reaction pathways and mechanisms in thermocatalytic biomass conversion ii: homogeneously catalyzed...

Marcel Schlaf Z. Conrad Zhang Editors
Reaction Pathways and Mechanisms in Thermocatalytic Biomass Conversion II Homogeneously Catalyzed Transformations, Acrylics from Biomass, Theoretical Aspects, Lignin Valorization and Pyrolysis Pathways
Green Chemistry and Sustainable Technology
Series editors Prof. Liang-Nian He State Key Laboratory of Elemento-Organic Chemistry, Nankai University, Tianjin, China
Prof. Robin D. Rogers Department of Chemistry, McGill University, Montreal, Canada
Prof. Dangsheng Su Shenyang National Laboratory for Materials Science, Institute of Metal Research, Chinese Academy of Sciences, Shenyang, China and Department of Inorganic Chemistry, Fritz Haber Institute of the Max Planck Society, Berlin, Germany
Prof. Pietro Tundo Department of Environmental Sciences, Informatics and Statistics, Ca’ Foscari University of Venice, Venice, Italy
Prof. Z. Conrad Zhang Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, China
Aims and Scope The series Green Chemistry and Sustainable Technology aims to present cutting- edge research and important advances in green chemistry, green chemical engineering and sustainable industrial technology. The scope of coverage includes (but is not limited to):
– Environmentally benign chemical synthesis and processes (green catalysis, green solvents and reagents, atom-economy synthetic methods etc.)
– Green chemicals and energy produced from renewable resources (biomass, carbon dioxide etc.)
– Novel materials and technologies for energy production and storage (biofuels and bioenergies, hydrogen, fuel cells, solar cells, lithium-ion batteries etc.)
– Green chemical engineering processes (process integration, materials diversity, energy saving, waste minimization, effi cient separation processes etc.)
– Green technologies for environmental sustainability (carbon dioxide capture, waste and harmful chemicals treatment, pollution prevention, environmental redemption etc.)
The series Green Chemistry and Sustainable Technology is intended to provide an accessible reference resource for postgraduate students, academic researchers and industrial professionals who are interested in green chemistry and technologies for sustainable development.
More information about this series at http://www.springer.com/series/11661
Reaction Pathways and Mechanisms in Thermocatalytic Biomass Conversion II Homogeneously Catalyzed Transformations, Acrylics from Biomass, Theoretical Aspects, Lignin Valorization and Pyrolysis Pathways
ISSN 2196-6982 ISSN 2196-6990 (electronic) Green Chemistry and Sustainable Technology ISBN 978-981-287-768-0 ISBN 978-981-287-769-7 (eBook) DOI 10.1007/978-981-287-769-7
Library of Congress Control Number: 2015951980
Springer Singapore Heidelberg New York Dordrecht London © Springer Science+Business Media Singapore 2016 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifi cally the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfi lms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specifi c statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made.
Printed on acid-free paper
Springer Science+Business Media Singapore Pte Ltd. is part of Springer Science+Business Media (www.springer.com)
Editors Marcel Schlaf The Guelph-Waterloo-Centre for Graduate Work in Chemistry (GWC)2
Department of Chemistry University of Guelph Guelph , ON , Canada
Z. Conrad Zhang Dalian National Laboratory
for Clean Energy Dalian Institute of Chemical Physics Dalian , China
Pref ace
Short carbon chain molecules (C 2 –C 9 ) obtained either directly from sugars, the hydrolysis of starch, or preferably by the controlled breakup of lignocellulosic biomass into soluble components are the only conceivable sustainable source of carbon on the planet that could ultimately replace the fossil hydrocarbons that currently form the base of the chemical industry and hence our technological civilization at large. In particular, the production of polymer components and polymers that are chemically or at least functionally equivalent to those derived from the refi ning of crude oil would offer ecologic and environmental as well as economic advantages.
The use of sugars, starch, and ultimately lignocellulosic biomass, i.e., forestry (e.g., wood and bark chips, etc.) and agricultural (e.g., straws, husks, stovers, etc.) residues, as a renewable carbon resource will, however, require careful life-cycle analyses of the processes involved. This in turn is critically dependent on a deep and detailed understanding of the mass and energy fl ows in these processes and hence their mechanisms at the molecular level. Almost “by defi nition” these processes will have to be catalytic in nature to be ecologically sustainable and economically viable.
The development of new catalysts and catalytic processes that are specifi cally designed for and adapted to the unique properties of the biomass-derived carbon substrates poses a unique challenge. Due to the abundance of oxygen-containing functional groups, the pentose and hexose sugars and their furanic condensates obtainable from (hemi-)cellulose as well as the phenol propanoid units of lignin are characterized by a high polarity and reactivity that is very different – one could say almost opposite – to that of the traditionally employed alkane and arene sources available from refi ned crude oil. The fundamental study of the reaction cascades and mechanisms involved in the transformation of oxygenated biomass to value-added chemicals is the fi rst step to meet this challenge.
Focusing on the use of thermochemical and acid-/base- or metal-catalyzed processes only, the two volumes of this book attempt to give an overview of and insights into the specifi c aspects of this challenge as perceived and formulated by expert contributors research-active in this fi eld.
vi
Volume I is comprised of chapters that address the nanoscale structure of lignocellulose, the application of acid-base reactions and catalysts to the depolymerization of cellulose, the use of heterogeneous hydrogenation catalysts for its direct conversion to polyols, as well as chapters that explore pathways for the metal- catalyzed dehydration and oxidation of sugars and sugar alcohols to furans and carboxylic acids, respectively.
The chapters of Volume II cover the hydrodeoxygenation of sugar-derived substrates by homogenous catalysts systems; the valorization of carboxylic acids, notably lactic acid and its derivatives; a theoretical approach to the elucidation of the conversion pathways of sugars and sugars condensates and their decomposition to humins; as well as mechanistic and practical aspects of the conversion and pyrolysis of lignin to functionalized monocyclic aromatics and the pyrolysis of biomass to synthesis gas.
We hope that the insights provided by the different and varied perspectives offered here will convince the readers that a switch to renewable biomass as a key carbon source for the chemical industry will be feasible and does indeed offer a way forward to a more sustainable future.
Guelph , Canada Marcel Schlaf Dalian, China Z. Conrad Zhang
Preface
vii
Contents
1 Deoxydehydration (DODH) of Biomass- Derived Molecules ................. 1 Shuo Liu , Jing Yi , and Mahdi M. Abu-Omar
2 Homogeneous Catalysts for the Hydrodeoxygenation of Biomass- Derived Carbohydrate Feedstocks ...................................... 13 Marcel Schlaf
3 Valorization of Lactic Acid and Derivatives to Acrylic Acid Derivatives: Review of Mechanistic Studies .................................. 39 Elodie Blanco , Stéphane Loridant , and Catherine Pinel
4 Computational Chemistry of Catalytic Biomass Conversion ............... 63 Guanna Li , Emiel J. M. Hensen , and Evgeny A. Pidko
5 Humin Formation Pathways .................................................................... 105 Jacob Heltzel , Sushil K. R. Patil , and Carl R. F. Lund
6 Catalytic Hydrodeoxygenation of Lignin Model Compounds .............. 119 Basudeb Saha , Ian Klein , Trenton Parsell , and Mahdi M. Abu-Omar
7 Oxidation of Lignins and Mechanistic Considerations ......................... 131 Adilson R. Gonçalves , Priscila Benar , and Ulf Schuchardt
8 Pyrolysis Mechanisms of Lignin Model Compounds Using a Heated Micro-Reactor ................................................................ 145 David J. Robichaud , Mark R. Nimlos , and G. Barney Ellison
9 Catalytic Gasification of Lignocellulosic Biomass ................................. 173 C. V. Pramod and K. Seshan
ix
Contributors
Mahdi M. Abu-Omar Department of Chemistry and the Center for Catalytic Conversion of Biomass to Biofuels (C3Bio) , Purdue University , West Lafayette , IN , USA
Spero Energy, Inc. , West Lafayette , IN , USA
Priscila Benar Instituto Agronômico de Campinas , Campinas , SP , Brazil
Elodie Blanco Institut de Recherches sur l’Environnement et la Catalyse de Lyon (IRCELYON) , UMR 5256, CNRS – Université Lyon 1 , Villeurbanne Cedex , France
G. Barney Ellison Department of Chemistry and Biochemistry , University of Colorado , Boulder , CO , USA
Adilson R. Gonçalves Pontifícia Universidade Católica de Campinas, PUCCAMP , Campinas , SP , Brazil
Jacob Heltzel Department of Chemical and Biological Engineering , University at Buffalo , Buffalo , NY , USA
Emiel J. M. Hensen Inorganic Materials Chemistry Group, Schuit Institute of Catalysis, Department of Chemical Engineering and Chemistry , Eindhoven University of Technology , Eindhoven , The Netherlands
Ian Klein Department of Chemistry and the Center for Catalytic Conversion of Biomass to Biofuels (C3Bio) , Purdue University , West Lafayette , IN , USA
Spero Energy, Inc. , West Lafayette , IN , USA
G. Li Inorganic Materials Chemistry Group, Schuit Institute of Catalysis, Department of Chemical Engineering and Chemistry , Eindhoven University of Technology , Eindhoven , The Netherlands
Guanna Li Inorganic Materials Chemistry Group, Schuit Institute of Catalysis, Department of Chemical Engineering and Chemistry , Eindhoven University of Technology , Eindhoven , The Netherlands
x
Shuo Liu Department of Chemistry , Purdue University , West Lafayette , IN , USA
Stéphane Loridant Institut de Recherches sur l’Environnement et la Catalyse de Lyon (IRCELYON) , UMR 5256, CNRS – Université Lyon 1 , Villeurbanne Cedex , France
Carl R. F. Lund Department of Chemical and Biological Engineering , University at Buffalo , Buffalo , NY , USA
Mark R. Nimlos National Renewable Energy Laboratory, National Bioenergy Center , Golden , CO , USA
Trenton Parsell Spero Energy, Inc. , West Lafayette , IN , USA
Sushil K. R. Patil Advanced Module Engineering , Globalfoundries , Malta , NY , USA
Evgeny A. Pidko Inorganic Materials Chemistry Group, Schuit Institute of Catalysis, Department of Chemical Engineering and Chemistry , Eindhoven University of Technology , Eindhoven , The Netherlands
Institute for Complex Molecular Systems , Eindhoven University of Technology , Eindhoven , The Netherlands
Catherine Pinel Institut de Recherches sur l’Environnement et la Catalyse de Lyon (IRCELYON) , UMR 5256, CNRS – Université Lyon 1 , Villeurbanne Cedex , France
C. V. Pramod University of Twente , Enschede , The Netherlands
David J. Robichaud National Renewable Energy Laboratory, National Bioenergy Center , Golden , CO , USA
Basudeb Saha Department of Chemistry and the Center for Catalytic Conversion of Biomass to Biofuels (C3Bio) , Purdue University , West Lafayette , IN , USA
Marcel Schlaf The Guelph-Waterloo-Centre for Graduate Work in Chemistry (GWC)2, Department of Chemistry , University of Guelph , Guelph , ON , Canada
Ulf Schuchardt Instituto de Química – UNICAMP , Campinas , SP , Brazil
K. Seshan University of Twente , Enschede , The Netherlands
Jing Yi Department of Chemistry , Purdue University , West Lafayette , IN , USA
Contributors
1© Springer Science+Business Media Singapore 2016 M. Schlaf, Z.C. Zhang (eds.), Reaction Pathways and Mechanisms in Thermocatalytic Biomass Conversion II, Green Chemistry and Sustainable Technology, DOI 10.1007/978-981-287-769-7_1
Chapter 1 Deoxydehydration (DODH) of Biomass-Derived Molecules
Shuo Liu , Jing Yi , and Mahdi M. Abu-Omar
Abstract Deoxygenation of vicinal diols and polyols, common moieties in biomass- derived molecules, represents an important chemical pathway for making chemicals from renewable biomass resources. Catalytic deoxydehydration (DODH) is a promising deoxygenation reaction that removes two adjacent hydroxyl (−OH) groups from vicinal diols in one step to generate alkenes. Since the fi rst catalytic DODH with Cp*Re(O) 3 report by Cook and Andrews in 1996, a number of metal complexes based on rhenium, ruthenium, vanadium, and molybdenum have been investigated. High-valent oxorhenium complexes are among the most effi cient catalysts for DODH reactions and have been studied using various reductants including organic phosphines, molecular hydrogen (H 2 ), sulfi te, and alcohols. These complexes exhibit intriguing oxophilic performance, which facilitates selective C-O bond cleavage of polyols. A fl urry of investigations have appeared in the literature over the past few years on the scope and mechanism of the DODH reaction in the context of biomass conversion and sustainable chemistry. In this chapter, we briefl y review the development of DODH reactions with a focus on homogenous Re-catalyzed transformations. Several heterogeneous and other metal catalysts are included for comparison.
