Alkane Separation DOI: 10.1002/anie.201205040 Computer-Assisted Screening of Ordered Crystalline Nanoporous Adsorbents for Separation of Alkane Isomers** David Dubbeldam,* Rajamani Krishna, Sofía Calero, and Ahmet Özgür Yazaydın The separation of linear, mono-branched, and di-branched isomers of alkanes is of significant importance in the petrochemical industry. For example, the di-branched alkanes in the 5–7 carbon number range are preferred components of high-octane gasoline. Their selective removal from the other isomers produced in an alkane isomerization reactor can be achieved using ordered crystalline nanoporous materials, such as zeolites, metal–organic frameworks (MOFs), covalent organic frameworks (COFs), and zeolitic imidazolate frame- works (ZIFs), by exploiting subtle differences in molecular configurations. Literally, several thousands of such materials have been synthesized, making the choice of adsorbent a daunting task. Our approach is to carry out molecular simulations on a pre-screened list of more than 100 nano- porous structures. Our screening methodology demonstrates that ZIF-77, whose synthesis was reported in 2008, [1] has significantly higher selectivities, by about two orders of magnitude, over other materials that are currently used and for which patents have been issued. In the petroleum industry, catalytic isomerization is used to convert linear alkanes into their branched isomers. [2] Isomerization processes generate a mixture of isomers that usually require separation and recycling of the non-isomer- ized components. For example, the effluent of a paraffin isomerization reactor may contain normal alkanes, mono- methyl alkanes and di-methyl alkanes. Traditionally, only the normal alkanes would be separated from the mixture by adsorption (e.g. using the LTA (Linde TypeA) type sieve which only adsorbs linear alkanes) and recycled to the isomerization reactor, and any mono-methyl alkanes would be collected with the di-methyl alkanes as product. However, it is the di-methyl alkanes that are the most desired because they have the highest octane numbers. Therefore, a more efficient approach would be to adsorptively separate only the di-methyl alkanes as product and recycle the normal and the mono-methyl alkanes to the isomerization reactor, [3, 4] sche- matically depicted in Figure 1. Ordered crystalline porous materials, such as zeolites, offer the potential for selective adsorption exploiting differences in molecular configurations. Zeolites are readily available, very stable, and cheap. The zeolite should have the right combination of high adsorption selectivity, combined with adequate capacity for use in traditionally used fixed-bed devices. CFI zeolite [5] and ATS zeolite [6] have been suggested for use in this separation in two patents. The undisclosed material in the patents [7, 8] is most likely to be MFI-type zeolite. [9] It is of vital importance to improve the energy efficiency of this process to generate more and cleaner gasoline from every barrel of oil. There has been a recent upsurge in next-generation nanoporous crystalline materials, such as MOFs, [10–15] COFs, [16, 17] and ZIFs. [18] There are almost unlimited structural possibilities because of the wide variety of combinations of metal atoms, organic linker molecules, and the building blocks used in self-assembly during synthesis. Each year the synthesis of several hundred new structures is reported and this has created the need to screen these efficiently. There is also a need to understand structure–property relations, such as the mechanisms of alkane separation as a function of shape and size of the pore system. Molecular simulations have suffi- ciently advanced in both speed and accuracy to allow rapid evaluation of (hypothetical) structures for storage and/or separation devices. [19, 20] The first step in our screening procedure is to pre-select, for further examination, only those nanoporous structures that have pore sizes that are large enough to accommodate alkanes molecules with seven or less carbon atoms. After this pre-screening, we performed a complete molecular simulation study on selected structures. The simulation model describes the system in full atomistic detail and is explained in detail in the Supporting Information (Section 1 on modeling and validation of the simulations). The force field we have used produces results in very good agreement with available experimental data and has good predictive capability (see Section 1 Supporting Information). The over one hundred structures we selected from zeolites, MOFs, COFs, and ZIFs provide a broad sample of available nanoporous materials. For each structure we have computed the nC6-2MP-3MP-22DMB-23DMB (nC6 = n-hexane, 2MP = 2-methylpentane, 3MP = 3-methylpentane, 22DMB = 2,2-dimetylbutane, 23DMB = 2,3-dimethylbutane) and nC7- 2MH-3MH-22DMP-23DMP (nC7 = n-heptane, 2MH = 2- methylhexane, 3MH = 3-metylhexane, 22DMP = 2,2-dime- thylpentane, 23DMP = 2,3-dimethylpentane) single compo- [*] Dr. D. Dubbeldam, Prof. Dr. R. Krishna Van’t Hoff Institute for Molecular Sciences, University of Amster- dam Science Park 904, 1098 XH Amsterdam (The Netherlands) E-mail: D.Dubbeldam@uva.nl Homepage: http://molsim.chem.uva.nl Prof. Dr. S. Calero Departamento de Sistemas Físicos, Químicos y Naturales, Uni- versidad Pablo de Olavide, Sevilla (Spain) Dr. A. Ö. Yazaydın Department of Chemical Engineering, University of Surrey (UK) [**] This work is supported by the Netherlands Research Council for Chemical Sciences (NWO/CW) through a VIDI grant (D.D.) and by the European Research Council through an ERC Starting Grant (S.C.). A.O.Y. acknowledges a Marie Curie International Reintegra- tion Grant from the European Commission. Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/anie.201205040. A ngewandte Chemi e 11867 Angew. Chem. Int. Ed. 2012, 51, 11867 –11871  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim