CHEMSUSCHEM CONCEPTS DOI: 10.1002/cssc.201402379
High-Performance Polymers from Nature: Catalytic Routes and Processes for Industry Guido Walther*[a] In memoriam of Dr. Michael Kant
It is difficult to imagine life today without polymers. However, most chemicals are almost exclusively synthesized from petroleum. With diminishing oil reserves, establishing an industrial process to transform renewables into high-value chemicals may be more challenging than running a car without gasoline. This is due to the difficulty in setting up processes that are novel, profitable, and environmentally benign at the same time. Additionally, the quest for sustainability of renewable resources should be based on incorporating ethical considerations in the development of plans that utilize feedstocks intended for human nutrition and health. Thus, it is important to use bio-energy containing renewable resources in the most efficient way. This Concept goes beyond the synthesis of monomers and provides insights for establishing an industrial pro-
cess that transforms renewable resources into high-value chemicals, and it describes careful investigations that are of paramount importance, including evaluations from an economical and an ecological perspective. The synthesis of monomers suitable for polymer production from renewable resources would ideally be accompanied by a reduction in CO2 emission and waste, through the complete molecular utilization of the feedstock. This Concept advocates the drop-in strategy, and is guided by the example of catalytically synthesized dimethyl 1,19-nonadecanedioate and its a,w-functionalized derivatives. With respect to the Twelve Principles of Green Chemistry, this Concept describes a technological leap forward for a sustainable green chemical industry.
1. Introduction Ever since our distant ancestors managed to light fire, humankind has been developing a continuously increasing quest for energy. With a world population greater than 7 billion, growing rapidly, and fossil resources that are diminishing, the energy demand needs to be covered by a smart energy mix. This might be a solvable challenge in due course, since many renewable energy resources are already in use (e.g., hydropower, geothermal energy, wind and solar energy, bioenergy).[1, 2] However, renewable resources, such as bio-oil and biomass, are scarce and limited because of their availability varies per season. These renewables are indeed the only resources providing both energy that is stored in chemical bonds and a useful molecular structure of carbon atoms. The quest for sustainability of these renewable resources should therefore be based on incorporating ethical considerations in the development of plans that utilize feedstocks intended for human nutrition and health. Thus, it is important to use bioenergy containing renewable resources in the most efficient way. With a view to petroleum-based polymers, there is a quest for finding ways to synthesize polymers that possess properties similar to, for example, polyethylene.[3, 4] Many polymers are sophisticated, product-oriented and high-performance materials, and these all have in common outstanding thermal,
[a] Dr. G. Walther Leibniz Institute for Catalysis at the University of Rostock Albert-Einstein-Str. 29 A, 18059 Rostock (Germany) E-mail: [email protected]
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
mechanical, electrical, optical, surface, and rheological properties. Petroleum-based examples of such specialty polymers, produced on a scale of several 100 Mton annually, are polyether ether ketone (PEEK), polytetrafluoroethylene (PTFE), and polyamide 12 (PA12). The current size of the market for biobased high-performance polymers [e.g., polyphthalamide (PPA), PA10.10 or PA10.12][5] covers about 0.7 Mton annually, and this figure is rapidly increasing. These materials, mainly polyamides, have found application in areas such as the aerospace industry, extrusion processes, and microelectronic compounds.[6] Indeed, the molecular structure of carbon atoms in plant oils is particularly well-suited to the synthesis of long-chain polymerizable segments. Even though the hydroformylation of alkenes is a well-established synthesis route, it cannot achieve a yield of linear products greater than 36 %.[7] This is due to a thermodynamically unfavorable intermediate reaction step, in which a terminal C=C double bond is temporarily formed. Linear monomers used for polymerization require a very high grade of purity, which is why so far long-chain hydrocarbons (longer than C12) have been used only rarely. A pioneering discovery in this field was made Cole-Hamilton et al. in 2005, using the catalytically active and highly selective palladium complex Pd/o-C6H4(CH2PtBu2)2 in a methoxycarbonylation route towards linear reaction products.[8] In the presence of methane sulfonic acid (MeHSO3) this catalyst was found to convert highly reactive ethylene into methyl propanoate (cf. Scheme 1 a), and was hence commercialized for the petroleChemSusChem 2014, 7, 2081 – 2088
2081
CHEMSUSCHEM CONCEPTS
www.chemsuschem.org
Besides the promising developments in methoxycarbonylation reactions, there are various alternative routes to obtain long-chain hydrocarbons that can be functionalized at their terminal position. A recent critical Review presented four reactions, taking the role of key reactions for processing renewable resources: Metathesis, which was reported to shorten the carbon chain length while fuctionalizing them at the same time; hydroformylation, which provides access to aldehydes, intermediates from which other a,wfunctionalized derivatives can be obtained using well-known technologies; degradation, which defunctionalizes renewables to access new active reaction Scheme 1. Inspired by the Lucite process (a), we developed a combined transesterificasites; and also isomerization, in combination with tion, isomerization, and methoxycarbonylation reaction of plant oil toward 1,19-diester (b), using Pd/o-C6H4(CH2PtBu2)2. R, R’ and R’’ consist of linear mono- or polyunsaturated other techniques such as methoxycarbonylation.[21] hydrocarbon chains of 17 carbon atoms. FAME denotes fatty acid methyl esters of polyWith a view to making long-chain crystallizable unsaturated hydrocarbon chains. a,w-functionalized segments from sunflower oil, olefin metathesis would provide a complex mixture of a,w-functionalized monomers. More promising results were described for hydroformylation reactions.[21] Howevum-based Lucite process.[8–10] Inspired by the Lucite process, an isomerizing methoxycarbonylation route was developed to er, the selectivity towards a,w-dialdehydes was reported as still transform methyl oleate (derived from sunflower or rapeseed being very challenging. In addition, all reactions required transoil) into dimethyl 1,19-nonadecanedioate (1,19-diester).[3, 4] A ersterification of the sunflower oil. Thus, all reactions are key finding we made is that this catalyst also promotes the tandem reactions in different vessels.[21] The combined reaction direct conversion of sunflower oil into 1,19-diester, (cf. of transersterification, isomerization, and methoxycarbonylaScheme 1 b),[11–13] which may open the door to new catalytic tion may provide both high yield and high selectivity of the routes and processes for industry. desired product, and proceeds as a one-pot reaction. This would make the formation and distillation of intermediate methyl esters redundant.[11–13]
2. Results 2.1 Plant oil utilization
2.2 a,w-Diester—A novel platform chemical
At present, mainly fruits that consist of oil and sugar are used for energy conversion. The rest of the plant matter, even though it contains energy stored in chemical bonds, is only used scarcely, because the technology is still in its infancy.[14] The distribution of vegetable oil consumption between food, feed, and industrial use was 74:6:20 in 2008.[15] This ratio is continuously shifting toward higher industrial use, of which approximately 30 % accounts for biobased transportation fuel— this equals 6 % in total.[15] Within the last decade, the methoxycarbonylation of lower olefins as well as unsaturated carboxylic acids and esters has received much attention; besides experimental investigations,[4, 8–13, 16–19] tremendous efforts have been undertaken to elucidate a potential mechanism.[9, 10, 17, 20] However, prior to processing olefins according to the Lucite process (cf. Scheme 1 a), the plant oil needs to be converted into its corresponding fatty acid methyl esters and glycerol by transesterification. Even though the transesterification of plant oil, which is the first step for utilizing this resource as a fuel (e.g. biodiesel) for the transportation sector, is available commercially on a large scale, it would be desirable to obtain a long nonbranched (CH2)n a,w-functionalized crystallizable segment directly from the raw material. Such a segment may consist of n = 19 or 23 CH2 units, which may be formed from oleic or erucic acid moieties, respectively.
Recently, a possibility has been developed to directly synthesize 1,19-diester from high-oleic sunflower oil via a highly selective combined reaction involving transesterification, isomerization, and methoxycarbonylation.[11–13] This reaction proceeds already at 80 8C in an acidic environment (9.6 mol % CH3SO3H) and pCO = 30 bar, using 2.4 mol % Pd/o-(C6H4)(CH2PtBu2)2 as catalyst. The molecular ratio of palladium to ligand was always kept at 1:5. When the volumetric ratio of methanol to plant oil equaled 3, we obtained a yield (Y) of 91 % at a selectivity (S) of 94 % of the 1,19-diester. The best result (Y = 97 %, S = 97 %) was achieved when increasing the volumetric ratio of methanol to plant oil to 4 and the catalyst quantity to 4.8 mol %.[13] However, the latter set of parameters does not meet the best result when considering its “greenness”. This is associated with the high quantities of catalyst and solvent used, and will be revisited in this Concept when discussing environmental aspects. This direct synthesis process was found to be suitable for scale-up by several orders of magnitude, without any significant decrease of yield and selectivity.[11–13] Traces of water do not influence the reaction, which allows the use of technicalgrade methanol. However, greater quantities of water do hinder the reaction’s progress. With a view to the acidic reaction environment required, H2SO4 was found to be suitable in slightly greater quantities to obtain comparable yields of > 80 %.[18] Another important finding is that polyunsaturated
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2081 – 2088
2082
CHEMSUSCHEM CONCEPTS and branched hydrocarbon moieties are not converted into the desired a,w-diesters,[18] instead, a,x-diesters, with x < w, are obtained.[19] An comparison, albeit a simplistic one, of a two-stage process from sunflower oil via methyl oleate toward a,w-diesters and the direct synthesis route shows the potential advantage of the latter process, as it may be lead to a reduction of CO2 emissions and of waste. Permitting that any erroneous assumptions are consistently made for both processes, this is particularly the case for the power consumption when compressing CO and heating/cooling the reaction inventory based either on renewable or on fossil energy. Moreover, the direct synthesis route does not require any additional chemicals, such as NaOH and KOH, which are needed for the two-stage process. Additionally, it should be noted that the glycerol obtained from the transesterification of the plant oil into the methyl ester is salt-contaminated and, in the least favorable case, needs to be considered as a waste. The direct synthesis process may provide salt-free glycerol as a by-product. Both glycerol[22] and the desired a,w-functionalized long-chain hydrocarbon diester[13] have indeed been identified as being platform chemicals, as illustrated in Scheme 2. Knowledge of the reaction mechanism is crucial for understanding and improving the properties of catalysts. To overcome limitations such as reactivity and stability, we examined this reaction experimentally and developed a microkinetic
www.chemsuschem.org model from extensive density functional theory (DFT) calculations reflecting the current state of science.[20] An important fact to consider is that the active palladium species is present as a hydride (cf. Scheme 3), H-[Pd] (0), where [Pd] denotes the organometallic Pd/o-(C6H4)(CH2PtBu2)2 catalyst. This was confirmed by electron paramagnetic resonance (EPR) spectroscopy. The thermal stability of the catalyst was found to be of minor importance, since the decomposition of the ligand is provoked more by the methanolysis toward the desired a,w-diester. This elementary step was identified as being the rate-determining step and having a Gibbs free energy of 137.5 kJ mol1. A high-pressure NMR cell[23] equipped with a gas flow and gas saturation system and gas recycling unit[24] was used to corroborate the findings on the making and breaking of bonds, especially with regard to the PdP bond of the organometallic complex. In situ UV-vis spectroscopy provided insight into the reversibility of a change in the ligand sphere of the [Pd] catalyst while the combined isomerization–methoxycarbonylation reaction of methyl oleate toward a,w-diester proceeded. This result agrees with the findings from activity measurements.[20] Hence, a potential mechanism may be suggested for the combined transesterification, isomerization, and methoxycarbonylation of plant oils, cf. Scheme 3. A thorough thermodynamic analysis provided detailed insight into the elementary steps. Some intermediate steps were
Scheme 2. Principle routes of the value-added chain from renewable resources (plant oils) that are directly converted into platform chemicals (a,w-diester and glycerol), via oleochemical building blocks such as the corresponding a,w-diacid, the a,w-diol, and the a,w-diamine, toward the desired polymers using existing technologies for polymerization.