Keywords Biomass • Deoxygenation • Deoxydehydration • Sustainable chemistry • Polyols • Rhenium • Alkenes
S. Liu • J. Yi Department of Chemistry , Purdue University , West Lafayette , IN 47907 , USA
M. M. Abu-Omar (*) Department of Chemistry and the Center for Catalytic Conversion of Biomass to Biofuels (C3Bio) , Purdue University , West Lafayette , IN 47907 , USA
Spero Energy, Inc. , 1281 Win Hentschel Blvd. , West Lafayette , IN 47906 , USA e-mail: [email protected]
1.1 Introduction
Biomass-derived molecules can be used as a sustainable platform feedstock to produce high-value chemicals (HVCs) and biofuels [ 1 – 6 ]. Since carbohydrates, the major component of biomass, contain one oxygen atom per carbon, deoxygenation of carbohydrates and their derivatives such as sugar alcohols (polyols) represents a major path forward for making renewable compounds that can potentially replace petroleum-based chemicals. Current deoxygenation methods include high- temperature pyrolysis [ 7 – 9 ], acid-catalyzed dehydration [ 10 – 13 ], hydrogenolysis reactions [ 14 , 15 ], and deoxydehydration (DODH) reactions (Fig. 1.1 ). This chapter focuses on deoxydehydration (DODH), which removes two adjacent hydroxyl (−OH) groups from vicinal diols in one step and represents an effi cient approach to making olefi ns from renewable feedstock [ 16 , 17 ].
1.2 DODH of Diols and Polyols Catalyzed by Rhenium Complexes
1.2.1 High-Valent Oxorhenium and Rhenium Carbonyl Catalysts
In a typical DODH reaction, the substrate (vicinal diol or polyol) is reduced to produce an alkene or allylic alcohol, while a reducing agent is oxidized via oxygen atom transfer (OAT). Rhenium complexes are among the most effi cient and well- studied catalysts. Cook and Andrews reported in 1996 the fi rst catalytic DODH of aromatic diols catalyzed by Cp*ReO 3 , employing stoichiometric amount of PPh 3 as the reducing agent (Fig. 1.2a ) [ 18 ]. The conversion was quite low; nevertheless, the addition of Brønsted acid enhanced the rate of reaction signifi cantly. In the presence of p -toluenesulfonic acid (TsOH), glycerol was reduced to allylic alcohol using
Fig. 1.1 General strategies for deoxygenation of biomass-derived sugar alcohols (polyols)
S. Liu et al.
3
Cp*ReO 3 , PPh 3 , and chlorobenzene in biphasic medium at 125 °C. When unpro- tected tetritol was used as substrate in this system, the main product was the fully deoxygenated butadiene (80 % yield), with the rest as 3-butene-1,2-diol and cis-2- butene-1,4-diol in an 85:15 ratio (Fig. 1.2b ). The cis-2-butene-1,4-diol isomerized to 3-butene-1,2 diol under the catalysis conditions. Previously, the Gable [ 19 – 25 ] and Herrmann groups [ 26 , 27 ] independently synthesized and characterized a number of oxorhenium(V) diolate complexes, investigated alkene extrusion from rhenium diolate complexes, and shed light on the mechanistic pathway of these reactions. The mechanism of the DODH reaction of diols by oxorhenium was suggested to proceed through three steps: (1) reduction of Cp*ReO 3 by phosphine to generate Cp*ReO 2 , (2) diol condensation with Cp*ReO 2 to generate Re(V)-diolate intermediate and water, and (3) extrusion of alkene from the Re(V)-diolate to regen- erate Cp*ReO 3 and complete the catalytic cycle.
In 2009 the Abu-Omar group reported on methyltrioxorhenium (MTO)-catalyzed DODH/deoxygenation of vicinal diols and epoxides to alkenes and alkanes with the more practical reductant H 2 (Fig. 1.3 ) [ 28 ]. Under lower H 2 pressure, the products were dominated by alkenes, whereas under higher pressure, by the alkane. This indicated an extension of the DODH reaction to generate saturated products. Several biomass-derived substrates were also tested in the system. 1,4-anhydroerytheritol, formed by acid-catalyzed ring closure of erythritol, yielded 25 % 2,5-dihydrofuran and 5 % tetrahydrofuran. However, erythritol gave signifi cant charring. The authors also noted that only cis-cyclohexane diol gave the desired products, while trans-
a
b
Fig. 1.2 DODH reactions of aromatic diol ( a ) and erythritol ( b ) catalyzed by Cp*ReO 3 employing PPh 3 as a reductant
Fig. 1.3 DODH/deoxygenation of diols and epoxides by MTO using H 2
1 Deoxydehydration (DODH) of Biomass-Derived Molecules
4
cyclohexane diol did not react, consistent with observations in other reports [ 18 ]. The specifi c stereoselectivity indicates the formation of a metal diolate intermediate, which would require cis-vicinal diols in the two adjacent hydroxyl groups.
Soon afterwards, the Nicholas group reported MTO and perrhenate salts catalyzed DODH of diols using sulfi te as the reductant (Fig. 1.4 ) [ 29 , 30 ]. Both aromatic and aliphatic diols were converted to the corresponding alkenes with good to moderate yields, while the latter required longer reaction times. Addition of the crown ether 15-crown-5 as a phase transfer agent was found to signifi cantly shorten the reaction time and increase conversions. [NBu 4 ][ReO 4 ] was a superior catalyst to MTO under these conditions in terms of conversion of erythritol. 1,3-butadiene (27 % yield), 2,5-dihydrofuran (6 % yield), and cis-2-betene-1,4-diol (3 % yield) were obtained using [NBu 4 ][ReO 4 ] as the catalyst, while substantial charring was observed for MTO.
The Bergman group employed rhenium carbonyl as the catalyst (Re 2 (CO) 10 or BrRe(CO) 5 ) and a secondary alcohol as solvent and reductant (Fig. 1.5 ) [ 31 ]. In the presence of simple alcohols such as 3-octanol or 5-nonanol, both terminal and inter- nal vicinal diols were deoxygenated to olefi ns with good yields, and a stoichiometric amount of the alcohol was oxidized to the corresponding ketone. In the presence of TsOH, erythritol could be converted to 2,5-dihydrofuran (62 % yield). Interestingly, the system required air and high temperature for activation, which indicated that the actual active catalyst is probably an oxidized rhenium species and shares a similar reaction mechanism with high-valent oxorhenium complexes.
Our group applied the MTO-catalyzed DODH reaction to glycerol, under either neat conditions or in the presence of a sacrifi cial alcohol (Fig. 1.6 ) [ 32 ]. The substrate, glycerol, participated in transfer hydrogenation and deoxygenation to form volatile products, for example, allyl alcohol, propanal, and acrolein, leaving the nonvolatile dihydroxyacetone (DHA) as by-product in the residual liquid reaction mixture. Under neat conditions, the glycerol itself can function as reductant. Utilization of NH 4 ReO 4 as the catalyst gave similar results to MTO, with allyl alcohol as the major product. The addition of acid (HCl, NH 4 Cl) increased the conversion rate and yield.
Fig. 1.4 DODH of diols by MTO using sulfi te
Fig. 1.5 DODH of diols by rhenium-carbonyl complexes in the presence of sacrifi cial alcohol
S. Liu et al.
5
We have also investigated the MTO-catalyzed DODH reaction with sacrifi cial alcohol as reductant and proposed a reaction mechanism based on detailed kinetics and spectroscopic studies [ 33 ]. The reaction kinetics were zero order in [diol] and half order in [MTO]. The different induction periods of MTO and MDO (methyldioxorhenium(V)) indicated that the active form of the catalyst was MDO, which was formed by reduction of MTO by alcohol or via a novel C-C bond cleavage of an MTO-diolate complex. The rate-determining step involved reaction with alcohol, and the majority of the MDO-diolate complex was present in dinuclear form, giving rise to the [Re catalyst] 1/2 (half-order) dependence. Furthermore, the Re(V)-diolate did not extrude alkene at rates that are consistent with the catalytic rate unless a reductant was present. Thus we proposed a catalytic cycle in which MDO-diolate was reduced further by 3-octanol to a transient rhenium(III) diolate, which was responsible for alkene extrusion and regeneration of MDO (Fig. 1.7 ). It
ReO O
Fig. 1.6 Rhenium-catalyzed DODH reactions of neat glycerol and erythritol
Fig. 1.7 Mechanism of the major pathways for MTO-catalyzed DODH reaction of vicinal diols in the presence of sacrifi cial alcohol
6
is worth noting that depending on the reducing agent, the rate-limiting step may shift under different conditions.
