Scholarly article on topic 'Catalysis by Design: Well-Defined Single-Site Heterogeneous Catalysts'

Catalysis by Design: Well-Defined Single-Site Heterogeneous Catalysts Academic research paper on "Chemical sciences"

0
0
Share paper
Academic journal
Accounts of Chemical Research
OECD Field of science
Keywords
{""}

Academic research paper on topic "Catalysis by Design: Well-Defined Single-Site Heterogeneous Catalysts"

ACCOUNTS

pubs.acs.org/accounts

—of chemical research—

Catalysis by Design: Well-Defined Single-Site Heterogeneous Catalysts

Jeremie D. A. Pelletier and Jean-Marie Basset*

KAUST Catalysis Center, King Abdullah University of Science and Technology, Thuwal 23955-6900, Saudi Arabia

Article

CONSPECTUS: Heterogeneous catalysis, a field important industrially and scientifically, is increasingly seeking and refining strategies to render itself more predictable. The main issue is due to the nature and the population of catalytically active sites. Their number is generally low to very low, their "acid strengths" or " redox properties" are not homogeneous, and the material may display related yet inactive sites on the same material. In many heterogeneous catalysts, the discovery of a structure—activity reationship is at best challenging. One possible solution is to generate single-site catalysts in which most, if not all, of the sites are structurally identical. Within this context and using the right tools, the catalyst structure can be designed and well-defined, to reach a molecular understanding. It is then feasible to understand the structure—activity relationship and to develop predictable heterogeneous catalysis. Single-site well-defined heterogeneous catalysts can be prepared using concepts and tools of surface organometallic chemistry (SOMC). This approach operates by reacting organometallic compounds with surfaces of highly divided oxides (or of metal nanoparticles). This strategy has a solid track record to reveal structure—activity relationship to the extent that it is becoming now quite predictable. Almost all elements of the periodical table have been grafted on surfaces of oxides (from simple oxides such as silica or alumina to more sophisticated materials regarding composition or porosity).

Considering catalytic hydrocarbon transformations, heterogeneous catalysis outcome may now be predicted based on existing mechanistic proposals and the rules of molecular chemistry (organometallic, organic) associated with some concepts of surface sciences. A thorough characterization of the grafted metal centers must be carried out using tools spanning from molecular organometallic or surface chemistry. By selection of the metal, its ligand set, and the support taken as a X, L ligands in the Green formalism, the catalyst can be designed and generated by grafting the organometallic precursor containing the functional group(s) suitable to target a given transformation (surface organometallic fragments (SOMF)). The choice of these SOMF is based on the elementary steps known in molecular chemistry applied to the desired reaction. The coordination sphere necessary for any catalytic reaction involving paraffins, olefins, and alkynes also can thus be predicted. Only their most complete understanding can allow development of catalytic reactions with the highest possible selectivity, activity, and lifetime. This Account will examine the results of SOMC for hydrocarbon transformations on oxide surfaces bearing metals of group 4—6. The silica-supported catalysts are exhibiting remarkable performances for Ziegler—Natta polymerization and depolymerization, low temperature hydrogenolysis of alkanes and waxes, metathesis of alkanes and cycloalkanes, olefins metathesis, and related reactions. In the case of reactions involving molecules that do not contain carbon (water—gas shift, NH3 synthesis, etc.) this single site approach is also valid but will be considered in a later review.

■ INTRODUCTION

The chemical challenges facing the planet in terms of energy and environment increase the pressure on the scientific community to find alternative catalytic solutions. In this changing landscape, catalysis remains pivotal since, in principle, it can improve the efficiency, the rate, or the selectivity of any chemical reaction. Heterogeneous catalysts are considered as the ideal choice for a broad range of chemical industrial transformations due to their robustness and operational practicability. Homogeneous and, to a greater extent, enzymatic catalysis are currently less common in industry mostly because of their fragility, short lifetime, thermal instability, and difficulty of separation after reaction.

Heterogeneous catalysis, preferred by industry, is nonetheless far from perfect: high temperatures lead to energetic over-consumption, low selectivity raises environmental concerns,

ACS Publications © XXXX American Chemical Society A

and deactivation can render processes complex, unecological, and uneconomical. Although, the following statement is an oversimplification, the high temperature and the low selectivity reflect the fact that only a small number of sites are catalytically active (the concept of "active site" was introduced by Sir H. S. Taylor).1 This makes it very difficult to accurately characterize these catalysts. Consequently structure/activity relationships are difficult to establish, and their synthesis can also be challenging to reproduce. This may explain why great discoveries were not obtained by a predictive approach but were constructed iteratively by "trial and error". One of the first examples (in the early 20th century) was the synthesis of ammonia, by Fritz Haber and Carl Bosch, whose development

Received: November 24, 2015

Figure 1. Surface organometallic fragments (SOMF) evidenced by surface organometallic chemistry (SOMC).

required thousands of experiments.2 Its mechanism and the nature of the active site were elucidated later after decades of research by G. Ertl, thanks to the spectacular development of surface science.3

In contrast to heterogeneous catalysis, homogeneous systems are better understood regarding both their active site (structure) and the elementary steps involved (mechanism). Because this discipline strongly overlaps with molecular and organometallic chemistry, it has benefited from the tools available for structure determination and reactivity studies of organometallic molecules. This unbalanced situation has led to an important transfer of concepts between homogeneous and heterogeneous catalysis.4

The following concept is now evident and widely accepted by the community: When a carbon-containing molecule reacts with a surface to give products, bonds are broken and others are formed. During these catalytic processes, "surface organometallic fragments" (SOMFs) are generated (Figure 1). For molecules not containing carbons, (e.g., H2O, N2, NO, etc.), the rules of surface coordination chemistry (SCC) prevail and will be considered in another review. Metal hydrides have been included in this Account because they are key in most of the catalysis tackled.5