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2081 – 2088
2083
CHEMSUSCHEM CONCEPTS
www.chemsuschem.org methoxycarbonylation reaction, such as o-(C6H4)(CH2PPh2)2, which was inactive, o-(C6H4)(CH2PtBu2) (CH2PPh2).[4, 10] Surprisingly, the latter (nonsymmetric) ligand was found to be as active as o-(C6H4)(CH2PtBu2)2 but much more robust against decomposition. From the DFT study, it may be assumed that the tBu moieties bound to the phosphorus atom of the ligand provide charge transfer, necessary for the catalytic reaction. On the other hand, tBu moieties are less robust than Ph moieties. However, Ph groups do not provide enough steric hindrance to fully isomerize the C=C double bond toward the terminal position. With respect to the costs of the catalyst, the question arises whether a ligand that is symmetric but contains a mix of tBu and Ph moieties, for example, o-(C6H4)(CH2PtBuR)2, where R denotes Ph or Ad, meets the properties of o-(C6H4)(CH2PtBu2)(CH2PPh2).
Scheme 3. Catalytic cycle of the synthesis of a,w-diesters from methyl oleate (n + m = L = 14) or methyl erucate (n + m = L = 18) present in triglycerides of the plant oil. Elementary steps are described in the text.
examined in greater accuracy; for example, the isomerization (1) of the C=C double bond proceeds in both directions with low barriers (2); however, a solvent effect was found to favor the formation of an n-alkyl (3). According to the consecutive addition of CO and methanol, as interpreted from the experimental results, CO insertion was not found to be an elementary step that is reversible; it rather undergoes a quasi-equilibrated elementary step (4). Nevertheless, the rate-determining step is the methanolysis, which proceeds favorably via intramolecular attack of methanol (5). This elementary step might be attended by PdP bond cleavage, which may lead to precipitation of Pd0. A microkinetic model was developed to calculate the apparent reaction barrier for the formation of 1,19diester from sunflower oil. This result was found to agree well with the experimental findings, and may thus be taken as indirect corroboration of the proposed reaction mechanism.[20] Additionally, the same data predicted this reaction to also occur when using other feedstocks and acids. To verify this hypothesis, different plant oils (e.g. sunflower oil, rapeseed oil, soybean oil, castor oil, olive oil, peanut oil) and acids were studied. Experimental results showed that Pd/o-(C6H4)(CH2PtBu2)2 is a very selective catalyst for the synthesis of a,wdiesters. However, the a,w-diesters are always obtained from oleic or erucic acid moieties present in the triglyceride of the plant oil. Other fatty acid moieties, independent of whether they are saturated or polyunsaturated, or even OH-branched monounsaturated ricinoleic acid moieties do not influence the reaction.[18] Surprisingly, these simple facts were not perceived before the reaction mechanism was elucidated; although recently a,w-diesters were already found to be obtained from other plant oils as well.[16] Finally, with the use of (technical) H2SO4 as co-catalyst, the catalytic system was found to open new routes for industrial applications.[18] Besides palladium-based organometallic catalysts containing o-(C6H4)(CH2PtBu2)2 as ligand, other ligands were tested for the 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2.3 Other a,w-functionalized monomers—Oleochemical building blocks Other a,w-functionalized derivatives of the a,w-diesters that are presumably of interest for polymerization can not be obtained directly from plant oils. Such derivatives include highly purified a,w-diacids, a,w-diols, and a,w-diamines. However, we synthesized them starting at the a,w-diesters. For example, highly purified a,w-diacids were obtained from a,w-diesters by alkaline hydrolysis and one subsequent re-crystallization (purity > 99 %).[13] The catalytic hydrogenation of nonactivated esters is challenging. Alternatively, esters have been reduced using stoichiometric quantities of metal hydrides, such as LiAlH4, but this has been found to be ineffective for the synthesis of long-chain hydrocarbon a,w-diols.[4] However, a few systems that catalyze the hydrogenation of non-activated esters have been reported.[25–27] Surprisingly, the ruthenium hydride complex [(PNN)RuH(CO)], based on the pincer ligand PNN (2-(di-tert-butylphosphinomethyl)-6-(diethylaminomethyl)-pyridine),[27] was found to catalyze the reduction of a,w-diester towards its corresponding a,w-diol.[13] In actual fact, this catalyst operates already under mild conditions, such as # = 115 8C and pH2 10 bar, and we obtained a yield of the a,w-diol of 98 % using 0.5 mol % catalyst. More sophisticated is the synthesis of linear but also primary a,w-diamines. Inspired by the selective synthesis of primary amines directly from alcohols and ammonia,[28] we found that this reaction also proceeds from a,w-diol toward a,w-diamine. This reaction was catalyzed at rather mild conditions using the dearomatized acidine-based pincer complex [RuHCl(A-iPr-PNP) (CO)], where PNP denotes 4,6-(di-iso-propylphosphinomethyl)acridine, developed by Milstein et al.[28] At already 140 8C, we obtained a yield of primary a,w-diamines of 68 %, using only 0.25 mol % catalyst, an excess of ammonia, and 2-methyl-2-butanol as solvent.[13, 29] Within this context, the a,w-diester occupies a pivotal role as a platform chemical, with its derivative monomers, such as linear aliphalic a,w-diacids, a,w-diols, and a,w-diamines, acting as oleochemical building blocks for the polymer industry, cf. Scheme 2.[13] ChemSusChem 2014, 7, 2081 – 2088
2084
CHEMSUSCHEM CONCEPTS 2.4 Polymer synthesis The presented oleochemical building blocks as well as the identified platform chemical a,w-diester were found to have similar properties as other polymerizable monomers of shorter segments.[4, 30] Thus, for polymerizing these novel monomers existing technologies may be used. Quinzler et al. synthesized polyesters made up of alternating a,w-diesters and a,w-diols that contain 19 or even 23 (CH2) segments. The polycondesation of stoichiometric quantities of a,w-diester and a,w-diol was reported being catalyzed using Ti(OBu)4 at a temperature range of 110 8C–150 8C and at a pressure of 0.01 mbar. After a reaction time of 17 h and cooling, a quantitative yield was obtained.[4] Polyamides, such as PAx.19, where x E {6, 9, 10, 12, 19},[5] were synthesized using melt polycondensation on a pilot plant scale, where equimolar amounts of a,w-diamines and a,w-diacids were employed. Typical reaction conditions were # = 240– 260 8C, p = 18–26 bar, and in some cases an admixture of 50 % distilled water and were kept invariant for about 2 h.[30] At first glance, we found that the melting temperature decreases with increasing length of the monomers. Additionally, thermal and mechanical properties are very close to those of the established PA12. A unique selling point of these novel high-performance polymers—especially PA6.19 and PA10.19—is their very high translucence compared to that of PA12, which is beneficial for many applications, but also improved thermostability and high impact strength. This trend could be confirmed for PAx.23, where x E {6, 9, 10}, as well.[30] Polyurethanes are formed by reacting a polyol with an isocyanate. The length of the linear hydrocarbon chain of the polyol is expected to mainly contribute to the polyurethanes’ properties, such as elasticity and good hydrolysis resistance. The reaction of long-chain a,w-diacids [e.g., (CH2)n,with n = 19 or 23] and short-chain a,w-diols (e.g. diethyleneglycole) results in polyesterpolyols which, when further treated with diisocyanites using polycondensation, yield polyurethanes. However, since highly purified a,w-diols of a chain length of either 19 or 23 (CH2) units are available, these monomers may directly react with isocyanates to form polyurethanes. Since the 1,19-diol has a high melting point, and the synthesis of polyurethanes re-