The reaction mechanism(s) of the MTO-catalyzed DODH of diols was also investigated by density functional theory (DFT) calculations [ 34 , 35 ]. The DFT results supported the original reaction mechanism proposed by Gable and Cook, alkene extrusion from rhenium(V) diolate being the key rate-determining step.
The Toste group extended the reaction scope to larger sugar alcohols and sugars, demonstrating the general effi ciency and high selectivity of DODH [ 36 , 37 ]. C 3 -C 6 sugar alcohols can be readily obtained by pretreatment of naturally abundant sugars, such as hydrogenation, fermentation, and decarbonylation. These sugar alcohols were converted to the corresponding olefi nic products with good yields using either secondary alcohols or primary alcohols (1-butanol) as reductants, albeit the former is more favorable than the latter (Fig. 1.8 ). The linear alkene products can be used as drop-in chemicals because they can serve as precursors for polymers and fuels that are already in use in the chemical industry. MTO displayed a higher activity than rhenium-carbonyl complexes previously reported by the Bergman group. When using a primary alcohol as the reducing agent, rhenium-carbonyl catalysts didn’t generate any product, while high-valent oxorhenium complexes afforded alkene products with high yield and selectivity.
Following their initial work (Fig. 1.8 ), Toste and Shiramizu expanded the reaction to 1,4-DODH and 1,6-DODH via tandem [1, 3]-OH shift-DODH process [ 36 ]. This process merged different reaction intermediates into one product and thereby increased product selectivity, providing a new strategy for DODH reaction development for real biomass-derived sugar conversion (Fig. 1.9 ). The reaction was also
Fig. 1.8 DODH reactions of C 3 -C 6 sugar alcohols catalyzed by MTO
S. Liu et al.
7
applied to the conversion of sugar acids into unsaturated esters, employing HReO 4 as both DODH catalyst and Brønsted acid. The DODH/acid dual-catalyst strategy was applied to produce plasticizer precursors from tartaric acid and erythritol.
1.2.2 Heterogeneous Rhenium Catalysts
There are few reports on heterogeneous DODH reactions. Jentoft and Nicholas reported on heterogeneous polyol-into-olefi n DODH reactions catalyzed by carbon- supported perrhenate, employing both H 2 and hydrogen-transfer reductants with moderate yield [ 38 ]. Interestingly, in 2011 the Schlaf group reported that stainless steel reactors could catalyze the deoxygenation of glycerol and levulinic acid in aqueous acidic medium [ 39 ]. Ferdi Schuth reported an iron oxide-catalyzed conversion of glycerol to allylic alcohol and proposed a mechanism through dehydration and consecutive hydrogen transfer [ 40 ]. Andreas Martin also reported glycerol deoxygenation reaction in the gas phase using a series of heteropolyacid catalysts [ 41 ]. Ir-ReOx/SiO 2 was also reported as the catalyst for the production of butanediol from erythritol via hydrogenolysis, by the Tomishige group [ 42 ]. More recently, Nicholas reported on deoxydehydration of glycols using heterogeneous elemental reductants such as zinc, iron, manganese, and carbon [ 43 ]. The molecular identity of the catalyst is more diffi cult to discern in these systems and the contribution from the metal surface in the reactor per Schlaf’s study remains an open question. Nevertheless, these studies demonstrate the feasibility to translating the DODH reaction into a more practical process in which continuous fl uidized bed reactors can be employed.
1.3 Other Transition Metal Catalysts for DODH
In addition to rhenium, other transition metal complexes based on vanadium [ 44 ], molybdenum [ 45 , 46 ], and ruthenium [ 47 , 48 ] have been used in DODH of vicinal diols. Nicholas reported a DODH reaction of diols to olefi ns catalyzed by
Fig. 1.9 1,4-DODH and 1,6-DODH reactions via tandem [1, 3]-OH shift-DODH process
8
inexpensive metavanadate (VO 3 − ) and chelated dioxovanadium derivatives, with phosphine or sulfi te as reductants [ 44 ]. Dioxomolybdenum(VI) complexes with acylpyrazolonate ligands were synthesized by the Pettinari group, and these complexes showed moderate activity toward diol DODH reactions with PPh 3 as reductant [ 45 ]. The Fristrup group also reported another DODH reaction catalyzed by a series of Mo-oxo complexes under neat conditions [ 46 ]. In addition, Schlaf and Bullock have pioneered the use of organometallic ruthenium catalysts for the deoxygenation of alcohols [ 49 , 50 ]. Two other Ru-catalyzed DODH reactions were reported by Srivastava and Nicholas, independently, using [Cp*Ru(CO) 2 ] 2 or Ru(II)- sulfoxides as catalysts for hydrodeoxygenation (HDO) and hydrocracking of diols and epoxides [ 47 , 48 ].
1.4 Conclusion
As one of the promising strategies to selective removal of oxygen atoms from biomass- derived molecules, DODH shows tremendous potential to producing important fi ne and commodity chemicals from renewable feedstock. Compared to conventional deoxygenation methods such as hydrogenation and hydrogenolysis, the advantages of DODH reaction are that the products are deoxygenated while remaining synthetically versatile through the retention of a double bond, making the most use of every atom in biomass. The successful development of stable and highly active homogenous and heterogeneous catalysts provides access to various synthetic procedures to produce specifi c molecules with tailored properties to be used as drop-in chemicals for making polymers and other useful products. However, many challenges remain in utilizing biomass-derived molecules and feedstock. Some of these challenges include the ability to use carbohydrates and polysaccha- rides directly and developing cost-effective, effi cient, and active earth-abundant catalysts that can be employed under continuous reaction conditions with facile product/catalyst separation.
Acknowledgment Our research in deoxygenation of biomass-derived molecules has been support by the Department of Energy (DOE), Basic Energy Sciences (BES) grant no. DE-FG-02-06ER15794.
References
1. Huber GW, Iborra S, Corma A (2006) Synthesis of transportation fuels from biomass: chemistry, catalysts, and engineering. Chem Rev 106(9):4044–4098. doi: 10.1021/cr068360d
2. Serrano-Ruiz JC, Luque R, Sepulveda-Escribano A (2011) Transformations of biomass- derived platform molecules: from high added-value chemicals to fuels via aqueous-phase processing. Chem Soc Rev 40(11):5266–5281. doi: 10.1039/C1CS15131B
S. Liu et al.
3. Alonso DM, Bond JQ, Dumesic JA (2010) Catalytic conversion of biomass to biofuels. Green Chem 12(9):1493–1513. doi: 10.1039/C004654J
4. Gallezot P (2012) Conversion of biomass to selected chemical products. Chem Soc Rev 41(4):1538–1558. doi: 10.1039/C1CS15147A
5. Huber GW, Corma A (2007) Synergies between bio- and oil refi neries for the production of fuels from biomass. Angew Chem Int Ed 46(38):7184–7201. doi: 10.1002/anie.200604504
6. Chheda JN, Huber GW, Dumesic JA (2007) Liquid-phase catalytic processing of biomass- derived oxygenated hydrocarbons to fuels and chemicals. Angew Chem Int Ed 46(38):7164– 7183. doi: 10.1002/anie.200604274
7. Zhang Q, Chang J, Wang T, Xu Y (2007) Review of biomass pyrolysis oil properties and upgrading research. Energy Convers Manage 48(1):87–92. doi: http://dx.doi.org/10.1016/j. enconman.2006.05.010
8. Mohan D, Pittman CU, Steele PH (2006) Pyrolysis of wood/biomass for bio-oil: a critical review. Energy Fuel 20(3):848–889. doi: 10.1021/ef0502397
9. Mettler MS, Vlachos DG, Dauenhauer PJ (2012) Top ten fundamental challenges of biomass pyrolysis for biofuels. Energy Environ Sci 5(7):7797–7809. doi: 10.1039/C2EE21679E
10. Jin F, Enomoto H (2011) Rapid and highly selective conversion of biomass into value-added products in hydrothermal conditions: chemistry of acid/base-catalysed and oxidation reactions. Energy Environ Sci 4(2):382–397. doi: 10.1039/C004268D
11. Tong X, Ma Y, Li Y (2010) Biomass into chemicals: conversion of sugars to furan derivatives by catalytic processes. Appl Catal Gen 385(1–2):1–13. doi: http://dx.doi.org/10.1016/j. apcata.2010.06.049
12. Rinaldi R, Schuth F (2009) Design of solid catalysts for the conversion of biomass. Energy Environ Sci 2(6):610–626. doi: 10.1039/B902668A
13. Akien GR, Qi L, Horvath IT (2012) Molecular mapping of the acid catalysed dehydration of fructose. Chem Commun 48(47):5850–5852. doi: 10.1039/C2CC31689G
14. Corma A, Iborra S, Velty A (2007) Chemical routes for the transformation of biomass into chemicals. Chem Rev 107(6):2411–2502. doi: 10.1021/cr050989d
15. Du X-L, Bi Q-Y, Liu Y-M, Cao Y, He H-Y, Fan K-N (2012) Tunable copper-catalyzed che- moselective hydrogenolysis of biomass-derived [gamma]-valerolactone into 1,4-pentanediol or 2-methyltetrahydrofuran. Green Chem 14(4):935–939. doi: 10.1039/C2GC16599F
16. Dutta S (2012) Deoxygenation of biomass-derived feedstocks: hurdles and opportunities. ChemSusChem 5(11):2125–2127. doi: 10.1002/cssc.201200596
17. Metzger JO (2013) Catalytic deoxygenation of carbohydrate renewable resources. ChemCatChem 5(3):680–682. doi: 10.1002/cctc.201200796
18. Cook GK, Andrews MA (1996) Toward nonoxidative routes to oxygenated organics: stereo- specifi c deoxydehydration of diols and polyols to alkenes and allylic alcohols catalyzed by the metal oxo complex (C5Me5)ReO3. J Am Chem Soc 118(39):9448–9449. doi: 10.1021/ ja9620604
