Computationally-Guided Synthesis of the 8-Ring Zeolite AEITopics in Catalysis

About

Authors
Joel E. Schmidt, Michael W. Deem, Christopher Lew, Tracy M. Davis
Year
2015
DOI
10.1007/s11244-015-0381-1
Subject
Chemistry (all) / Catalysis

Text

ORIGINAL PAPER

Computationally-Guided Synthesis of the 8-Ring Zeolite AEI

Joel E. Schmidt1,2 • Michael W. Deem3 • Christopher Lew1 • Tracy M. Davis1

Published online: 2 April 2015  Springer Science+Business Media New York 2015

Abstract A computational method capable of predicting chemically-synthesizable organic structure directing agents (OSDAs) for targeted microporous material frameworks has been applied to the zeolite SSZ-39 (AEI framework topology). The top predicted OSDA has been found to have a more favorable stabilization energy than any of the

OSDAs previously reported to form SSZ-39. This result was verified experimentally, demonstrating that this computational method is capable of predicting successful

OSDAs for zeolite synthesis mixtures containing a large number of inorganic variables such as heteroatoms, inorganic cations, hydroxide media and high water content.

This is a significant improvement over the first experimental validation of this computational method.

Keywords Computational guidance  Zeolite synthesis 

SSZ-39 (AEI)  Structure directing agents 1 Introduction

Microporous materials (aluminosilicates are denoted zeolites) are used in a variety of processes including separations, adsorption, catalysis and ion exchange, where their utility depends on the unique properties of each material.

Some of these properties include the pore size and dimensionality, material composition, hydrothermal stability and cost [1–3]. Much research is currently directed towards creating new material frameworks and compositions in combination with lowering the cost of known materials for targeted applications [4]. Molecular sieves are commonly synthesized in the presence of an organic structure directing agent (OSDA), which strongly influences zeolite phase selectivity. In recent decades, there has been much investigation into OSDAs, leading to the discovery of many new frameworks and compositions [5–7]. In large part, this investigation has been based on trial and error approaches that are directed by a number of guiding principles [8].

More recently, computational guidance has become increasingly reliable [4]. Past computational approaches relied on scoring the interaction energies of the OSDA and framework, and commonly leveraged methods developed for the pharmaceutical industry [9–11]. These methods suffered from numerous shortcomings, most notably that they produced OSDAs that were often difficult or impossible to synthesize [11].

A new method to predict chemically synthesizable

OSDAs for crystalline molecular sieves was recently reported, and resulted in the experimentally confirmed synthesis of pure-silica STW (HPM-1)1 using one of the computationally generated OSDAs [4, 12]. In this preliminary case, a pure-silica, fluoride-mediated inorganic synthesis system was selected in an effort to limit the number of variables influencing the synthesis product.

However, commercial microporous materials are typically produced in hydroxide-mediated syntheses in the presence of heteroatoms (e.g., aluminum, boron, titanium, etc.) and & Michael W. Deem mwdeem@rice.edu 1 Chevron Energy Technology Company, Richmond,

CA 94802, USA 2 Chemical Engineering, California Institute of Technology,

Pasadena, CA 91125, USA 3 Departments of Bioengineering and Physics & Astronomy,

Rice University, Houston, TX 77005, USA 1 Three-letter framework type codes (boldface capital letters) for all zeolites mentioned in the text are given in parentheses. 123

Top Catal (2015) 58:410–415

DOI 10.1007/s11244-015-0381-1 inorganic cations (e.g., sodium, potassium, etc.), which can also influence product selectivity [13–17]. These inorganic synthesis parameters introduce variables that add complexity to computational methods and can hamper the ability to predict OSDAs for a given framework.

Herein, we report the computational prediction of an

OSDA for zeolite AEI (SSZ-39), and its experimental validation in a hydroxide-mediated synthesis mixture containing both aluminum and inorganic cations. This refined computational method is able to predict OSDAs that are soluble in the synthesis mixture, stable under reaction conditions (i.e., high temperature and high pH) and that can be produced using known chemical reactions [4, 12]. This is accomplished by using an evolutionary growth process to create organic molecules that fit inside the desired framework with a favorable van der Waals’ interaction energy.

These organic compounds are evolved by computational synthesis from simpler molecules that are available from common chemical suppliers. Computational synthesis is carried out via known chemical reactions, most notably the

Menshutkin reaction with the goal of producing quaternary amines [4, 18].

The zeolite AEI has a three-dimensional system of pores with 8-membered rings (8MRs) that open into larger cages, and has been proposed for applications including NOX reduction and the methanol-to-olefins reaction [19]. Cagebased microporous materials with 8MRs are useful for these applications as their pore size limits coking, provides high hydrothermal stability and leads to desired products.

AEI has been synthesized previously using a wide range of piperidinium based OSDAs (Fig. 2), and under a variety of inorganic synthesis conditions [19–22]. In order to favor the formation of AEI, inorganic conditions similar to those used previously in the synthesis of AEI were selected. 2 Experimental 2.1 OSDA Synthesis

N-ethyl-N-methyl-2,2,6,6-tetramethylpiperidinium was synthesized by reacting 2,2,6,6-tetramethylpiperidine with a stoichiometric amount of iodoethane and an excess of

KHCO3 in acetonitrile at reflux overnight. The KHCO3 was filtered and then the solvent was removed using rotary evaporation. Next, the product tertiary amine was reacted with a large excess of methyl iodide (as both the solvent and reactant) at reflux for several days. The excess methyl iodide was removed under vacuum and the product was washed with a small amount of hot acetone, followed by washing with ethyl ether (Caution: Toxic vapors present, use appropriate precautions). Product characterization: 13C