www.chemsuschem.org quires the reactants being in liquid form, the polyurethane formation is accelerated so that the processing time for forming a certain compound is limited. However, the reaction behavior was controlled best at # = 100–110 8C when using hexamethylene diisocyanate and an organo-metal catalyst.[30] It was found that the hydrophobic domain of the polyurethanes based on C19 monomers is more beneficial for hydrolysis resistance (even at 70 8C) than the one of C12 and shorter monomers.
3. Discussion 3.1 Drop-in strategy—Synthesis route of high-performance polymers for industry Establishing an industrial process that transforms renewable resources into high-value chemicals may primarily focus on the potential value chains in the utilization of the feedstocks. In a more general sense, producing high-performance polymers from nature follows the drop-in strategy. The drop-in strategy is characterized by an existing value chain, utilizing existing infrastructure and a mature market.[17] Hence, existing technologies for the final polymerization process do not need to be changed, when platform chemicals (here a,w-diester and glycerol) and intermediates obtained from them “drop into” an existing process, illustrated in Figure 1 by interlocking the renewable production route (top) with the petroleum-based route (bottom). From the left to the right, the bottom tiles Figure 1 depict crude oil extraction, refinery, platform chemicals, monomers (petroleum-based building blocks), existing industrial polymerization process, and final products. By analogy, renewable resources are turned into higher-value chemicals along the value-adding chain using a catalyst; these chemicals are already basic units (platform chemicals) that after functionalization serve as initial feedstock for polymerization, cf. top tiles. Polymers are then synthesized from such monomers, using existing fossil fuel-based technologies. Additionally, the simple up-scalable direct synthesis process of a,w-diesters from plant oils is essential to make such highperformance polymers competitive for the global market.
Figure 1. Drop-in strategy for platform chemicals (here: a,w-diester and glycerol) into existing processes.
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2081 – 2088
2085
CHEMSUSCHEM CONCEPTS
www.chemsuschem.org
3.2 Environmental aspects: Polymer, plate, or fuel tank? A sensitive issue is the ethical question of the competition of renewable bio-resources between “polymer, plate, and fuel tank”. In order to assess the severity of the consequences, objective criteria can be found, on the one hand, in the quantity of plant oil required to meet the demand of high-performance polymers, and, on the other hand, in the fact that the desired a,w-diesters are exclusively made from nonbranched, monounsaturated alkyl chains present in plant oils. High-oleic sunflower oil contains only 4 % of polyunsaturated moieties, known as w3 and w6 fatty acid derivatives. However, this special breeding of sunflowers does not contain any w3 fatty acid moieties at all. Such plant oils are less suitable for human consumption than those containing greater quantities of polyunsaturated moieties. In the European Union, Article 13 Health Claims says that a nutrition product is a source of w3 fatty acids, when it contains at least 0.3 g alpha-linolenic acid per 100 g. In contrast, rapeseed oil contains more than 35 % polyunsaturated moieties.[18] From this perspective, there might not be the assumed ethical conflicts. This again is affirmed by the costs of such high-performance polymers. Another important issue arises with the question of how much of the cultivated area may be used for sunflower breeding and which location suits best in, for example, Europe to maximize the annual harvest. This is of course dependent on conditions of the soil and the climate. In Germany, the crop yield of rapeseed oil is 1.7–1.8 m3 Ha1, while that of sunflower oil is only 0.7 m3 Ha1. With a view on Europe, the conditions for sunflowers to thrive and prosper are best in the south of France and in Spain. Additionally, edibles are high-quality products from agriculture but quite low in their appraisal, which is provoked politically in Germany. Increasing food prices would raise the price of all products that have an added value. However, politicians are currently pursuing the goal of blending and substituting fossil fuels by bioethanol and biodiesel, respectively. With a view to the role biofuel may play in the future energy supply,[31] the energy density stored in the chemical bonds of the resource should be considered. This is a function of the effective H/C ratio,[1] H=C ratio ¼
nðHÞ 2nðOÞ 3nðNÞ nðCÞ
ð1Þ
For a direct comparison, assuming a copolymer such as PA19.19, made of alternating 1,19-diamine and 1,19-diacid monomers, the H/C ratios for the 1,19-diamine, 1,19-diacid, and PA19.19 are 36:19, 28:19 and 64:38 (= 32:19), respectively; while the H/C ratio of biodiesel is less than 34:19. The reason that biodiesel does not achieve 34:19 is due to polyunsaturated hydrocarbon chains present in the fuel. The quantity of energy stored in the chemical bonds does not differ significantly when comparing biodiesel with the sequence of polymers made of the same raw material. Hence, the question arises: why not utilize renewable resources for the production of higher value-added contaminant-free polymers, before reusing them for zero-emission power generation. To complete this 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 2. Comparison of the value-added chain starting at rapeseed and following the goal currently pursued by politicians, blending and substituting fossil fuels by bioethanol and biodiesel, respectively, with that of the proposed concept for the production of high-performance polymers from nature. The smallest environmental and ethical conflict may be assumed when using sunflowers as feedstock.
view of zero emission, methanol and CO, also required within this process, may be synthesized from syngas and by gasification of the residual biomass after harvesting the plant oil from the seeds, respectively. Figure 2 provides a detailed overview of this concept. Going beyond the CO2 liberated into our atmosphere, one has to distinguish between a small quantity of CO2 that is recycled forming C and O2 by the chlorophyll of the plants, and an excess quantity of CO2. It is widely accepted that the excess CO2 causes the serious environmental problem known as “global warming” or “greenhouse effect”, but there is an even worse environmental problem being considered: The excess CO2 in our atmosphere (that is no longer recycled by plants) is an admission of O2 depletion; the O2 covered in the excess CO2 was originally breathable and thus an increasing extremely serious environmental problem for mankind. On a metric scale, recalling the atomic weight of CO2 (44 u) and O2 (32 u), the daily loss of O2 equals 32/44 73 % of the excess CO2. Even though the global warming is generally considered, it seems that governments and industries do not face the fact that we need oxygen to breathe. Instead, CO2 capturing and storage in depositories underground is believed to solve the environmental problem, which in the author’s opinion only shifts the main challenge of this generation to one of the next generation. It is an important task—maybe a basic need for the survival of mankind—to find solutions to reduce the quantity of excess CO2. On a broader perspective, this is why resources need to be used in their most efficient way. The “greenness of chemical processes and the products”, as promoted by Jimnez-Gonzlez et al.,[32] shall be evaluated for the proposed strategy using different metrics for resources, solvents, renewability, and efficiency. Such metrics are the mass intensity, P
Im ¼
i mðmi Þ mð1,19-diesterÞ
ð2Þ
where m(mi ) is the mass of material mi ; the solvent intensity, P
Is ¼
i mðsi Þ mð1,19-diesterÞ
ð3Þ
ChemSusChem 2014, 7, 2081 – 2088
2086
CHEMSUSCHEM CONCEPTS
www.chemsuschem.org
where m(si ) is the mass of solvent si ; the renewables intensity, P
Ir ¼
i mðri Þ mð1,19-diesterÞ
ð4Þ
where m(ri ) is the mass of renewable ri ; and the atom economy, Mð1,19-diesterÞ P ae ¼ i MðRi Þ
ð5Þ
where M(Ri ) is the molar mass of the reactants Ri, respectively.[32] For a typical batch reaction according to direct synthesis reaction process of 1,19-diester (VMeOH = 180 mL, Vplant oil = 60 mL, ncatal = 2.4 mol %, nCH3SO3H = 9.6 mol %, pCO = 30 bar, # = 80 8C, and t = 32 h), calculations estimate Im 3.76 kg/ kg1,19-diester, Is 2.48 kg/ kg1,19-diester, Ir 3.65 kg/ kg1,19-diester, and ae 62 %, where CO and methanol are assumed to be synthesized from the residual biomass after harvesting the plant oil from the seeds. Another metric is the E-factor, which provides a measure of the quantity of waste produced per quantity of product.[33] Hence, the author recommends considering glycerol as a coproduct that does not need to be treated as waste. The direct synthesis process according to Scheme 1b was found to provide salt-free glycerol,[30] which again confirms the complete utilization of the feedstock. However; a life cycle assessment would provide an in-depth analysis, especially with a view to zero-emission CO and methanol synthesis from biomass and power generation.
Conclusions This Concept provides a strategy synthesizing high-performance polymers from renewable resources using catalytic routes and suggests a process for industry. It seems that currently the wish of serving a continuously increasing economy dominates the choice in which concept being invested. A cheaper producible alternative catalyst may be based on a symmetric ligand, such as Pd/o-(C6H4)(CH2PtBuR)2, where R denoted Ph or Ad. However, there is a lack of the catalytic characterization of such systems. From an environmental point of view, producing high-performance polymers from nature allows the use of safer auxiliaries (e.g. less hazardous acid), catalysts and renewable feedstocks, and provides an energy efficient method for preventing waste material by almost completely utilizing the feedstock. In my opinion, this would contribute to increasing the yield of the desired product and reduce significantly the production costs and the CO2 emission. With respect to the Twelve Principles of Green Chemistry,[34] this concept is a technological leap forward for a sustainable green chemical industry. 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Acknowledgements The author is grateful for an interview with Martin Schulz (President of the European Parliament). Fruitful discussions with Bernhand Conzen (President of the Farming Community Heinsberg, Germany), Angela Kçckritz (LIKAT) and Franz-Erich Baumann (Evonik Ind.) are acknowledged. The TOC graphic was designed by Duncan J. Mowbray. Keywords: catalysis · plant oils · polymers · renewables · sustainable chemistry
[1] P. N. R. Vennestrøm, C. M. Osmundsen, C. H. Christensen, E. Taarning, Angew. Chem. Int. Ed. 2011, 50, 10502 – 10509; Angew. Chem. 2011, 123, 10686 – 10694. [2] J. J. Bozell, Science 2010, 329, 522 – 523. [3] D. J. Cole-Hamilton, Angew. Chem. Int. Ed. 2010, 49, 8564 – 8566; Angew. Chem. 2010, 122, 8744 – 8746. [4] D. Quinzler, S. Mecking, Angew. Chem. Int. Ed. 2010, 49, 4306 – 4308; Angew. Chem. 2010, 122, 4402 – 4404. [5] PAx.19: polyamide made of alternating monomer units, where x represents the chain length of the a,w-diamine and 19 the chain length of a,w-nonadecanediacid used for synthesis; PUR19: polyurethane made of a,w-nonadecanediol. [6] http://plasticeurope.de. [7] A. Behr, D. Obst, A. Westfechtel, Eur. J. Lipid Sci. Technol. 2005, 107, 213 – 219. [8] C. Jimnez-Rodriguez, G. R. Eastham, D. J. Cole-Hamilton, Inorg. Chem. Commun. 2005, 8, 878 – 881. [9] V. de La Fuente, M. Waugh, G. R. Eastham, J. A. Iggo, S. Castillon, C. Claver, Chem. Eur. J. 2010, 16, 6919 – 6932. [10] T. Fanjul, G. Eastham, N. Fey, A. Hamilton, A. G. Orpen, P. G. Pringle, M. Waugh, Organometallics 2010, 29, 2292 – 2305. [11] F.-M. Petrat, F.-E. Baumann, H. Hger, G. Walther, A. Martin, A. Kçckritz, DE 10 2010 002 809A1. [12] F.-M. Petrat, F.-E. Baumann, H. Hger, G. Walther, A. Martin, A. Kçckritz, WO 2011/110249A1. [13] G. Walther, J. Deutsch, A. Martin, F.-E. Baumann, D. Fridag, R. Franke, A. Kçckritz, ChemSusChem 2011, 4, 1052 – 1054. [14] G. Berndes, M. Hoogwijk, R. van den Broek, Biomass Bioenergy 2003, 25, 1 – 28. [15] F. D. Gunstone, Lipid Technol. 2008, 20, 48. [16] M. R. L. Furst, R. Le Goff, D. Quinzler, S. Mecking, C. H. Botting, D. J. Cole-Hamilton, Green Chem. 2012, 14, 472 – 477. [17] P. Roesle, C. J. Durr, H. M. Mçller, L. Cavallo, L. Caporaso, S. Mecking, J. Am. Chem. Soc. 2012, 134, 17696 – 17703. [18] G. Walther, A. Martin, A. Kçckritz, J. Am. Oil Chem. Soc. 2013, 90, 141 – 145. [19] M. R. L. Furst, T. Seidensticker, D. J. Cole-Hamilton, Green Chem. 2013, 15, 1218 – 1225. [20] G. Walther, L. R. Knçpke, J. Rabeah, M. P. Che˛cin´ski, H. Jiao, U. Bentrup, A. Brckner, A. Martin, A. Kçckritz, J. Catal. 2013, 297, 44 – 55. [21] A. Behr, A. J. Vorhold, K. A. Ostrowski, T. Seidensticker, Green Chem. 2014, 16, 982 – 1006. [22] J. A. Kenar, Lipid Technol. 2007, 19, 249 – 253. [23] G. Walther, A. Martin, A. Kçckritz, D. Selent, DE 10 2011 007 527A1. [24] D. Selent, W. Baumann, A. Bçrner, DE 10 3331 143 B4. [25] S. N. Ege, Organic Chemistry, D. C. Heath&Company, Lexington, 1989, p. 596. [26] Y. Pouilloux, F. Autin, J. Barrault, Catal. Today 2000, 63, 87 – 100. [27] J. Zhang, G. Leitus, Y. Ben-David, D. Milstein, Angew. Chem. Int. Ed. 2006, 45, 1113 – 1115; Angew. Chem. 2006, 118, 1131 – 1133. [28] C. Gunanathan, D. Milstein, Angew. Chem. Int. Ed. 2008, 47, 8661 – 8664; Angew. Chem. 2008, 120, 8789 – 8792. [29] F.-E. Baumann, A. Martin, M. Ullrich, M. Roos, P. Hannen, F.-M. Petrat, H. Hger, A. Kçckritz, G. Walther, J. Deutsch, EP 2 484 659A1.
ChemSusChem 2014, 7, 2081 – 2088
2087
CHEMSUSCHEM CONCEPTS [30] F.-E. Baumann, U. Fehrenbacher, T. Hoffmann, D. Fridag, A. Kçckritz, M. Kugler, F.-M. Petrat, B. Tbke, G. Walther, Final report: Neuartige Polymere auf Basis von Triglyceriden; Eigenverlag des Fraunhofer ICT, Germany, 2012. See www.fnr-server.de/ftp/pdf/berichte/22001308.pdf (accessed June 2014). [31] L. Caspeta, N. A. A. Buijs, J. Nielsen, Energy Environ. Sci. 2013, 6, 1077 – 1082. [32] C. Jimnez-Gonzlez, D. J. C. Constable, C. S. Ponder, Chem. Soc. Rev. 2012, 41, 1485 – 1498.