19. Gable KP, Phan TN (1994) Extrusion of alkenes from rhenium(V) diolates: energetics and mechanism. J Am Chem Soc 116(3):833–839. doi: 10.1021/ja00082a002
20. Gable KP (1994) Condensation of vicinal diols with the oxo complex {Cp*Re(O)}2(.mu.-O)2 giving the corresponding diolate complexes. Organometallics 13(6):2486–2488. doi: 10.1021/ om00018a048
21. Gable KP, Juliette JJJ (1996) Hammett studies on alkene extrusion from rhenium(V) diolates and an MO description of metal alkoxide−alkyl metal oxo interconversion. J Am Chem Soc 118(11):2625–2633. doi: 10.1021/ja952537w
22. Gable KP, Zhuravlev FA (2002) Kinetic Isotope effects in cycloreversion of rhenium (V) diolates. J Am Chem Soc 124(15):3970–3979. doi: 10.1021/ja017736w
23. Gable KP, AbuBaker A, Zientara K, Wainwright AM (1998) Cycloreversion of rhenium(V) diolates containing the hydridotris(3,5-dimethylpyrazolyl)borate ancillary ligand. Organometallics 18(2):173–179. doi: 10.1021/om980807o
24. Kevin PG, Brian R (2006) Improved catalytic deoxygenation of vicinal diols and application to alditols. In: Feedstocks for the future, vol 921, ACS symposium series. American Chemical Society, Washington, DC, pp 143–155. doi: 10.1021/bk-2006-0921.ch011
25. Gable KP, Juliette JJJ (1995) Extrusion of alkenes from rhenium(V) diolates: the effect of substitution and conformation. J Am Chem Soc 117(3):955–962. doi: 10.1021/ja00108a012
26. Herrmann WA, Marz D, Herdtweck E, Schäfer A, Wagner W, Kneuper H-J (1987) Glycolate and thioglycolate complexes of rhenium and their oxidative elimination of ethylene and of glycol. Angew Chem Int Ed Engl 26(5):462–464. doi: 10.1002/anie.198704621
27. Herrmann WA, Marz DW, Herdtweck E (1990) Mehrfachbindungen zwischen hauptgrup- penelementen und übergangsmetallen: LXXVIII. Über oxo- und methylimido-komplexe des rheniums mit sauerstoff-, schwefel- und stickstoffchelaten: Synthese, spaltungsreaktionen und strukturchemie. J Organomet Chem 394(1–3):285–303. doi: http://dx.doi. org/10.1016/0022-328X(90)87239-A
28. Ziegler JE, Zdilla MJ, Evans AJ, Abu-Omar MM (2009) H2-driven deoxygenation of epoxides and diols to alkenes catalyzed by methyltrioxorhenium. Inorg Chem 48(21):9998–10000. doi: 10.1021/ic901792b
29. Vkuturi S, Chapman G, Ahmad I, Nicholas KM (2010) Rhenium-catalyzed deoxydehydration of glycols by sulfi te. Inorg Chem 49(11):4744–4746. doi: 10.1021/ic100467p
30. Ahmad I, Chapman G, Nicholas KM (2011) Sulfi te-driven, oxorhenium-catalyzed deoxydehydration of glycols. Organometallics 30(10):2810–2818. doi: 10.1021/om2001662
31. Arceo E, Ellman JA, Bergman RG (2010) Rhenium-catalyzed didehydroxylation of vicinal diols to alkenes using a simple alcohol as a reducing agent. J Am Chem Soc 132(33):11408– 11409. doi: 10.1021/ja103436v
32. Yi J, Liu S, Abu-Omar MM (2012) Rhenium-catalyzed transfer hydrogenation and deoxygenation of biomass-derived polyols to small and useful organics. ChemSusChem 5(8):1401– 1404. doi: 10.1002/cssc.201200138
33. Liu S, Senocak A, Smeltz JL, Yang L, Wegenhart B, Yi J, Kenttämaa HI, Ison EA, Abu-Omar MM (2013) Mechanism of MTO-catalyzed deoxydehydration of diols to alkenes using sacrifi cial alcohols. Organometallics 32(11):3210–3219. doi: 10.1021/om400127z
34. Bi S, Wang J, Liu L, Li P, Lin Z (2012) Mechanism 1: mechanism of the MeReO3-catalyzed deoxygenation of epoxides. Organometallics 31(17):6139–6147. doi: 10.1021/om300485w
35. Qu S, Dang Y, Wen M, Wang Z-X (2013) Mechanism 2: mechanism of the methyltrioxorhenium- catalyzed deoxydehydration of polyols: a new pathway revealed. Chem Eur J 19(12):3827– 3832. doi: 10.1002/chem.201204001
36. Shiramizu M, Toste FD (2012) Deoxygenation of biomass-derived feedstocks: oxorhenium- catalyzed deoxydehydration of sugars and sugar alcohols. Angew Chem Int Ed 51(32):8082– 8086. doi: 10.1002/anie.201203877
37. Shiramizu M, Toste FD (2013) Expanding the scope of biomass-derived chemicals through tandem reactions based on oxorhenium-catalyzed deoxydehydration. Angew Chem Int Ed 52(49):12905–12909. doi: 10.1002/anie.201307564
38. Denning AL, Dang H, Liu Z, Nicholas KM, Jentoft FC (2013) Deoxydehydration of glycols catalyzed by carbon-supported perrhenate. ChemCatChem 5(12):3567–3570. doi: 10.1002/ cctc.201300545
39. Di Mondo D, Ashok D, Waldie F, Schrier N, Morrison M, Schlaf M (2011) Stainless steel as a catalyst for the total deoxygenation of glycerol and levulinic acid in aqueous acidic medium. ACS Catal 1(4):355–364. doi: 10.1021/cs200053h
40. Liu Y, Tuysuz H, Jia C-J, Schwickardi M, Rinaldi R, Lu A-H, Schmidt W, Schuth F (2010) Iron oxide: from glycerol to allyl alcohol: iron oxide catalyzed dehydration and consecutive hydrogen transfer. Chem Commun 46(8):1238–1240. doi: 10.1039/B921648K
41. Atia H, Armbruster U, Martin A (2008) Poly acid: dehydration of glycerol in gas phase using heteropolyacid catalysts as active compounds. J Catal 258(1):71–82. doi: http://dx.doi. org/10.1016/j.jcat.2008.05.027
S. Liu et al.
43. Michael McClain J, Nicholas KM (2014) Elemental reductants for the deoxydehydration of glycols. ACS Catal 4(7):2109–2112. doi: 10.1021/cs500461v
44. Chapman G, Nicholas KM (2013) Vanadium-catalyzed deoxydehydration of glycols. Chem Commun 49(74):8199–8201. doi: 10.1039/C3CC44656E
45. Hills L, Moyano R, Montilla F, Pastor A, Galindo A, Álvarez E, Marchetti F, Pettinari C (2013) Dioxomolybdenum(VI) complexes with acylpyrazolonate ligands: synthesis, structures, and catalytic properties. Eur J Inorg Chem 2013(19):3352–3361. doi: 10.1002/ejic.201300098
46. Dethlefsen JR, Lupp D, Oh B-C, Fristrup P (2014) Molybdenum-catalyzed deoxydehydration of vicinal diols. ChemSusChem 7(2):425–428. doi: 10.1002/cssc.201300945
47. Stanowski S, Nicholas KM, Srivastava RS (2012) [Cp*Ru(CO)2]2-catalyzed hydrodeoxygenation and hydrocracking of diols and epoxides. Organometallics 31(2):515–518. doi: 10.1021/ om200447z
48. Murru S, Nicholas KM, Srivastava RS (2012) Ruthenium (II) sulfoxides-catalyzed hydrogenolysis of glycols and epoxides. J Mol Catal Chem 363–364(0):460–464. doi: http://dx.doi. org/10.1016/j.molcata.2012.07.025
49. Thibault ME, DiMondo DV, Jennings M, Abdelnur PV, Eberlin MN, Schlaf M (2011) Cyclopentadienyl and pentamethylcyclopentadienyl ruthenium complexes as catalysts for the total deoxygenation of 1,2-hexanediol and glycerol. Green Chem 13(2):357–366. doi: 10.1039/ C0GC00255K
50. Ghosh P, Fagan PJ, Marshall WJ, Hauptman E, Bullock RM (2009) Synthesis of ruthenium carbonyl complexes with phosphine or substituted Cp ligands, and their activity in the catalytic deoxygenation of 1,2-propanediol. Inorg Chem 48(14):6490–6500. doi: 10.1021/ic900413y
Chapter 2 Homogeneous Catalysts for the Hydrodeoxygenation of Biomass- Derived Carbohydrate Feedstocks
Marcel Schlaf
Abstract The use of homogeneous rather than heterogeneous catalysts for the hydrodeoxygenation of sugars, sugar alcohols, and their condensates such as furfural, 5-hydroxymethylfurfural, levulinic acid, and iso sorbide may offer reaction pathways that have distinct advantages, notably with respect to catalyst deactivation by coking and fouling as observed on many heterogeneous systems with these highly reactive and polar substrates. Homogeneous systems, however, also face unique challenges in ligand, catalyst, and process design. The catalyst systems employed will have to be stable to the required aqueous acidic high-temperature (T > 150 °C) reaction conditions while exhibiting activities that make them kinetically competent over acid-catalyzed decomposition and oligo- and polymerization reactions leading to humin formation. For each of the hydrodeoxygenation reaction cascades for the C3 (glycerol), C4 (erythritol), C5 (xylose and derivatives or levulinic acid), and C6 (glucose and derivatives) value chains, comparatively few homogeneous catalyst systems have been evaluated to date. Key issues remain the thermal and redox stability of the complexes employed against decomposition and reduction to bulk metal acting as a heterogeneous catalysts and the recovery and recycling of the catalyst from the often very complex reaction and product mixtures.
Keywords Homogeneous catalysis • Hydrodeoxygenation • Sugar alcohols • Carbohydrate biomass • Ruthenium-based complexes
M. Schlaf (*) The Guelph-Waterloo-Centre for Graduate Work in Chemistry (GWC)2, Department of Chemistry , University of Guelph , Guelph , ON , Canada e-mail: [email protected]
2.1.1 The Fundamental Challenges of Biomass Hydrodeoxygenation
Carbohydrate biomass, i.e., essentially sugars and their polymers and derivatives, is, as the name implies, constituted of molecules of the general composition C n + n H 2 O and therefore characterized by a high oxygen content – mainly in the form of hydroxyl functions with a lower relative abundance of carbonyls and ether or acetal and ketal linkages. This high oxygen content imparts a high and typically nonspecifi c reactivity and at the same time comparatively low energy density on the material. Hydrodeoxygenation effects the net removal of oxygen from a biomass-derived substrate to yield products of lower oxygen content and reactivity and higher energy density that – for a complete removal of oxygen – is equivalent to that of fossil hydrocarbons used as liquid fuels.
An actual hydrodeoxygenation of carbohydrate-derived substrates faces two fundamental challenges that have to be met for an economically viable and ecologically sustainable use of biomass as a source of chemicals and fuels:
(a) The depolymerization and separation of recalcitrant biomass, in particular the most abundant carbohydrates cellulose and hemicellulose, into molecularly well-defi ned and soluble substrates that can then be targeted for conversion to value-added products.
(b) The partial selective or total catalytic hydrodeoxygenation of soluble carbohydrate monomers – glycerol (C3), erythrose/erythritol (C4), xylose/xylitol (C5), and glucose/sorbitol (C6) – or their condensates, e.g., furfural, levulinic acid or 5-hydroxymethylfurfural (HMF), and isosorbide, respectively, to less polar and less reactive products that can serve directly as polymer components, solvents, or fuels and so seamlessly integrate into existing large-scale (petro-) chemical feed streams.
The application of transition metal-based homogeneous catalyst systems to the second challenge is the subject of this chapter. From a purely chemical viewpoint, this challenge is limited to only a handful of reactions detailed below that for a given substrate must typically occur in an iterative reaction cascade and specifi c value chain to yield the desired deoxygenated products.
2.1.2 Reaction Patterns for the Hydrodeoxygenation of Carbohydrate Biomass-Derived Substrates to Chemicals and Fuels
An effective removal of oxygen from highly oxygenated substrates is only possible by rejection of either H 2 O or CO 2 . The former is applicable to (poly-)alcohols by acid-catalyzed dehydration and the latter to carboxylic acids, e.g., levulinic acid,
M. Schlaf
15
which is ultimately obtainable from cellulose via 5-hydroxymethylfurfural ( vide infra ). In a second step the unsaturation or ring formation caused by the loss of water is then followed by metal-catalyzed hydrogenation or hydrogenolysis resulting in a net removal of an oxygen atom from the substrate. Using the linear sugar alcohols as an example, the reaction principle is illustrated in Fig. 2.1 and can – when carried out in an iterative reaction cascade – result in the total deoxygenation of the substrates to alkenes or alkanes.
The type of functional groups present in carbohydrates and substrates derived from them by acid-catalyzed dehydration is limited to hydroxyl, carbonyl, carboxyl, alkene, and cyclic ether functions, which in turn limits the number of fundamental reaction patterns required for hydrodeoxygenation to the comparatively short and deceptively simple (!) list presented in Fig. 2.2 [ 1 ]. Ideally any catalyst system tar- geting the hydrodeoxygenation of such substrate should be promiscuous, i.e., be at least be capable of catalyzing reactions i–iii shown in this fi gure, as this would enable an actual process in a single reaction vessel. An obvious potential limitation of hydrodeoxygenation reaction cascades assembled from these fundamental reactions is the propensity of the carbonyl and alkene intermediates formed to react with themselves leading to uncontrolled oligomerization or polymerization side reactions resulting in humin formation. The side reactions are catalyzed by the necessarily present acid, which then effectively reverse the depolymerization of (hemi-) cellulose required to obtain soluble hydrodeoxygenation target substrates in the fi rst place, i.e., negating any effort expended to meet challenge (a) discussed previously.
Any viable catalyst system and process must therefore be kinetically competent with respect to these side reactions by converting the reactive alkene and carbonyl intermediates to more stable oxacycles, primary alcohols and alkanes faster than they can nonspecifi cally (re-)condense to insoluble macromolecules. Regardless of their homo- or heterogeneous nature, this constitutes the main problem in the design and development of catalyst systems and processes for a successful hydrodeoxygenation of highly reactive biomass-derived carbohydrate substrates.
Fig. 2.1 Hydrodeoxygenation as an iterative reaction of acid-catalyzed dehydrations and metal- catalyzed hydrogenations resulting in the net loss of oxygen
2 Homogeneous Catalysts for the Hydrodeoxygenation of Biomass-Derived…
16
2.2 Why Homogeneous?
2.2.1 Historic Perspective
Beginning with the historic development of processes for the production of xylitol and sorbitol by hydrogenation of xylose and glucose over Raney Nickel [ 2 ], the reductive cyclization of levulinic acid to γ-valerolactone over PtO 2 by Sabatier and Schuette [ 3 , 4 ], and the more recent seminal paper by Descotes describing the hydrogenation of 5-hydroxymethylfurfural in aqueous medium over Ni, CuO•CrO 3 , Pd/C, Pt/C, Ru/C, PtO 2 , as well as Ru and Pt metal [ 5 ], the development of hydrodeoxygenation catalysts has mainly focused on heterogeneous systems employing various combinations of the same hydrogenating metals. Typically the metals are supported on metal oxides, e.g., Al 2 O 3 , SiO 2 , ZrO 2 , or Nb 2 O 3 , which often also play the role of the acid component of the system, as can carbon supports bearing acidic surface functionalities (e.g., carboxylic or sulfonic acids). Alternatively a solid acid such as WO 3 can be added separately . Substantial progress has been made in this endeavor and the fi eld has been extensively reviewed [ 6 – 13 ]. Comparatively much
Fig. 2.2 Fundamental reactions for the hydrodeoxygenation of carbohydrate-derived substrates
M. Schlaf
17
fewer efforts have been directed at the development of hydrodeoxygenation catalyst systems, in which both the acid and the hydrogenating metal are present in homogeneous molecular dispersed phase [ 14 ].
2.2.2 Heterogeneous vs. Homogeneous Catalysts for Hydrodeoxygenation
As in any catalytic reaction or process, the use of either hetero- or homogeneous systems has distinct advantages and disadvantages. Adopting and expanding a comparison originally given by Cornils and Herrmann [ 15 ], Table 2.1 lists these with special consideration given to the unique properties of the highly polar carbohydrate substrates targeted.
As a consequence of the necessarily aqueous acidic reaction conditions encountered in acid/metal-catalyzed hydrodeoxygenation reaction cascades (Fig. 2.1 ) and the undifferentiated high reactivity of the substrates, heterogeneous catalysts are susceptible to challenges that are typically not – or only to a much lesser extent – encountered in their applications to fossil carbon-derived hydrocarbon feeds for which most of them were originally developed. In particular, they are vulnerable to leaching of the active hydrogenating metal into solution, degradation, or complete destruction of oxide supports such as SiO 2 , Al 2 O 3 or similar by the corrosive aqueous acidic reaction medium under what are effectively hydrothermal conditions and/or rapid fouling/coking and hence deactivation due to condensation of the highly polar substrates on the also polar catalyst surface [ 13 ]. The latter is in principle reversible
Table 2.1 Advantages and disadvantages of hetero- vs. homogeneous catalysts applied to the hydrodeoxygenation of biomass-derived carbohydrate substrates
Heterogeneous catalysis Homogeneous catalysis
Activity (rel. to metal content) Variable High Selectivity Variable Variable Reaction conditions Harsh Moderate to harsh Service life Long Variable Sensitivity to poisons High Low Diffusion limitations Can be a limiting factor None Catalyst recycling Not necessary Very challenging Temperature stability High Needs to be “designed in” Stability against aqueous acids
Typically low due to leaching and support degradation
Needs to be “designed in”
Sensitivity to coking High None Sensitivity to fouling High None Coordinative inhibition None Can be a limiting factor Electronic and steric design Not possible Rationally changeable Mechanistic understanding Very diffi cult Plausible and achievable
18
by oxidative reactivation (i.e., burn-off of deposited carbonaceous material followed by reduction), but is however – by defi nition – not applicable to the very promising carbon-supported catalysts systems, e.g., Ru/C.
Homogeneous catalysts will not suffer from any of these limitations, but face their own challenges when applied to hydrodeoxygenation reaction cascades. Empirically the acid-catalyzed loss of water from polyalcohol substrates (reaction i in Fig. 2.2 ) or hydrolysis of furan rings (as encountered, e.g., in HMF and its derivatives) to α,γ-diones requires temperatures in excess of 150 °C, while a direct ring opening of THF-type substrates will require 200–300 °C at elevated pressures [ 16 ], with all of these reactions occurring in an aqueous acidic medium. Together these reaction conditions – listed as “moderate to harsh” rather than “mild” in Table 2.1 – will require the rational design and synthesis of transition metal complex catalysts of unprecedented thermal , acid , and general chemical stability while at the same time offering satisfactory activity toward the hydrogenation of C = C and C = O bonds at reasonably accessible pressures of hydrogen gas. This constitutes the fi rst major challenge to the use of homo- rather than heterogeneous catalyst systems for hydrodeoxygenation and can be addressed through the use of highly chelating ligands that result in very high complex formation constants [ 17 ]. The main role of the ligand in this instance is the stabilization of the metal center against reduction to oxidation state zero and precipitation of bulk metal (effectively generating a heterogeneous catalyst) rather than the induction of chemo-, regio-, or stereoselectivity as usually sought after in homogeneously catalyzed reactions.
The second major challenge to an economically and technically viable use of homogeneous catalysts is the requirement for their effective and facile reuse and recycling, in particular when considering that the catalysts will likely have to be based on platinum group metals, e.g., Ru, Ir, Pd, or Pt. The desired hydrodeoxygenation reaction products will – especially if a total deoxygenation to alkenes or alkanes is realized – have a much lower polarity and solubility in polar solvents than the sugar(-derived) starting material. This very feature, identifi ed as a problem for heterogeneous systems, can work in favor of homogeneous catalyst systems, provided they are designed as polar solvent-soluble salts (ideally in water), e.g., by combining a cationic hydrogen-activating transition metal complex with a hydrolysis- stable non-coordinating counter anion such as phosphate, sulfate, or trifl uoromethanesulfonate (“trifl ate”) [ 18 ].
Figure 2.3 illustrates a – to date unrealized – vision, in which the polarity and solubility differences between the highly polar substrates and catalyst(s) and the nonpolar deoxygenated products are exploited leading to an “automatic” product isolation and catalyst system recovery and recycling by phase separation using the total deoxygenation of sorbitol to hexane as a conceptual example. Even in cases where the hydrodeoxygenation reaction cascades lead to high-value-added products that retain one or two oxygen atoms, e.g., tetrahydrofuran, α,ω-diols, or mono-ols, this strategy may still be viable, if the deoxygenates can be recovered by (azeo- tropic) distillation from the aqueous mixture regenerating the acid/catalyst solution. With product isolation and catalyst recycling by phase separation, even a continuous process may become possible.
M. Schlaf
2.3.1 Hydrodeoxygenations of C2 to C6 Substrates
2.3.1.1 Ethanol
Due its high production volume, to date mainly from corn starch (US) and sugar- cane (Brazil), but anticipated to shift to cellulosic feeds in the near future, ethanol is an attractive starting material. While not strictly a hydrodeoxygenation reaction in the sense defi ned in Fig. 2.1 , the recent revival of the Guerbet reaction [ 19 – 21 ], which is conducted under basic rather than acidic conditions, opens a pathway for the conversion to n-butanol. As a fuel this is substantially superior to ethanol and arguably more importantly can serve as an entry point into the C4 petrochemical product manifold. A conceivable simplifi ed catalytic cycle for the Guerbet reaction of ethanol to 1-butanol with [Ru(H) 2 (dppm) 2 ] as an actually used catalyst is proposed in Fig. 2.4 [ 22 ]. The reaction employs the concept of borrowed hydrogen [ 23 ], in which a transition metal complex fi rst acts as the dehydrogenation catalysts for the
Fig. 2.3 A vision for a homogeneously catalyzed hydrodeoxygenation of sugar(-derived) substrates by homogeneous catalysts systems with product isolation and catalyst recycling effected by phase separation using the sorbitol to hexane transformation as an example
20
conversion of the alcohol to acetaldehyde, temporarily stores hydrogen in form of hydride ligands, and then acts as a hydrogenation catalyst for the re-addition of hydrogen to the α,β-unsaturated crotonaldehyde resulting from the aldol condensation of two equivalents of acetaldehyde. Iridium – as well as ruthenium-based diphosphine and phenanthroline chelate complexes – has been successfully employed realizing selectivities > 90 % at up to 20 % conversion [ 22 , 24 – 26 ].
2.3.1.2 Glycerol
Due to the large amounts of glycerol generated as the by-product of biodiesel production by the transesterifi cation of triglycerides with methanol or ethanol, glycerol is a very attractive target for conversion to value-added deoxygenated, carbonylated, or otherwise modifi ed products. A hydrodeoxygenation reaction cascade and the products obtainable in principle from glycerol (excluding polymerization) are shown in Fig. 2.5 . Glycerol (in very pure, i.e., food-grade form) itself has direct applications, e.g., in cosmetics; however, its singly deoxygenated derivatives, 1,2-propanediol and in particular 1,3-propanediol [ 27 ], the key component for the manufacture of polytrimethylene terephthalate (PTT) and polytrimethylene glycol, offer substantial value addition, while the doubly or totally deoxygenated products
Fig. 2.4 Proposed mechanism of the Guerbet reaction of ethanol to 1-butanol using [Ru(H) 2 (dppm) 2 ] as an example homogeneous catalyst
M. Schlaf
21
propanol and propene/propane are, while still useful, less desirable due the relatively higher amount of hydrogen required for their production and lower relative value.
Evident from Fig. 2.5 is that a successful acid-/metal-catalyzed glycerol hydrodeoxygenation process resulting in conversion and selectivity to a single product in high yield is challenging and requires a very careful adjustment and tuning of the catalyst and reaction parameters in order to select a single dominant reaction chan- nel from the cascade. Under aqueous acidic conditions, the double or total deoxygenation via the acrolein pathway leading beyond the desired diol products is of special concern. Research efforts have again been mainly focused on heterogeneous catalysts and processes [ 28 – 32 ]; however, glycerol was also the subject of the earli- est work on a homogeneously catalyzed hydrodeoxygenation with a solid acid and the seminal work of Braca et al. employing ruthenium iodocarbonyl complexes (Fig. 2.6 ) [ 33 , 34 ]. For both reactions relatively high pressures of carbon monoxide serve as a source of carbonyl ligands stabilizing the metal against reduction to oxidation state zero. A further similar example is a system composed of a palladium diphosphine complex and (trifl uro)methane sulfonic acid patented by Shell (E. Drent) that achieves comparable conversions to 1,3-propanediol in water/sulfolane mixtures under similar reaction conditions (170 °C, 6 MPa CO/H 2 = 1:2) [ 35 ].
The generation of small amounts of THF by the ruthenium system reported by Braca (Fig. 2.6 , bottom) is intriguing, as it points the way for a possible synthesis of a very valuable C4 product manifold by the deoxycarbonylation rather hydrodeoxygenation of cheap and abundant glycerol. This approach was originally described in a Japanese patent by Nakamura and has recently been further explored using
Fig. 2.5 Reaction cascade and pathways for the hydrodeoxygenation of glycerol to value-added products
22
rhodium or iridium iodocarbonyl complexes generated in situ from the corresponding chlorides and methyl iodide in acetic acid/water mixtures at 130–180 °C and CO pressures of 30–120 bar. Depending on the actual conditions using various blends of allyl acetate, (iso-)butyric acid and vinyl acetic and vinyl crotonic acid can be obtained [ 36 – 38 ].
A series of ruthenium-based complexes known or postulated to operate as ionic hydrogenation catalysts [ 39 ], i.e., by a heterolytic activation of hydrogen gas into a hydride ligand and, under aqueous conditions, H 3 O + as the strongest possible solvent- leveled acid in the reaction mixture was developed by Bullock and the author and tested as catalysts against terminal diols as model systems for glycerol, achieving – depending on reaction conditions – partial or total hydrodeoxygenation to either primary alcohols or alkenes/alkanes, respectively. The deoxygenation of diols proceeds by the reactions (i) and (ii) in Fig. 2.2 . Figure 2.7 proposes a heterolytic dihydrogen activation mechanism, in which an aqua ligand, e.g., in the complexes cis -[(6,6′-Cl 2 -2,2′-bipyridine) 2 Ru(H 2 O) 2 ][OTf] 2 [ 40 – 42 ] or [(4′-Ph-terpyridine)Ru(H 2 O) 3 ](OTf) 2 [ 43 ], is displaced by a dihydrogen ligand forming a transient highly acidic (η 2 -H 2 ) complex that is deprotonated by water to give the corresponding hydride as the active reductant. In this context it should be noted that the metal–ligand bond enthalpy of the M-(η 2 -H 2 ) unit is estimated to be very similar to that of M-(OH 2 ) and that the direct generation of nonclassical dihydrogen complexes from aqua complexes has been demonstrated even in water as the solvent [ 44 , 45 ]. Water-soluble cationic aqua complexes with weakly coordinating
Fig. 2.6 First examples of homogeneously catalyzed hydrodeoxygenation transformations of glycerol
M. Schlaf
23
counter ions may therefore be regarded as almost ideal pro-catalysts for the homogeneously catalyzed hydrodeoxygenation of polyalcohols in aqueous medium. Displacement of a second water ligand (in either sequence) gives a hydride ketone/ aldehyde (or alkene as relating to reaction ii in Fig. 2.2 ) complex that reacts by insertion of the C = O (or C = C) bond into the metal-hydride bond. Protonation of the resulting alkoxide (or alkyl) complexes releases the product restarting the cycle.
The complexes tested however all failed to achieve the most desirable conversion of glycerol to 1,3-propanediol [ 40 , 43 , 46 – 49 ]. Instead the complex [(4′-Ph-terpyridine)Ru(H 2 O) 3 ](OTf) 2 effects – depending on reaction temperature, hydrogen pressure, solvent, and added HOTf acid concentration – a partial deoxygenation to n -propanol in up to 35 % or total deoxygenation of glycerol to propene/ propane in quantitative yield [ 43 ].
A successful selective deoxygenation to 1,3-propanediol was ultimately realized by heterogeneous systems developed by Tomishige [ 50 ] and relevant to this chapter using an iridium pincer system patented by Heinekey and Goldberg [ 51 , 52 ]. In an aqueous acidic 1,4-dioxane solution at 200 °C and 8.1 MPa, the complex [(κ 3 -C 6 H 3 - 1,3-[OP( t Bu) 2 ] 2 )Ir(CO)] reaches conversions of up to 45 % with a product selectivity of 1:4 of 1,3-propanediol/ n -propanol. As with the ruthenium complexes, this system is postulated to operate as an ionic hydrogenation catalyst with a heterolytic activation of hydrogen gas split into a proton, taken up by a base from the reaction medium (water, solvent, substrate) and hydride ligand on the iridium center. The proposed catalytic cycle with and without explicit formulation of an acidic η 2 - H 2 ligand is shown in Fig. 2.8 and begins with the protonation of the neutral pro- catalyst to what is formally an Ir III complex as the hydrogen-activating species.
Fig. 2.7 Proposed mechanism for the heterolytic activation of dihydrogen and hydrogenation of carbonyl and alkene bonds by ruthenium aquo complexes
24
A remarkable metal-free glycerol hydrodeoxygenation, in which formic acid acts as both the acid catalyst and the reductant, achieves a conversion to allyl alcohol in 80 % yield at 230 °C [ 53 ]. The reaction is also applicable to the conversion of erythritol to 2,5-dihydrofuran, and a proposed mechanism is shown in Fig. 2.9 .
2.3.1.3 Erythritol
The comparatively lower availability and hence higher price of erythritol, which is typically prepared by yeast fermentation of glucose (e.g., by Moniliella pollinis ), appear to date have limited its use as a hydrodeoxygenation target. However, if
Fig. 2.8 Proposed mechanism for the conversion of glycerol to 1,3-propanediol by [(κ 3 -C 6 H 3 -1,3- [OP( t Bu) 2 ] 2 )Ir(CO)] in aqueous acidic solution under hydrogen pressure
Fig. 2.9 Proposed mechanism of the metal-free hydrodeoxygenation of glycerol to allyl alcohol
M. Schlaf
25
competitive sources of erythritol became available, its hydrodeoxygenation would offer a direct route to the important C4 building block 1,4-butanediol, a key component of the elastic polyester–polyurethane copolymers LYCRA TM and SPANDEX TM and of course THF solvent (Fig. 2.10 ).
Again most efforts toward the hydrodeoxygenation reactions shown in Fig. 2.10 have been focused on the use of heterogeneous systems, e.g., the production of THF in ~ 50 % yield over a Re/Pd/SiO 2 /Nafi on catalyst patented by Manzer [ 54 ] or the more recent development of the already mentioned Ir–ReOx/SiO 2 systems by Tomishige [ 43 ]. The only homogeneous hydrodeoxygenation of this substrate to the potentially very valuable 2,5-dihydrofuran (Fig. 2.11 ) was achieved by formic acid via 1,4-anhydroerytrhitol in analogy to the already discussed glycerol transformation [ 53 ].
2.3.1.4 Xylose, Furfural, and Xylitol
The high abundance of hemicellulose, its comparatively facile hydrolysis to xylose followed by hydrogenation to xylitol, and its well-established direct conversion to furfural make these three C5 units attractive hydrodeoxygenation targets [ 2 , 55 ]. However, starting with a carbon-chain length of fi ve (or longer – see below), an overall very complex hydrodeoxygenation reaction can cascade result, in which the actual pathways and observed intermediates followed and observed depend on (a) whether the reaction cascade begins with hydrogenation of xylose to xylitol or dehydration and rearrangement to furfural and (b) the relative concentrations of substrate, water, acid, and catalysts employed. Excluding any substrate or product dimerization/oligomerization/polymerization pathways, e.g., those leading to humin formation [ 56 ], Fig. 2.12 shows the overall hydrodeoxygenation reaction cascades with the conceivable intermediates and pathways ending in 2-methy- tetrahydrofuran, tetrahydropyran, and the total hydrodeoxygenation product pen- tane as stable terminal products under reducing conditions [ 57 ].
Neither hetero- nor homogeneous catalysts have to date realized this entire C5 reaction cascade, value chain, and pathways. Intermediates and products observed with homogeneous systems are limited to furfural → furfuryl alcohol,
Fig. 2.10 Hydrodeoxyge- nation of erythritol to 1,4-butanediol and THF
Fig. 2.11 Hydrodeoxyge- nation of erythritol 2,5-dihydrofuran by formic acid
26
furfuryl alcohol → 2-methylfuran and furfural → tetrahydrofurfuryl alcohol, while the potentially very valuable 1,5-pentanediol has to date only been generated using heterogeneous catalysts [ 58 – 60 ]. In addition to the pathways shown in Fig. 2.12 , heterogeneous catalysts have also been used to generate 1,2-pentanediol and cyclo- pentanone directly from furfural [ 61 – 65 ].
Employing the complex cis -[(6,6′-Cl 2 -2,2′-bipyridine) 2 Ru(H 2 O) 2 ][BArF] as a homogeneous catalyst, Lapido et al. have demonstrated part of the reaction cascade as shown in Fig. 2.13 leading from furfural to furfuryl alcohol, 2-methylfuran, and tetrahydrofurfuryl alcohol [ 66 ]. This catalyst already mentioned above was originally developed as an alkene hydrogenation catalyst operating in aqueous medium by Lau [ 41 , 42 ] and has also (as the trifl ate salt) been successfully employed for the partial or complete hydrodeoxygenation of terminal diols to primary alcohols or alkanes in water [ 40 ]. In ethanol solvent and with perfl uoro tetra-aryl borate as the counterion, it achieves very high conversions of furfural to furfuryl alcohol (93 %) and essentially quantitative conversion to tetrahydrofurfuryl alcohol (99 %) under 5.16 MPa hydrogen at 85 and 130 °C, respectively. Under similar conditions, the trifl ate salt catalyst generates 20–25 % of 2-methylfuran, effectively a benzylic hydrogenolysis, underlining the potentially large infl uence of solvent and counterion effects on catalyst reactivity and selectivity in homogeneous systems assuming that the catalyst is in fact still homogeneous in this case [ 66 ].
Fig. 2.12 Hydrodeoxygenation reaction cascade for hemicellulose/xylose showing all conceivable intermediates and pathways
M. Schlaf
27
Sen has applied the RhCl 3 /HI (1/1.5) system to the conversion of xylose in a biphasic water/chlorobenzene mixture at 140 °C and 2 MPa hydrogen realizing 80 % yield of 2-methyl-tetrahydrofuran at 95 % substrate conversion. This system is further discussed in more detail in the context of the C6 hydrodeoxygenation targets in Sect. 2.3.1.6 .
2.3.1.5 Levulinic Acid
A second C5 value chain starting from levulinic acid – obtained by rehydration and deformylation of 5-hydroxymethylfurfural (HMF) – can lead to the same terminal stable products as when starting from hemicellulose. Figure 2.14 shows all conceivable intermediates and pathways for this substrate, again all of which have been realized with heterogeneous catalyst systems [ 4 , 67 – 69 ]. For homogeneous catalysts, the cascade has to date focused on the comparatively facile conversion to γ-valerolactone, which has been considered as a key platform molecule that can serve as solvent, fuel, or precursor to other value-added chemicals as laid out in Fig. 2.14 and has been generated in high yield by ruthenium- and iridium-based hydrogenation catalysts supported by phosphine and pincer ligands [ 70 – 73 ]. The complexes can be prepared in situ from “standard” precursors such as [Ru(acac) 3 ] or [Ir(COE) 2 Cl 2 ], which for the latter with the PNP pincer ligand 2,6-di(bis-tbutyl- phosphine)methyl-pyridine can reach TON of 70,000 [ 74 , 75 ]. Using formic acid as the reductant, which eliminates the need for hydrogen pressure, Horvath has successfully applied the metal–ligand bifunctional Shvo catalyst to this transformation achieving TON of ~ 1000 with repeated recycling of the catalyst without loss of activity [ 76 – 78 ].
A fi rst example of the use of a homogeneous catalyst system reaching beyond γ-valerolactone is part of the conceptual value chain from sucrose to alkanes developed by Horvath and given in Fig. 2.15 [ 79 ], which integrates the acid-catalyzed hydrolysis, dehydration, and deformylation of sucrose or HMF to levulinic acid with
Fig. 2.13 Actually realized (homo- and heterogeneously catalyzed) hydrodeoxygenation reaction cascades for furfural
28
Fig. 2.14 Hydrodeoxygenation reaction cascade for levulinic acid showing all conceivable intermediates and pathways
Fig. 2.15 Integrated levulinic acid centered value chain leading from sucrose and HMF to alkanes by combination of homo- and heterogeneously catalyzed reactions
M. Schlaf
29
three alternative homogeneously catalyzed pathways for the hydrodeoxygenation of levulinic acid and as a fi nal step the heterogeneously catalyzed conversion of 2-methyl-THF to alkanes using Pt(acac) 2 as the pro-catalyst that in the absence of supporting ligands likely forms Pt 0 particles as the active system. Starting from levulinic acid, the right branch of the reaction cascade in Fig. 2.15 describes the use of a water-soluble catalyst formed in situ from sulfonated triphenylphosphines P(m- C 6 H 4 SO 3 Na ) 3 and Ru(acac) 3 for the initial conversion of levulinic acid to γ-valerolactone followed – after product isolation – by a second hydrogenation to 2-methyl-THF under slightly more forcing and acidic (NH 4 PF 6 additive) solvent- free conditions using PBu 3 as the supporting ligand. At higher temperatures and starting directly from levulinic acid as the reactant and solvent (left branch of the reaction cascade in Fig. 2.15 ), the same catalyst realizes hydrogenolysis of the initial γ-valerolactone product to the potentially valuable 1,4-pentanediol in 63 % yield. Alternatively, and avoiding the requirement for H 2 pressure, [(η 6 -C 6 H 6 ) Ru(1,10-phenanthroline)]SO 4 can act as a transfer hydrogenation catalyst using sodium formate as the reductant and HNO 3 as the acid cocatalyst [ 80 ].
Another unique homogeneous system capable of deoxygenating levulinic acid beyond the initial product γ-valerolactone was developed by Leitner, Klankenmayer, and coworkers [ 81 , 82 ]. The combination of the complex [Ru(triphos)(CO)H 2 ], generated in situ under reducing conditions from [Ru(acac) 3 ] and the triphos ligand (1,1,1-tris(diphenylphosphinomethyl)ethane), and an acidic imidazolium-based ionic liquid and/or NH 4 PF 6 as acid cocatalysts enables the generation of 2-methyl- THF in up to 92 % yield. Figure 2.16 summarizes the transformation, catalyst structures, and reaction conditions.
The authors also performed a preliminary engineering and techno-economical analysis of a continuous process that indicated that with levulinic acid as the substrate and solvent, i.e., under effectively solvent-free conditions, effi cient catalyst recycling should be possible and that the overall costs of the process would be dominated by those of the catalyst and raw material making geographically distributed small-scale operations – a key point for any biorefi nery – potentially economically attractive.
Fig. 2.16 Homogeneously catalyzed hydrodeoxygenation of levulinic acid to 2-methyl-THF
30
2.3.1.6 Glucose, Fructose
With glucose units as the repeat unit of cellulose and therefore likely the most abundant biomolecule on the planet, the C6 substrates are arguably the most important hydrodeoxygenation targets. Similar to the C5 substrates, a very complex reaction cascade can result and Fig. 2.17 shows the intermediates and pathways ending in 2,5-dimethyl-THF, hexane and methylcyclopentane, and conceivably dianhydrosor- bitol [ 83 ] as the terminal stable products under reducing conditions with HMF, 2,5-dimethylfuran, 2,5-hexanedione, and iso sorbide as the key reactive intermediates.
As with the C5-based substrates, approaches to the hydrodeoxygenation of glucose and its derivatives have to date been dominated by the use of heterogeneous catalysts with only a handful of examples attempting to use homogeneous systems described in the literature.
A simple example of the application of a homogeneous catalyst to a biomass- derived substrate is the use of the metal–ligand bifunctional Shvo catalyst to the hydrogenation of HMF to 2,5-dihydroxymethyl-furan [ 76 , 77 , 84 ]. The reaction proceeds quantitatively (99 % yield) under comparatively mild conditions (90 °C, 1 MPa H 2 ), and the catalyst can be recycled without loss of activity at least nine
Fig. 2.17 Hydrodeoxygenation reaction cascade for cellulose/glucose showing all conceivable intermediates and pathways
M. Schlaf
31
times by isolating the product through precipitation and fi ltration from the toluene reaction solution.
Sen et al. have reported the use of HI as a reductant capable of effecting both dehydration and simultaneous partial reduction of C6 sugars, most notably fructose to 5-methylfurfural in up to 47 % with the concomitant production of I 2 . The HI reductant can then be regenerated in situ by reaction of I 2 with hydrogen over a supported noble metal (Pd/C, Rh/C, Ru/C) [ 85 ]. More relevant to this chapter is the extension of this concept to the use of a catalyst system consisting of RhCl 3 , HI, and NaI that results in the conversion of fructose, HMF, 5-methyl-furfural, 2,5- hexanedione, and 2,5-hexanediol to 2,5-dimethyl-THF in up to quantitative yield following the same reaction pathways as laid out in Fig. 2.17 [ 86 – 88 ]. While the true identity of the catalyst could not be established, fi ltration of the biphasic toluene/water reaction mixture gave an aqueous phase that maintained its catalytic activity and could be recycled multiple times by adding fresh fructose substrate. When HCl rather than HI was used as the acid cocatalyst, precipitation of Rh(0) with lower and different catalytic activity was observed with no formation of 2,5-dimethyl-THF. This suggests that in the presence of HI, the catalyst formed in situ from RhCl 3 is in fact homogeneous, likely a complex or soluble cluster of unknown structure supported by iodide ligands [ 86 ]. Notably the presence of HI rather than HCl also almost completely suppressed the hydrogenation of the toluene solvent phase to methylcyclohexane that is observed with the heterogeneous Rh(0) catalyst generated in situ from RhCl 3 /HCl.
The author’s group has demonstrated that the complex [(4′-Ph-terpyridine) Ru(H 2 O) 3 ](OTf) 2 , previously applied to terminal diols and glycerol (see Sect. 2.3.1.2 ), is also effective in the hydrogenation of 2,5-hexanedione, either targeted directly or generated in situ by hydrolysis of 2,5-dimethylfuran (Fig. 2.17 ) to 2,5-hexanediol and 2,5-dimethyl-tetrahydrofuran and also evaluated the iso - electronic iridium system [(4′-Ph-terpyridine)Ir(H 2 O) 3 ](OTf) 3 in the same reaction [ 89 ]. Depending on reaction conditions, the ruthenium system realizes yields of 2,5-hexanediol (69 % at 175 °C) or 2,5-dimethyl-tetrahydrofuran (80 % at 200–225 °C) and also generates small amounts of 2-hexanone and hexane, while the iridium system is slightly less active and also has a lower temperature stability. Both catalysts work best in water as the reaction medium and are inhibited by the addition of acid (e.g., HOTf) as the cocatalyst or the use of 1,4-dioxane or sulfolane as solvents. They are deactivated by formation of the substitutionally inert bis-chelate complexes [M(4′-Ph-terpy) 2 ] n+ (M = Ru, Ir; n = 2, 3) and a deposition of a metal coating in the reactor body, which for iridium is acting as a heterogeneous catalyst emphasizing the need for carefully executed control reactions in order to establish true homogeneous catalyst activity under the fairly harsh reaction conditions (T > > 150 °C, acidic medium) typically required for hydrodeoxygenation reactions.
An effective rapid total deoxygenation of various C6 sugars to hexane and its isomers and in some cases hexene(s) was achieved by using either the iridium phosphine pincer complex [(POCOP)Ir(H)(acetone)][B(C 6 F 5 ) 4 ]; POCOP = 2,6-[OP( t Bu) 2 ] 2 C 6 H 3 or – in a completely metal-free system – B(C 6 F 5 ) 3 as the catalyst and excess diethylsilane (Et 2 SiH 2 ) as the reductant [ 90 – 92 ]. This in
32
principle promising result is however completely negated by the high cost and extremely low atom effi ciency of both the use and synthesis of the reducing agent, which must be produced via the corresponding silyl chloride obtained through the Müller–Rochow process followed by metathesis with metal hydride or Grignard reagents [ 93 , 94 ]. While the authors also suggest that polymethylhydrosiloxane, a by-product of silicone manufacture, is also effective as a reductant, the overall production volumes of silicones and hence its by-products are miniscule compared to the actual needs for alkanes either as solvents, let alone fuels. An actual large-scale use of the process would therefore be prohibitive from both an economic and ecologic perspective. The proposed reaction sequence does however have enormous potential and merit for the synthesis of partially deoxygenated silylated polyalcohols, e.g., 1,2-deoxyglucitol obtainable when stoichiometric amounts of sterically more demanding silanes (e.g., Me 2 EtSiH) are used. Such deoxygenated sugars are otherwise very diffi cult to access and very valuable synthons for enantioselective synthesis with a well-defi ned stereochemistry of the starting material. By the use of different, naturally abundant hexoses, i.e., mannose, galactose, etc., different precursors with four stereocenters should be accessible by this method.
2.3.2 Homogeneous Catalysts for Hydrodeoxygenation Upgrading of Pyrolysis Bio-Oil
Pyrolysis bio-oil, obtained from the fast pyrolysis of lignocellulosic biomass with exclusion of oxygen, is characterized by the presence of a multitude of reactive carbonyl compounds, i.e., aldehydes, ketones, and hydroxyl aldehydes and hydroxy ketones, anhydrosugars, and phenolics (originating from the lignin content of the biomass) up to 10 % w/w of formic and acetic acid as well as a high and variable (15–30 % w/w) water content. The oil is therefore self-reactive, i.e., unstable against condensation and polymerization reactions. These properties make a typical bio-oil unusable as a fuel “as is,” requiring a reductive upgrading process that lowers the overall content of reactive oxygen functionalities and increases the energy density of the oil. The identifi cation of any catalyst capable of effi ciently hydrodeoxygenat- ing bio-oil is challenging and the subject of current intense research mainly employing heterogeneous systems [ 95 – 97 ]. The use of homogeneous systems for this purpose may arguably be even more diffi cult, in particular with a view to catalyst inhibition by the (acidic) substrates and catalyst recycling from an anticipated very complex reaction and product mixture. To date only two studies that attempt this approach have appeared in the literature. Heeres et al. reported the use of the well- established hydrogenation catalyst RuCl 2 (PPh 3 ) 2 [ 98 , 99 ] for the hydrogenation of the ketone and aldehyde model compounds acetol and hydroxyl-acetaldehyde (both typically major components of bio-oil) and the aqueous extract of a bio-oil in a biphasic mixture with toluene as the organic solvent. The catalyst achieved a selective conversion to the corresponding diols in ~ 60 % yield after < 30 min. at 60–90 °C
M. Schlaf
33
under 4 MPa H 2 and tolerated the presence of acetic acid with only a small decrease in activity [ 100 ]. The catalyst hydrogenates the aldehyde substrate faster that the ketone and the authors suggested that catalyst recycling from the organic phase of the reaction mixture may be possible with further process optimization.
The already mentioned Shvo catalyst (Sects. 2.3.1.5 and 2.3.1.6 ) was also applied to a set of model compounds consisting of lignin models such as vanillin, cinnam- aldehyde, and methylacetophenone as well as the acetol and hydroxyl-acetaldehyde also targeted by Heeres [ 101 ]. In toluene as the reaction medium at 90–145 °C under 1 MPa H 2 , the catalyst achieved high conversions of these substrates to corresponding alcohols and tolerated the presence of acetic acid as was also observed with RuCl 2 (PPh 3 ) 2 . When applied to an actual bio-oil (derived from white poplar), the catalyst was completely soluble in the oil and effected an almost complete conversion of the aldehyde components (by NMR analysis) indicating a substantial stabilization of the oil. However, catalyst recovery or reuse was not possible, which as previously stated is one of the major challenges of applying homogeneous systems to bio-oils and may in fact prevent their larger-scale use for this purpose.
2.4 Conclusion and Outlook
From the literature survey presented here, it is apparent that compared to the multitude of studies on heterogeneous catalyst systems, the use of homogeneous systems for biomass hydrodeoxygenation remains relatively unexplored, which may in part be rooted in the fact that the research fi eld is dominated by chemical engineers that often do not venture into the synthetic procedures typically required to make homogeneous catalysts available for study. However, from the author’s viewpoint, this represents an opportunity rather than a problem, as collaborations between synthetic (organometallic) chemists well-versed in ligand and catalyst design and synthesis and chemical engineers as experts in process design and optimization may indeed turn out to be very fruitful to the fi eld. In particular, the perennial problem of recycling homogeneous catalysts should benefi t from such collaborations, while the chemist’s role will lie in synthesizing catalysts that maintain their activity under the high-temperature aqueous acidic conditions necessarily required for the reaction cascades laid out in the Figs. 2.5 , 2.10 , 2.11 , 2.12 , 2.14 , and 2.17 given above. Since the overall goal of the hydrodeoxygenation reactions is the defunctionalization of the substrates to products that are thermodynamically stable and unreactive under the reducing conditions (hydrogen atmosphere), the main role of the ligand frame- work around metal center will in these reaction not be the induction of chemo- or stereoselectivity typically sought with homogeneous catalyst systems, but the maxi- mization of deoxygenation activity and the stabilization of the metal center against reduction to bulk metal leading to the in situ generation of heterogeneous catalysts defeating the purpose and potential advantages of maintaining a homogeneous system. A logical approach to achieve this goal is the design and use of highly chelating ligands that will result in very high complex formation constants due to the
34
macrocyclic effect while maintaining at least one free or labile coordination site for hydrogen and/or substrate activation [ 17 ]. The successful use of homogeneous catalysts for the hydrodeoxygenation of biomass-derived substrates to value-added chemicals and possibly alkane fuels will be contingent on meeting these criteria.
References
1. Schlaf M (2006) Selective deoxygenation of sugar polyols to alpha,omega-diols and other oxygen content reduced materials – a new challenge to homogeneous ionic hydrogenation and hydrogenolysis catalysis. J Chem Soc Dalton Trans 39:4645–4653
2. Schiweck H et al (2000), Sugar alcohols. In: Ullmann’s encyclopedia of industrial chemistry. Wiley-VCH Verlag GmbH & Co. KGaA
3. Sabatier P, Mailhe A (1909) New applications of the general method of hydrogenation on divided metals. Annales De Chimie Et De Phys 16:70–107
4. Schuette HA, Thomas RW (1930) Normal valerolactone. III. Its preparation by the catalytic reduction of levulinic acid with hydrogen in the presence of platinum oxide. J Am Chem Soc 52:3010–3012
5. Schiavo V, Descotes G, Mentech J (1991) Catalytic-hydrogenation of 5- hydroxymethylfurfural in aqueous-medium. Bull Soc Chim Fr 128(5):704–711
6. Huber GW, Iborra S, C

reaction pathways and mechanisms in thermocatalytic biomass conversion ii: homogeneously catalyzed...

Documents