These SOMFs can be synthesized using surface organo-metallic chemistry (SOMC) methods to study their reactivity. It is possible to transfer most concepts of molecular chemistry to design and rationalize heterogeneous catalysis bridging the existing gap between heterogeneous and homogeneous catalysis. The reactivity follows the elementary steps of molecular chemistry and in particular organometallic chemistry. In fact, the synthesis and isolation of such versatile and fragile intermediates has allowed investigation of their reactivity using the specific tools of SOMC (i.e., in situ and in operando FTIR, gas evolution, surface microanalysis of air-sensitive compounds, UV, EXAFS, XANES, SEM, TEM, and SS-NMR but also DFT).4 This Account will discuss the reality of this transfer of concepts and the understanding of catalytic mechanism using key SOMF fragments. Furthermore, the identification of the sequence of elementary steps and the reaction intermediates

from substrates to products can allow us to design a new generation of single-site heterogeneous catalysts.

The molecular understanding of homogeneous catalysis can be adapted to surfaces and heterogeneous catalysis, taking in account the properties inherent to the surface ligand (such as its bulk, its rigidity, and the complexity of grafting sites). The term surface organometallic chemistry (SOMC) covers the rich chemistry of organometallic compounds grafted onto oxides surfaces4 and nanoparticles4 of unsupported or supported zero-valent particles. When surface organometallic chemistry was initially developed,5 the fundamental consequences in terms of prediction of the catalytic activity of these SOMFs were not foreseen. In this Account, we aspire to demonstrate that a reasonable and simplified approach of predictable single-site heterogeneous catalysis can now be advanced for a wide range of catalytic reactions. We believe the emerging concepts exposed in this Account to be beneficial to all the heterogeneous catalysis community. We intend its terms to be accessible even to nonspecialized chemists.

■ CATALYSIS BY DESIGN: THE MAIN CONCEPTS

The SOMC approach considers the surface as a ligand for binding to organometallic complex, molecular in nature, by

X = Spectator ligand

Oxo, amido, imido, Cp,...

R = "Functional" I oT ""M ligands

I I v Hydride, Alkyl, Carbene,

0 y ......-o Carbyne, Alkylidene,

J^j^ iw _4m/cfo. Alkoxo_

M'= Oxide Support

Si02 (fused silica MCM-41. SBA15), Al203, Si02-Al203

Figure 2. Schematic presentation of the design of a surface organometallic fragment able to achieve any catalytic reaction (mono- or multifunctional).6,7

M = Transition Metal

Ti, Zr, Hf, V, Ta, Cr, Mo, W, Re. Fe, Co, Ni...

Scheme 1. "Catalysis by Design" Approach To Construct Catalytic Cycles

Scheme 2. Activation of Methane (or Other Alkanes) by a Bond Metathesis on Supported d0 Metals (Groups 4, 5, and 6) Followed by Various Elementary Steps (ß-Alkyl Transfer for Higher Alkanes and d0 Metals of Group 4) or a-Alkyl Transfers (for Higher Alkanes and d2 Metals of Groups 5 and 6)

Figure 3. Comparison of the low temperature rate of alkane hydrogenolysis with supported metal hydride.

covalent (or sometimes ionic) bonds. The metal center and its coordination sphere (e.g., formulation, electronic configuration, oxidation state, possible degree of coordinative unsaturation,

podality, and geometry) can be accessed by employing the SOMC methodology and concepts (Figure 2). This leads to a predictable catalytic reaction provided that a molecular

Scheme 3. Proposed Mechanisms of Hydrogenolysis of Alkanes, Waxes and Polyolefins by d° Zirconium Metal Hydrides (ß-Alkyl Transfer) and by d° Tantalum Metal Hydrides (a-Alkyl Transfer)

Scheme 4. Mechanism of Isomerization and Cleavage by Hydrogen of LLDPE by [Zr]-H.4

Figure 4. Structures of hafnium hydrides on silica (neutral) and alumina (cationic).4,17

mechanism is at the origin of the synthesis of such coordination sphere. The catalytic activity can be tuned not only via the ligands of the so-called "molecular" fragment but also through the surface ligand because it has steric properties, electronic properties, porosity, hydrophobic or hydrophilic properties, or acid—base or redox character. One certain

Figure 5. Metal alkyl surface organometallic fragments (SOMFs).

advantage provided by the surface is also to allow the active sites to remain isolated from each other and prevent bimolecular deactivation frequently observed in homogeneous catalysis especially for very electron deficient metals.

Our strategy aims to render catalysis predictable by selecting the SOMFs presumed to be intermediates in the catalytic cycle of the reaction targeted (or as close as possible from such

Scheme 5. Proposed Mechanism for Homologation of Propane to Isobutane

Scheme 6. Olefins Metathesis Mechanism Proposed by Herisson and Chauvin9

presumed intermediates; Scheme 1). The study of their stoichiometric reactivity allows rationalization of the identification of the elementary steps. The preparation and the characterization of the surface organometallic fragments (SOMFs) have been discussed in depth in other reviews.4'8 In brief, a metal oxide (mostly silica, alumina, silica—alumina, mesoporous oxide materials) is dehydroxylated at different

temperatures (200 to ca. 1000 °C). Increased temperature leads to rearrangement of the surface by lowering the hydroxyl population and increasing the number of strained metal oxide bonds. Both parameters are pivotal for the surface species generated by grafting because silanol reacts with the metal and immobilizes onto the surface while the surface oxygen can interact and either coordinate with the surface metals. Under the right conditions and using the SOMC methodology, the resulting surface complexes and their coordination sphere can be fully characterized. We, and others,4 have extended the field of SOMC to most of the metallic elements of the periodical table, using a large variety of organic ligands, and inorganic supports. In the following, we will consider SOM fragments and their insertion into several catalytic cycles (Figure 1). There are several cases depending on the number and type of fragment required for the catalytic cycle (multifunctional catalyst) but also on the presence of spectator ligand.