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org [33] R. A. Sheldon, Green Chem. 2007, 9, 1273 – 1283. [34] K. Sanderson, Nature 2011, 469, 18 – 20.
Received: May 3, 2014 Revised: May 22, 2014 Published online on July 22, 2014
ChemSusChem 2014, 7, 2081 – 2088
2088
High-Performance Polymers from Nature: Catalytic Routes and Processes for Industry Guido Walther*[a] In memoriam of Dr. Michael Kant
It is difficult to imagine life today without polymers. However, most chemicals are almost exclusively synthesized from petroleum. With diminishing oil reserves, establishing an industrial process to transform renewables into high-value chemicals may be more challenging than running a car without gasoline. This is due to the difficulty in setting up processes that are novel, profitable, and environmentally benign at the same time. Additionally, the quest for sustainability of renewable resources should be based on incorporating ethical considerations in the development of plans that utilize feedstocks intended for human nutrition and health. Thus, it is important to use bio-energy containing renewable resources in the most efficient way. This Concept goes beyond the synthesis of monomers and provides insights for establishing an industrial pro-
cess that transforms renewable resources into high-value chemicals, and it describes careful investigations that are of paramount importance, including evaluations from an economical and an ecological perspective. The synthesis of monomers suitable for polymer production from renewable resources would ideally be accompanied by a reduction in CO2 emission and waste, through the complete molecular utilization of the feedstock. This Concept advocates the drop-in strategy, and is guided by the example of catalytically synthesized dimethyl 1,19-nonadecanedioate and its a,w-functionalized derivatives. With respect to the Twelve Principles of Green Chemistry, this Concept describes a technological leap forward for a sustainable green chemical industry.
1. Introduction Ever since our distant ancestors managed to light fire, humankind has been developing a continuously increasing quest for energy. With a world population greater than 7 billion, growing rapidly, and fossil resources that are diminishing, the energy demand needs to be covered by a smart energy mix. This might be a solvable challenge in due course, since many renewable energy resources are already in use (e.g., hydropower, geothermal energy, wind and solar energy, bioenergy).[1, 2] However, renewable resources, such as bio-oil and biomass, are scarce and limited because of their availability varies per season. These renewables are indeed the only resources providing both energy that is stored in chemical bonds and a useful molecular structure of carbon atoms. The quest for sustainability of these renewable resources should therefore be based on incorporating ethical considerations in the development of plans that utilize feedstocks intended for human nutrition and health. Thus, it is important to use bioenergy containing renewable resources in the most efficient way. With a view to petroleum-based polymers, there is a quest for finding ways to synthesize polymers that possess properties similar to, for example, polyethylene.[3, 4] Many polymers are sophisticated, product-oriented and high-performance materials, and these all have in common outstanding thermal,
[a] Dr. G. Walther Leibniz Institute for Catalysis at the University of Rostock Albert-Einstein-Str. 29 A, 18059 Rostock (Germany) E-mail: [email protected]
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
mechanical, electrical, optical, surface, and rheological properties. Petroleum-based examples of such specialty polymers, produced on a scale of several 100 Mton annually, are polyether ether ketone (PEEK), polytetrafluoroethylene (PTFE), and polyamide 12 (PA12). The current size of the market for biobased high-performance polymers [e.g., polyphthalamide (PPA), PA10.10 or PA10.12][5] covers about 0.7 Mton annually, and this figure is rapidly increasing. These materials, mainly polyamides, have found application in areas such as the aerospace industry, extrusion processes, and microelectronic compounds.[6] Indeed, the molecular structure of carbon atoms in plant oils is particularly well-suited to the synthesis of long-chain polymerizable segments. Even though the hydroformylation of alkenes is a well-established synthesis route, it cannot achieve a yield of linear products greater than 36 %.[7] This is due to a thermodynamically unfavorable intermediate reaction step, in which a terminal C=C double bond is temporarily formed. Linear monomers used for polymerization require a very high grade of purity, which is why so far long-chain hydrocarbons (longer than C12) have been used only rarely. A pioneering discovery in this field was made Cole-Hamilton et al. in 2005, using the catalytically active and highly selective palladium complex Pd/o-C6H4(CH2PtBu2)2 in a methoxycarbonylation route towards linear reaction products.[8] In the presence of methane sulfonic acid (MeHSO3) this catalyst was found to convert highly reactive ethylene into methyl propanoate (cf. Scheme 1 a), and was hence commercialized for the petroleChemSusChem 2014, 7, 2081 – 2088
2081
CHEMSUSCHEM CONCEPTS
www.chemsuschem.org
Besides the promising developments in methoxycarbonylation reactions, there are various alternative routes to obtain long-chain hydrocarbons that can be functionalized at their terminal position. A recent critical Review presented four reactions, taking the role of key reactions for processing renewable resources: Metathesis, which was reported to shorten the carbon chain length while fuctionalizing them at the same time; hydroformylation, which provides access to aldehydes, intermediates from which other a,wfunctionalized derivatives can be obtained using well-known technologies; degradation, which defunctionalizes renewables to access new active reaction Scheme 1. Inspired by the Lucite process (a), we developed a combined transesterificasites; and also isomerization, in combination with tion, isomerization, and methoxycarbonylation reaction of plant oil toward 1,19-diester (b), using Pd/o-C6H4(CH2PtBu2)2. R, R’ and R’’ consist of linear mono- or polyunsaturated other techniques such as methoxycarbonylation.[21] hydrocarbon chains of 17 carbon atoms. FAME denotes fatty acid methyl esters of polyWith a view to making long-chain crystallizable unsaturated hydrocarbon chains. a,w-functionalized segments from sunflower oil, olefin metathesis would provide a complex mixture of a,w-functionalized monomers. More promising results were described for hydroformylation reactions.