., Elucidating CO2 Chemisorption in Diamine-Appended Metal-Organic Frameworks. J Am Chem Soc 2018, 140 (51), 18016-18031. 2. Li, J.; Wang, X. X.; Zhao, G. X.; Chen, C. L.; Chai, Z. F.; Alsaedi, A.; Hayat, T.; Wang, X. K., Metal-organic framework-based materials: superior adsorbents for the capture of toxic and radioactive metal ions. Chemical Society Reviews 2018, 47 (7), 2322-2356. 3. Kobielska, P. A.; Howarth, A. J.; Farha, O. K.; Nayak, S., Metal-organic frameworks for heavy metal removal from water. Coordination Chemistry Reviews 2018, 358, 92-107. 4. Kalmutzki, M. J.; Diercks, C. S.; Yaghi, O. M., Metal-Organic Frameworks for Water Harvesting from Air. Adv Mater 2018, 30 (37), e1704304.

5. Hobday, C. L.; Woodall, C. H.; Lennox, M. J.; Frost, M.; Kamenev, K.; Duren, T.; Morrison, C. A.; Moggach, S. A., Understanding the adsorption process in ZIF-8 using high pressure crystallography and computational modelling. Nat Commun 2018, 9 (1), 1429. 6. Shen, V. K.; Siderius, D. W.; Mahynski, N. A., Molecular simulation of capillary phase transitions in flexible porous materials. J Chem Phys 2018, 148 (12), 124115. 7. Dang, S.; Zhu, Q. L.; Xu, Q., Nanomaterials derived from metal-organic frameworks. Nature Reviews Materials 2017, 3 (1), 1707517088. 8. Lustig, W. P.; Mukherjee, S.; Rudd, N. D.; Desai, A. V.; Li, J.; Ghosh, S. K., Metal-organic frameworks: functional luminescent and photonic materials for sensing applications. Chemical Society Reviews 2017, 46 (11), 3242-3285. 9. Howarth, A. J.; Peters, A. W.; Vermeulen, N. A.; Wang, T. C.; Hupp, J. T.; Farha, O. K., Best Practices for the Synthesis, Activation, and Characterization of Metal-Organic Frameworks. Chemistry of Materials 2017, 29 (1), 26-39. 10. Choi, K. M.; Kim, D.; Rungtaweevoranit, B.; Trickett, C. A.; Barmanbek, J. T. D.; Alshammari, A. S.; Yang, P. D.; Yaghi, O. M., Plasmon-Enhanced PhotoCatalytic CO2 Conversion within Metal Organic Frameworks under Visible Light. J Am Chem Soc 2017, 139 (1), 356-362. 11. Demir, H.; Walton, K. S.; Sholl, D. S., Computational Screening of Functionalized UiO-66 Materials for Selective Contaminant Removal from Air. Journal of Physical Chemistry C 2017, 121 (37), 20396-20406. 12. Bien, C. E.; Chen, K. K.; Chien, S.-C.; Reiner, B. R.; Lin, L.C.; Wade, C. R.; Ho, W. S. W., Bioinspired Metal–Organic Framework for Trace CO2 Capture. J Am Chem Soc 2018, 140 (40), 12662-12666. 13. Thommes, M.; Kaneko, K.; Neimark, A. V.; Olivier, J. P.; Rodriguez-Reinoso, F.; Rouquerol, J.; Sing, K. S. W., Physisorption of gases, with special reference to the evaluation of surface area and pore size distribution (IUPAC Technical Report). Pure Appl Chem 2015, 87 (910), 1051-1069. 14. Liang, X.; Zhang, Y.; Qu, Y. Y.; Han, Y.; Wang, X. L.; Cheng, A.; Duan, A. J.; Zhao, L., Synthesis of HKUST-1 and zeolite beta composites for deep desulfurization of model gasoline. Rsc Adv 2018, 8 (25), 13750-13754. 15. Beckner, M.; Dailly, A., Adsorption Enthalpy Calculations of Hydrogen Adsorption at Ambient Temperature and Pressures Exceeding 300 bar. American Journal of Analytical Chemistry 2013, Vol.04No.10, 9.

ACS Paragon Plus Environment

Page 5 of 5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society 16. Chui, S. S. Y.; Lo, S. M. F.; Charmant, J. P. H.; Orpen, A. G.; Williams, I. D., A chemically functionalizable nanoporous material [Cu3(TMA)(2)(H2O)(3)](n). Science 1999, 283 (5405), 1148-1150. 17. Vishnyakov, A.; Ravikovitch, P. I.; Neimark, A. V.; Bülow, M.; Wang, Q. M., Nanopore Structure and Sorption Properties of Cu−BTC Metal−Organic Framework. Nano Letters 2003, 3 (6), 713-718. 18. Yakovenko, A. A.; Reibenspies, J. H.; Bhuvanesh, N.; Zhou, H. C., Generation and applications of structure envelopes for porous metal-organic frameworks. Journal of Applied Crystallography 2013, 46, 346-353. 19. Dubbeldam, D.; Calero, S.; Ellis, D. E.; Snurr, R. Q., RASPA: molecular simulation software for adsorption and diffusion in flexible nanoporous materials. Molecular Simulation 2016, 42 (2), 81-101. 20. Getzschmann, J.; Senkovska, I.; Wallacher, D.; Tovar, M.; Fairen-Jimenez, D.; Duren, T.; van Baten, J. M.; Krishna, R.; Kaskel, S., Methane storage mechanism in the metal-organic framework Cu3(btc)(2): An in situ neutron diffraction study. Microporous and Mesoporous Materials 2010, 136 (1-3), 50-58. 21. Martin, R. L.; Haranczyk, M., Construction and Characterization of Structure Models of Crystalline Porous Polymers. Crystal Growth & Design 2014, 14 (5), 2431-2440.

22. Pinheiro, M.; Martin, R. L.; Rycroft, C. H.; Jones, A.; Iglesia, E.; Haranczyk, M., Characterization and comparison of pore landscapes in crystalline porous materials. J Mol Graph Model 2013, 44, 208-219. 23. First, E. L.; Floudas, C. A., MOFomics: Computational pore characterization of metal–organic frameworks. Microporous and Mesoporous Materials 2013, 165, 32-39. 24. Sarkisov, L.; Harrison, A., Computational structure characterisation tools in application to ordered and disordered porous materials. Molecular Simulation 2011, 37 (15), 1248-1257. 25. Connolly, M. L., Analytical Molecular-Surface Calculation. Journal of Applied Crystallography 1983, 16 (Oct), 548-558. 26. Gelb, L. D.; Gubbins, K. E., Characterization of Porous Glasses:   Simulation Models, Adsorption Isotherms, and the Brunauer−Emmett−Teller Analysis Method. Langmuir 1998, 14 (8), 2097-2111. 27. Wu, Y.; Liu, D. F.; Chen, H. Y.; Qian, Y.; Xi, H. X.; Xia, Q. B., Enhancement effect of lithium-doping functionalization on methanol adsorption in copper-based metal-organic framework. Chem Eng Sci 2015, 123, 1-10.

Table of Contents

ACS Paragon Plus Environment