In this Account, catalytic reactions have been selected based on the clarity of their mechanism with the sequence of simple elementary steps identified (in the organometallic chemistry). They will be classified by the SOMF, which is expected to be inside the catalytic cycle or as close as possible from pivotal intermediates. For instance, the hydrogenolysis of alkanes occurs due to d0 electron deficient metal hydrides. These species are initially generated by c-bond metathesis between

Figure 6. Product distribution observed in propane metathesis with various supported metal hydrides.

Scheme 7. Proposed Mechanism of Alkane Metathesis by Tantalum Catalyst Starting Either from a Metal Hydride (right) or from a Metal Alkyl (left)

Scheme 8. Silica-Supported Tungsten Methyl Carbyne in Equilibrium with Bis-carbene

Figure 7. Supported molybdenum carbenes for olefin metathesis.

hydrogen and a metal alkyl. Another example is olefin metathesis. Most the known commercial catalysts are supported tungsten oxide. Yet their reactivity was assumed to be due to a presumed surface metallocarbene in agreement with the Chauvin mechanism.9 Evidence has been gathered that starting directly from a surface metallocarbene or its immediate precursors removes the induction period and the turn over numbers (TON) are the highest ever found (above 100 000) for this reaction.

In the following, for each proposed cycle, the SOMFs of interest are marked by a dotted circle. The supported metal will be written as [M] since we will focus more on the mechanistic approach. We obviously do not intend to cover all catalysis but

mostly the examples where this approach has been the most successful.

Metal Hydrides ([M]—H)

The simplest SOMFs are metal hydrides ([M]— H), which do not belong to the "organometallic" classification in the strict sense but deserve to be discussed here due to their frequent involvement. They can be conveniently obtained by hydro-genolysis of metal—alkyl ([M]—R) groups and can promote low temperature C—H bond activation of alkanes (i.e., methane activation; Scheme 2). Most of these metal hydrides are formally d0 and can only undergo c-bond metathesis. Tantalum hydrides stand apart by being able to activate alkanes either as Ta(III)H (d2) or as Ta(V)H3 (d0). In propane hydrogenolysis,

Scheme 9. Catalytic Cycle for the Conversion of 1-Butene to Propylene with the Precursor [W](=O)H3/Alumina Initiation

Scheme 10. Catalytic Cycle for the Conversion of trans-2-Butene to Propylene33

the first step of the mechanism, a c-bond metathesis, occurs more easily on [(=Si—O—)2Ta(V)(H)3] than on [(=Si— O—)2Ta(lIl)(H)].10 The reactions involving monofunctional SOMF metal hydrides are beyond the scope of this Account to examine each of them thoroughly, yet a rapid overview is required.

Most metal hydrides have the properties of low temperature hydrogenolysis of alkanes or waxes. The catalyst efficiency varies with the metal employed (Ti > Zr > Hf, Ta> W), but detailed investigation11 has revealed that the mechanism differs because group 5 and 6 hydrides cut the metal—alkyl by a-alkyl transfer while group 4 metals undergo ^-alkyl transfer (Figure

The alkane depolymerization12 was discovered almost simultaneously with the hydrogenolysis of alkanes and polyolefins.13 It is a fascinating transformation in which a group 4 metal-based catalyst converts alkane into smaller units by successive elementary steps of ^-alkyl transfer and c-bond metathesis under hydrogen. But it can also polymerize ethylene to polyethylene and break down polyethylene under hydrogen to diesel range gasoline. This reaction may be considered as the "microscopic reverse of Ziegler—Natta polymerization". This may not be a correct wording since hydrogen is introduced to render feasible the thermodynamically unfavored depolymeri-zation. Alkanes ranging from small molecular to long chain can be transformed in the presence of hydrogen to either methane or ethane. Interestingly group 5 and 6 metals can achieve such

Scheme 11. Tungsten Hydrido Carbene as a SOMF for 1-Butene Transformation to Propylene by a Succession of Catalytic Cycles (Self-metathesis, Cross-metathesis, and Isomerization)

Scheme 12. Elementary Steps in the Non-oxidative Coupling of Methane to Ethane and Hydrogen0

aThe NMR values were obtained by reacting 13CH W hydrides on silica alumina500 at 200 °C.

Scheme 13. Mechanism for Isometathesis of Cyclooctane from C8 to C16

hydrogenolysis, but in this case they must have a d2 configuration because the insertion mechanism requires two electrons to accommodate the formed metallacarbene (Scheme

[Zr]—H SOMFs can also isomerize and cleave linear low density polyethylene (LLDPE) into polyethylene by successive elementary steps of ^-alkyl eliminations (Scheme 4). In parallel, [Zr] —R fragments generated in the catalytic cycle can undergo "chain walking" by a series of ^-H elimination and olefin reinsertion ending with the release of alkyl branch chains by hydrogenolysis of [Zr] —R. The three processes work together to abstract branches from the polymer chain and also to cleave the polymer backbone.

Both neutral and cationic metal hydrides can be prepared by grafting Hf(CH2—C(CH3)3)4 on silica or alumina (Figure

4).14'15 The neutral form exhibits an activity of about 35 kg PE/ (mol Hf-h) whereas the cationic form has an activity of 400 kg PE/(mol Hf-h) in 3 bar at 50 °C and in heptane.4 The strong increase of activity by rendering the surface cationic is very general and was frequently observed to prepare floating cations using nickel or hafnium and either an alumina support or an external Lewis acid.16

Metal Alkyls ([M]—R)

Metal alkyls represent such a large family of surface complexes that it is difficult to cover even the most recent examples (Figure 5). The discovery of new polymerization catalysts by supported metal (alkyls) was an easy pathway reference.15'18

The case of bisalkyl SOMF is certainly more interesting since it promotes the cooperation between two alkyl groups and can

Scheme 14. Elementary Steps Occurring during Cyclooctane Isometathesis and Leading to C6-C7 and C12-C16

Figure 8. Product distribution for n-decane isometathesis.

produce higher alkanes from propane. A mechanism can be proposed starting from a bispropyl zirconium (generated in situ by propane activation on bis-zirconium hydride). For instance, competing processes of successive ^-H elimination, olefin

insertion, and ^-alkyl transfer allows the rearrangement of the hydrocarbon (Scheme 5).

This is effectively possible using a bis-zirconium hydride species, which upon reaction with propane gives the corresponding bis-zirconium propyl, and the resulting alkanes

are the linear and branched higher paraffins. Propane homologation has been reported for zirconium complexes supported on silica and silica—alumina when heated under supercritical conditions [(AlO)2ZrNp2]19 and [(=SiO)-ZrNp2(OAr)].20 Typically a bis-neopentyl zirconium SOMF (or a bis-zirconium hydride) is exposed to propane. At 210 and 290 °C, propane is catalytically transformed into its lower (methane and ethane) and higher homologues (mainly butane), though via a different mechanism than that of alkane metathesis (vide infra). The road to polyolefin from alkanes remains wide open.

Metal Carbenes (M=CR2): Olefin Metathesis

Since its discovery in 1964 by Banks and Bailey,21 olefin metathesis has been the focus of intense research for both fundamental and applied reasons.9,22 Chauvin and Herisson

first proposed the metallacarbene mechanism in 1971 based on the initial products of cross-metathesis of cyclic and acyclic olefins (Scheme 6).9 Heterogeneous catalysis was developed with the presumption that the active site was a surface carbene. The first evidence of surface carbene, [(=Si—O)2Nb=CH2], was reported by Iwasawa23 followed by Herrmann et al. [(= Si—O—)Mo(=CHCMe3)(=NH)Np] then by us [(=Si— O—)Re(Np)(=C-CMe3)(=CHCMe3)].24 More surface car-

benes or their precursors have been synthesized since the report of Iwasawa,23 validating the presence of [M]=CR2 as SOMF. Various levels of activity and selectivity were observed for olefin metathesis and could be correlated to both the nature of the metal employed and the spectator ligands (i.e., oxo, imido, amido, alkyl, etc.). Mechanistic studies have always been consistent with the metallacycle intermediate of Chauvin.9 They were also validating examples of the "Catalysis by design" approach by starting the catalysis from its immediate precursor or one reaction intermediate. Spectator ligands were found to be influential over the catalytic performance (oxo, imido, amido, alkyls, etc.). For instance, oxo spectators can induce higher lifetime than imido ones, while the latter lead to the formation of isobutene via a ^-H elimination from the metallacyclobutane.6,25

In the following sections, we will address the single-site cascade reactions, which require several SOMFs generated on the same metal. For example, a metal can exhibit two fragments (simultaneously or sequentially) such as hydride and carbene. The CH bond of paraffins can be activated by the metal hydride (see section Metal Hydrides ([M]—H)) to give a metal—alkyl plus H2, and the metallocarbene can then metathetize the olefin resulting from ^-H elimination from this metal—alkyl.

Scheme 16. Elementary Steps of the Hydroaminoalkylation of Olefins

Scheme 17. Elementary Steps of Imine Metathesis

Rf N + R3 N

LnM=NR

Metal Carbene Hydrides [M](H)(=CR2) or Metal Carbene Alkyls [M]R(=CR'2): n-Alkane, Isoalkane, and Cycloalkane Metathesis

Alkane metathesis was discovered in 1997 by using tantalum and tungsten hydride catalyst precursors supported on alumina or silica; their molybdenum counterparts were later revealed also to be active.27 This transformation, in which saturated hydrocarbons are rearranged to longer or shorter paraffins, was a real chemical challenge considering the inertness of the sp3 carbon—hydrogen or C—C bonds. For instance, n-propane can be converted into (C1; C2 C4, C5, etc.) paraffins under mild conditions. It was later demonstrated that both linear and branched alkanes can also be transformed into the next higher and lower branched paraffins (Figure 6).

The determination of the catalytic mechanism and the coordination sphere of the various intermediates was a progressive deduction of all the intermediates and primary products. These primary products (olefin and hydrogen)28 were isolated recently and consistent of a carbene SOMF associated with a hydride, [M](H)(=CR2), although the starting precursor was only a hydride. This implied the occurrence of a multifunctional catalyst including a metal hydride to activate the C—H bond of paraffins and a carbene able to carry-out metathesis of an olefin. These observations were in agreement with the low temperature hydrogenolysis properties of tungsten hydrides. It is only recently that the full observation of a tungsten carbene hydride could be observed by hydrogenolysis of (=Si—O—)WMe5 at —78 °C. The first

tantalum methylidene methyl [(=Si—O—)2Ta(=CH2)Me] had been selectively prepared, identified, and found active in alkane metathesis (Scheme 7).29 Thermal treatment of [(=Si— O—)TaMe4] (at ca. 80 °C) induces an a-H transfer from one methyl to another with formation of methane. This releases one equivalent of methane and forms a mixture of monopodal [Ta](=CH2)(CH3)2 and bipodal [Ta](=CH2)CH3. Heating [(=Si—O—)WMe5] above 100 °C promotes the unexpected

formation of tungsten carbyne products: a monopodal [(=Si— O—)W(=CH)Me2] and a bipodal [(=Si—O—)2W(= CH)Me]. ([w](=CH2)(CH3)) or ([W](=CH2)(H)) were the presumed products. An easy equilibrium between the carbyne dimethyl and the biscarbene monomethyl exist for the monopodal tungsten species (Scheme 8). A mechanism similar to the one proposed for tantalum also occurs.30

When [(=Si—O—)WMe5] is heated from —78 °C to room temperature under hydrogen, the obtained catalyst is extremely active for alkane metathesis (TON as high as 3000 can be reached at 150 °C). This controlled warming leads to the formation of an unique methylidene tungsten hydride [(=Si— O—)W(H)3(=CH2)]. The unprecedented characterization of both carbene and hydride functionalities by solid-state 1H and 13C NMR spectroscopies was also supported by DFT.31

Considering molybdenum analogues, the simultaneous presence of a carbene and an alkyl ligand on the atom center were a prerequisite to alkane metathesis. For instance, substituting the neopentyl ligand with diphenylamino pyrrole (or dimethylpyrrole) yields inactive catalysts. The imido ligand

plays the role of a spectator stabilizing tetracoordination around the tungsten (Figure 7).27

[W](=O)(H)x/Al2O3] could also transform branched alkanes (such as 2-methylpropane) into 2,3-dimethylbutane (42%), ethane (41%), and 2-methylpropane (3%).32 Due to the alkane substrate's branches, the abundance of the higher homologues can be ordered as follows: Cn+2 » Cn+1 > Cn+3. The overall mechanism is similar to that with linear alkanes, but the difference of selectivity observed was consistent with the steric properties of the tungsten center: the less hindered tungsten alkylidene or resulting metallacycle promotes faster [2 + 2] cycloaddition.

The duality of two functions one for C—H bond activation and the second for olefin metathesis is a unique property of SOMFs to the best of our knowledge. Several catalyses can be explained by the presence simultaneously on the same metal of dual functions, a carbene/hydride or carbene/alkyl. The reactions can be described by a variety of "cascade" elementary steps. This working hypothesis based on an accumulation of positive results has been at the origin of the discovery of numerous new reactions.

[M](H)(=CR2): Direct Transformation of Ethylene to Propylene. Ethylene is selectively transformed into propylene in a continuous-flow reactor using alumina-supported tungsten hydrides [W](=O)(H)3/Al2O3 (TON > 760 at 150 °C). Since the catalyst is also active for olefin metathesis, the reaction was assumed to proceed at a "trifunctional catalytic site" by ethylene dimerization to 1-butene, 1-butene isomerization to 2-butene, and cross-metathesis of ethylene with 2-butene. Although this reaction had been observed initially by employing [(=Si— O—)W(=O)H 3], the real catalyst coordination sphere was postulated to be [(=Si—O—)W(=O)(C2H5)(=CHCH3)] resulting from repeated insertions of ethylene into the hydrides followed by a-H transfer from one a-carbon of the alkyl to the other one with liberation of propane (Scheme 9).

[M](H)(=CR2): Production of Propylene from 2-Butene. The first example of propylene production from trans-2-butene over [W](=O)(H)x/a12O3 at 150 °C was observed in a continuous flow reactor at different pressures (TON > 6000 at atmospheric pressure). The active SOMF is assumed to be a tungsten carbene hydride, which operates as a "bifunctional catalyst" through the disfavored 2-butene isomer-ization to 1-butene on W-hydride and 2-butene/1-butene cross-metathesis on the W-carbene (Scheme 10).

[M](H)(=CR2)/Alumina]: Transformation of 1-Butene to Propylene. The transformation of 1-butene to propylene obeys a "tetrafunctional" mechanism. This means that the 1-butene is first isomerized to 2-butene. 1-Butene can undergo self-metathesis to produce ethylene and 3-hexene; then cross metathesis can occur between 1-butene and 2-butene, and ethylene may undergo cross metathesis with 2-butene. All these steps are assumed to occur on the [W] carbene hydride (Scheme 11).34

[M](H)(=CR2): Direct Coupling of Methane to Ethane and Hydrogen. Direct nonoxidative coupling transformation of methane into ethane and hydrogen occurs at T > 250 °C with a high selectivity (>98%) but low TON (<70), using precursor silica-supported tantalum hydride, [(=SiO)2Ta-(H)x], x = 1 or 3, or silica—alumina supported [W](= O)(H)3/AlO.35 The key intermediate in this reaction is postulated to be [Ta](CH3)(=CH2); the methylene group can insert into the tantalum methyl bond of the [Ta](CH3) affording [Ta](CH2—CH3). Furthermore, W—Me, W(=CH2),

and W(=CH) can be detected by reacting CH4 with silica— alumina supported W(=O)hydrides (Scheme 12). Interestingly, the elementary step of methylidene insertion into a W— Me group can be considered similar to chain growth initiation in Fischer—Tropsch synthesis.36

[M](H)(=CR2): Isometathesis of Cyclohexane. With the precursor [(=Si—O—)WMe5], cyclooctane can undergo metathesis to produce cycloalkanes, cyclohexadecane, cyclotetraco-sane, cycloheptane, and cyclohexane (Schemes 13 and 14; conv. 70% and TON 450 at 150 °C). The intriguing observation was the formation of cyclopentane to cycloheptane, as well as cyclodecane, cycloundecane, cyclododecane, cyclo-tridecane, and cyclotetradecane up to cyclohexatriacontane. A mechanism was proposed in which the double bond migrates along the SOMF and leads to an intramolecular olefin ring closing metathesis step. All these observations indicate the duality of the coordination sphere of the tungsten with [M] —H (C—H bond activation step) and [M](C=R)2 (olefin metathesis).

[M](H)(=CR2): Isometathesis of n-Decane. Alkane metathesis on substrates with a carbon number higher than five results in a product distribution like in the previous case, which unexpectedly goes much further than Cn+2 starting from Cn. The simplest explanation is that each time an olefin formed by decomposition of a metallacyclobutane is adjacent to a metal hydride, a reinsertion of the newly formed olefin may occur on this hydride (Figure 8). Then a succession of reinsertion and ¡-alkyl elimination occurs leading to an increase of the number of carbons. This was observed at the beginning with a bimetallic Zr/W catalyst, but it appeared very general. We have decided to call it "isometathesis" of alkanes (ISOMET-alkanes).

[M](=CR): Alkyne Cyclotrimerization

Silica-supported tungsten carbyne complexes (=SiO)xW(= CH)(Me)y (x =1, y = 2; x = 2, y =1) are efficient precatalysts for terminal alkyne cyclotrimerization with high TONs (e.g., for 1-hexyne, conv. 95%, TON 11116 at 150 °C). This catalyst is active for alkyne cyclotrimerization (Scheme 15) without the formation of significant alkyne metathesis products.

Metal-Aziridines: Hydroaminoalkylation of Olefins and Imine Metathesis

The hydroaminoalkylation of olefins is a reaction of C—C bond forming strategy, involving the addition of an amino akyl fragment to an olefin. It occurs by the addition of an a-C—H bond adjacent to nitrogen across a C=C bond. Although the reaction has been known since the report of Nugent et al.,37 its mechanism was not yet fully understood.38 Silica-supported bis-amido zirconium SOMF was prepared and allowed to isolate both the metallaaziridine 139 and the five-membered zircona-cycle metallacycle intermediate 2, after treatment of 1 with 1-propene (Scheme 16). The regeneration of the catalyst was conducted by dimethylamine protonolysis and affords pure amines. The catalytic cycle of a homogeneous catalytic reaction was almost fully demonstrated by SOMC. This example shows that our strategy catalysis by design can be applied also to organic synthesis. All reaction intermediates could be observed by entering into the catalytic cycle by a reaction intermediate and completing the full catalytic cycle step-by-step.

Interestingly starting from a silica supported Zr(=NEt2)-(NEt2), the metathesis of imines (Scheme 17) could be observed as expected from our strategy (conv. 76% can be obtained at 80 °C with N-(3,4 methylenedioxybenzylidene)-

benzamide).40

■ CONCLUSIONS

This Account attempted to explain the concept of "catalysis by design" applied to the field of heterogeneous catalysis. In principle, this model could be applied to any catalytic reaction occurring on a surface. Most of them occur via a surface organometallic fragment, and the best discipline to "make" such presumed intermediates is surface organometallic chemistry. It is true that the structure of nanocatalysts are certainly more complicated than our single-site systems: they show diversity in size, shape, support, and composition. These "physical" characteristics can be used to change catalytic properties of existing catalysts. We intend to tune our single site catalysts to improve their catalytic properties and compete with existing heterogeneous catalysts. We also intend to generalize the concept to reactions involving C—O, C—N, and C—X bonds.

■ AUTHOR INFORMATION Notes

The authors declare no competing financial interest. Biographies

Jeremie D. A. Pelletier received a doctoral degree in homogeneous catalysis in 2006 from the University of Leicester (U.K.). He pursued related research interests as a postdoctoral fellow at SASOL technology (St. Andrews, U.K.) and research associate at University of Liverpool. He moved back to France as a postdoctoral fellow with Jean Marie Basset at the University of Lyon to explore surface organometallic chemistry. He was appointed research engineer in 2010 of the Catalysis Center in King Abdullah University of Science and Technology (Saudi Arabia). He has been the current Lab Manager since 2011 and attempts to extend the range of reactions covered by SOMC (hydroaminalkylation, CO2 valorization, etc.).

Jean Marie Basset received his Ph.D. in 1969 from the University of Lyon, France. He came to Centre National de la Recherche Scientifique (CNRS) in 1971 and has occupied several positions, including vice director of the Institute of Catalysis (Lyon). He was appointed research Director at the CNRS in 1987 and funded his laboratory of "Surface Organometallic Chemistry" (University of Lyon) that he held from 1994 to 2008. He is a member of five Academies worldwide. His main research interests are the relations between homogeneous and heterogeneous catalysis and the metathesis of olefins and alkanes. He developed classical and modern aspects of surface organometallic chemistry. Currently, he is the Director of the KAUST Catalysis Center (KCC) in Saudi Arabia. He has developed this area thanks to his colleagues currently in Lyon and in Saudi Arabia. A special mention and thanks to all who he trained for the first time in surface organometallic chemistry in Lyon and in Saudi Arabia and who later developed this field in various universities worldwide including inter alia: M. Bouhrara (Kz), Y. Chen (Cn), V. Caps (Fr), C. Coperet (CH), R. Gauvin (Fr.), S. Scott (US), Y. Leroux (Norway), V. Delia (Th.), the late R. Sanchez Delgado (Ven, US), A.K. Smith (UK), W.Xuxu (CN).

■ ACKNOWLEDGMENTS

We thank KAUST for financial and human support and all our friends who helped us to develop this field and continue to do so in Europe, China, Japan, the Americas, and the Middle East including Saudi Arabia.

■ REFERENCES

(1) Taylor, H. S. The Heterogeneity of Catalyst Surfaces for Chemisorption. Adv. Catal. 1948, 1, 1—26.

(2) Bosch, C. Process of producing ammonia. U.S. Patent US 990191 A, 1911.

(3) Ertl, G. Oscillatory Catalytic Reactions at Single-Crystal Surfaces. In Advances in Catalysis; Eley, D. D., Pines, H., Weisz, P. B., Eds.; Academic Press: San Diego, CA, 1990; Vol. 37; pp 213-277.

(4) Basset, J.-M.; Psaro, R.; Roberto, D.; Ugo, R Modern Surface Organometallic Chemistry; Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2009.

(5) Basset, J. M.; Choplin, A. Surface Organometallic Chemistry - a New Approach to Heterogeneous Catalysis. J. Mol. Catal. 1983, 21, 95-108.

(6) Basset, J. M.; Coperet, C.; Soulivong, D.; Taoufik, M.; Cazat, J. T. Metathesis of Alkanes and Related Reactions. Acc. Chem. Res. 2010, 43, 323-334.

(7) Popoff, N.; Mazoyer, E.; Pelletier, J.; Gauvin, R M.; Taoufik, M. Expanding the scope of metathesis: a survey of polyfunctional, singlesite supported tungsten systems for hydrocarbon valorization. Chem. Soc. Rev. 2013, 42, 9035-9054.

(8) Coperet, C.; Chabanas, M.; Saint-Arroman, R. P.; Basset, J. M. Homogeneous and heterogeneous catalysis: Bridging the gap through surface organometallic chemistry. Angew. Chem., Int. Ed. 2003, 42, 156-181.

(9) Herisson, P. J. L.; Chauvin, Y. Transformation Catalysis of Olefins by Tungsten Complexes 0.2. Telomerization of Cyclic Olefins in Presence of Acyclic Olefins. Makromol. Chem. 1971, 141, 161-176.

(10) Pasha, F. A.; Cavallo, L.; Basset, J. M. Mechanism of n-Butane Hydrogenolysis Promoted by Ta-Hydrides Supported on Silica. ACS Catal. 2014, 4, 1868-1874.

(11) Chabanas, M.; Vidal, V.; Coperet, C.; Thivolle-Cazat, J.; Basset, J. M. Low-temperature hydrogenolysis of alkanes catalyzed by a silica-supported tantalum hydride complex, and evidence for a mechanistic switch from group IV to group V metal surface hydride complexes. Angew. Chem., Int. Ed. 2000, 39, 1962-1965.

(12) Dufaud, V. R.; Basset, J. M. Catalytic hydrogenolysis at low temperature and pressure of polyethylene and polypropylene to diesels or lower alkanes by a zirconium hydride supported on silica-alumina: A step toward polyolefin degradation by the microscopic reverse of Ziegler-Natta polymerization. Angew. Chem., Int. Ed. 1998, 37, 806810.

(13) Corker, J.; Lefebvre, F.; Lecuyer, C.; Dufaud, V.; Quignard, F.; Choplin, A.; Evans, J.; Basset, J. M. Catalytic cleavage of the C-H and C-C bonds of alkanes by surface organometallic chemistry: An EXAFS and IR characterization of a Zr-H catalyst. Science 1996, 271, 966-969.

(14) Tosin, G.; Delgado, M.; Baudouin, A.; Santini, C. C.; Bayard, F.; Basset, J. M. Surface Organometallic Chemistry of Hf(CH2-tBu)4 on Silica and Silica-Alumina: Reaction of the Resulting Grafted Hafnium Neopentyl with Dihydrogen. Organometallics 2010, 29, 1312-1322.

(15) Millot, N.; Soignier, S.; Santini, C. C.; Baudouin, A.; Basset, J. M. Synthesis, characterization, and activity in ethylene polymerization of silica supported cationic cyclopentadienyl zirconium complexes. J. Am. Chem. Soc. 2006, 128, 9361-9370.

(16) Dorcier, A.; Merle, N.; Taoufik, M.; Bayard, F.; Lucas, C.; de Mallmann, A.; Basset, J. M. Preparation of a Well-Defined Silica-Supported Nickel-Diimine Alkyl Complex - Application for the GasPhase Polymerization of Ethylene. Organometallics 2009, 28, 21732178.

(17) Delgado, M.; Santini, C. C.; Delbecq, F.; Baudouin, A.; De Mallmann, A.; Prestipino, C.; Norsic, S.; Sautet, P.; Basset, J. M. Characterization of Surface Hydride Hafnium Complexes on Alumina by a Combination of Experiments and DFT Calculations. J. Phys. Chem. C 2011, 115, 6757-6763.

(18) Williams, L. A.; Guo, N.; Motta, A.; Delferro, M.; Fragala, I. L.; Miller, J. T.; Marks, T. J. Surface structural-chemical characterization of a single-site d(0) heterogeneous arene hydrogenation catalyst having 100% active sites. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 413-418.

(19) Thieuleux, C.; Maraval, A.; Veyre, L.; Coperet, C.; Soulivong, D.; Basset, J. M.; Sunley, G. J. Homologation of propane catalyzed by oxide-supported zirconium dihydride and dialkyl complexes. Angew. Chem., Int. Ed. 2007, 46, 2288-2290.

(20) Guillemot, G.; Thieuleux, C.; Coperet, C.; Soulivong, D.; Spitzmesser, S.; Basset, J. M. Modifying the reactivity in the homologation of propane by introducing aryloxide ligands on a silica supported zirconium alkyl system. Dalton Trans. 2007, 4589-4593.

(21) Banks, R. L.; Bailey, G. C. Olefin Disproportionation - New Catalytic Process. Ind. Eng. Chem. Prod. Res. Dev. 1964, 3, 170-173.

(22) Hoveyda, A. H.; Zhugralin, A. R. The remarkable metal-catalysed olefin metathesis reaction. Nature 2007, 450, 243-251.

(23) Asakura, K.; Nishimura, M.; Iwasawa, Y. Spectroscopic Evidence for a Surface Nb Carbene in a New Sio2-Attached Nb Catalyst Active for Propene Metathesis. J. Mol. Catal. 1989, 55, 159-169.

(24) Herrmann, W. A.; Wagner, W.; Flessner, U. N.; Volkhardt, U.; Komber, H. Methyltrioxorhenium as Catalyst for Olefin Metathesis. Angew. Chem., Int. Ed. 1991, 30, 1636-1638.

(25) Mazoyer, E.; Merle, N.; de Mallmann, A.; Basset, J. M.; Berrier, E.; Delevoye, L.; Paul, J. F.; Nicholas, C. P.; Gauvin, R. M.; Taoufik, M. Development of the first well-defined tungsten oxo alkyl derivatives supported on silica by SOMC: towards a model of WO3/SiO2 olefin metathesis catalyst. Chem. Commun. 2010, 46, 8944-8946.

(26) Vidal, V.; Theolier, A.; ThivolleCazat, J.; Basset, J. M. Metathesis of alkanes catalyzed by silica-supported transition metal hydrides. Science 1997, 276, 99-102.

(27) Blanc, F.; Coperet, C.; Thivolle-Cazat, J.; Basset, J. M. Alkane metathesis catalyzed by a well-defined silica-supported mo imido alkylidene complex: [(SiO)Mo(= NAr)(= CHtBu) (CH(2)tBu)]. Angew. Chem., Int. Ed. 2006, 45, 6201-6203.

(28) Basset, J. M.; Coperet, C.; Lefort, L.; Maunders, B. M.; Maury, O.; Le Roux, E.; Saggio, G.; Soignier, S.; Soulivong, D.; Sunley, G. J.; Taoufik, M.; Thivolle-Cazat, J. Primary products and mechanistic considerations in alkane metathesis. J. Am. Chem. Soc. 2005, 127, 8604-8605.

(29) Chen, Y.; Abou-hamad, E.; Hamieh, A.; Hamzaoui, B.; Emsley, L.; Basset, J. M. Alkane Metathesis with the Tantalum Methylidene [(SiO)Ta(=CH2)Me-2]/[(SiO)(2)Ta(=CH2)Me] Generated from Well-Defined Surface Organometallic Complex [(SiO) (TaMe4)-Me-V]. J. Am. Chem. Soc. 2015, 137, 588-591.

(30) Callens, E.; Abou-Hamad, E.; Riache, N.; Basset, J. M. Direct observation of supported W bis-methylidene from supported W-methyl/methylidyne species. Chem. Commun. 2014, 50, 3982-3985.

(31) Maity, N.; Barman, S.; Callens, E.; Samantaray, M.; Abou-Hamad, E.; Minenkov, Y.; D'Elia, V.; Hoffman, A.; Widdifield, C.; Cavallo, L.; Gates, B.; Basset, J.-M. Controlling the Hydrogenolysis of Silica-Supported Tungsten Pentamethyl Leads to a Class of Highly Electron Deficient Partially Alkylated Metal Hydrides. Chem. Sci. 2016, 7, 1558-1568.

(32) Merle, N.; Stoffelbach, F.; Taoufik, M.; Le Roux, E.; Thivolle-Cazat, J.; Basset, J. M. Selective and unexpected transformations of 2-methylpropane to 2,3-dimethylbutane and 2-methylpropene to 2,3-dimethylbutene catalyzed by an alumina-supported tungsten hydride. Chem. Commun. 2009, 2523-2525.

(33) Mazoyer, E.; Szeto, K. C.; Basset, J. M.; Nicholas, C. P.; Taoufik, M. High selectivity production of propylene from 2-butene: non-degenerate pathways to convert symmetric olefins via olefin metathesis. Chem. Commun. 2012, 48, 3611-3613.

(34) Mazoyer, E.; Szeto, K. C.; Norsic, S.; Garron, A.; Basset, J. M.; Nicholas, C. P.; Taoufik, M. Production of Propylene from 1-Butene on Highly Active "Bi-Functional Single Active Site" Catalyst: Tungsten Carbene-Hydride Supported on Alumina. ACS Catal. 2011, 1, 1643-1646.

(35) Soulivong, D.; Norsic, S.; Taoufik, M.; Coperet, C.; Thivolle-Cazat, J.; Chakka, S.; Basset, J. M. Non-oxidative coupling reaction of methane to ethane and hydrogen catalyzed by the silica-supported tantalum hydride: (SiO)(2)Ta-H. J.Am. Chem. Soc. 2008, 130, 50445045.

(36) Schulz, H. Short history and present trends of Fischer-Tropsch synthesis. Appl. Catal., A 1999, 186, 3-12.

(37) Nugent, W. A.; Ovenall, D. W.; Holmes, S. J. Catalytic C-H Activation in Early Transition-Metal Dialkylamides and Alkoxides. Organometallics 1983, 2, 161-162.

(38) Chen, T. Y.; Tsutsumi, R.; Montgomery, T. P.; Volchkov, I.; Krische, M. J. Ruthenium-Catalyzed C-C Coupling of Amino Alcohols with Dienes via Transfer Hydrogenation: Redox-Triggered Imine Addition and Related Hydroaminoalkylations. J. Am. Chem. Soc. 2015, 137, 1798-1801.

(39) Hamzaoui, B.; El Eter, M.; Abou-Hamad, E.; Chen, Y.; Pelletier, J. D. A.; Basset, J. M. Well-Defined Single-Site Monohydride Silica-Supported Zirconium from Azazirconacyclopropane. Chem. - Eur. J.

2015, 21, 4294-4299.

(40) Hamzaoui, B.; Pelletier, J. D. A.; Abou-Hamad, E.; Basset, J.-M. Well-defined silica-supported zirconium-imido complexes mediated heterogeneous imine metathesis. Chem. Commun. (Cambridge, U. K.)

2016, DOI: 10.1039/C6CC00471G.