[21] Howevum-based Lucite process.[8–10] Inspired by the Lucite process, an isomerizing methoxycarbonylation route was developed to er, the selectivity towards a,w-dialdehydes was reported as still transform methyl oleate (derived from sunflower or rapeseed being very challenging. In addition, all reactions required transoil) into dimethyl 1,19-nonadecanedioate (1,19-diester).[3, 4] A ersterification of the sunflower oil. Thus, all reactions are key finding we made is that this catalyst also promotes the tandem reactions in different vessels.[21] The combined reaction direct conversion of sunflower oil into 1,19-diester, (cf. of transersterification, isomerization, and methoxycarbonylaScheme 1 b),[11–13] which may open the door to new catalytic tion may provide both high yield and high selectivity of the routes and processes for industry. desired product, and proceeds as a one-pot reaction. This would make the formation and distillation of intermediate methyl esters redundant.[11–13]
2. Results 2.1 Plant oil utilization
2.2 a,w-Diester—A novel platform chemical
At present, mainly fruits that consist of oil and sugar are used for energy conversion. The rest of the plant matter, even though it contains energy stored in chemical bonds, is only used scarcely, because the technology is still in its infancy.[14] The distribution of vegetable oil consumption between food, feed, and industrial use was 74:6:20 in 2008.[15] This ratio is continuously shifting toward higher industrial use, of which approximately 30 % accounts for biobased transportation fuel— this equals 6 % in total.[15] Within the last decade, the methoxycarbonylation of lower olefins as well as unsaturated carboxylic acids and esters has received much attention; besides experimental investigations,[4, 8–13, 16–19] tremendous efforts have been undertaken to elucidate a potential mechanism.[9, 10, 17, 20] However, prior to processing olefins according to the Lucite process (cf. Scheme 1 a), the plant oil needs to be converted into its corresponding fatty acid methyl esters and glycerol by transesterification. Even though the transesterification of plant oil, which is the first step for utilizing this resource as a fuel (e.g. biodiesel) for the transportation sector, is available commercially on a large scale, it would be desirable to obtain a long nonbranched (CH2)n a,w-functionalized crystallizable segment directly from the raw material. Such a segment may consist of n = 19 or 23 CH2 units, which may be formed from oleic or erucic acid moieties, respectively.
Recently, a possibility has been developed to directly synthesize 1,19-diester from high-oleic sunflower oil via a highly selective combined reaction involving transesterification, isomerization, and methoxycarbonylation.[11–13] This reaction proceeds already at 80 8C in an acidic environment (9.6 mol % CH3SO3H) and pCO = 30 bar, using 2.4 mol % Pd/o-(C6H4)(CH2PtBu2)2 as catalyst. The molecular ratio of palladium to ligand was always kept at 1:5. When the volumetric ratio of methanol to plant oil equaled 3, we obtained a yield (Y) of 91 % at a selectivity (S) of 94 % of the 1,19-diester. The best result (Y = 97 %, S = 97 %) was achieved when increasing the volumetric ratio of methanol to plant oil to 4 and the catalyst quantity to 4.8 mol %.[13] However, the latter set of parameters does not meet the best result when considering its “greenness”. This is associated with the high quantities of catalyst and solvent used, and will be revisited in this Concept when discussing environmental aspects. This direct synthesis process was found to be suitable for scale-up by several orders of magnitude, without any significant decrease of yield and selectivity.[11–13] Traces of water do not influence the reaction, which allows the use of technicalgrade methanol. However, greater quantities of water do hinder the reaction’s progress. With a view to the acidic reaction environment required, H2SO4 was found to be suitable in slightly greater quantities to obtain comparable yields of > 80 %.[18] Another important finding is that polyunsaturated
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2081 – 2088
2082
CHEMSUSCHEM CONCEPTS and branched hydrocarbon moieties are not converted into the desired a,w-diesters,[18] instead, a,x-diesters, with x < w, are obtained.[19] An comparison, albeit a simplistic one, of a two-stage process from sunflower oil via methyl oleate toward a,w-diesters and the direct synthesis route shows the potential advantage of the latter process, as it may be lead to a reduction of CO2 emissions and of waste. Permitting that any erroneous assumptions are consistently made for both processes, this is particularly the case for the power consumption when compressing CO and heating/cooling the reaction inventory based either on renewable or on fossil energy. Moreover, the direct synthesis route does not require any additional chemicals, such as NaOH and KOH, which are needed for the two-stage process. Additionally, it should be noted that the glycerol obtained from the transesterification of the plant oil into the methyl ester is salt-contaminated and, in the least favorable case, needs to be considered as a waste. The direct synthesis process may provide salt-free glycerol as a by-product. Both glycerol[22] and the desired a,w-functionalized long-chain hydrocarbon diester[13] have indeed been identified as being platform chemicals, as illustrated in Scheme 2. Knowledge of the reaction mechanism is crucial for understanding and improving the properties of catalysts. To overcome limitations such as reactivity and stability, we examined this reaction experimentally and developed a microkinetic
www.chemsuschem.org model from extensive density functional theory (DFT) calculations reflecting the current state of science.[20] An important fact to consider is that the active palladium species is present as a hydride (cf. Scheme 3), H-[Pd] (0), where [Pd] denotes the organometallic Pd/o-(C6H4)(CH2PtBu2)2 catalyst. This was confirmed by electron paramagnetic resonance (EPR) spectroscopy. The thermal stability of the catalyst was found to be of minor importance, since the decomposition of the ligand is provoked more by the methanolysis toward the desired a,w-diester. This elementary step was identified as being the rate-determining step and having a Gibbs free energy of 137.5 kJ mol1. A high-pressure NMR cell[23] equipped with a gas flow and gas saturation system and gas recycling unit[24] was used to corroborate the findings on the making and breaking of bonds, especially with regard to the PdP bond of the organometallic complex. In situ UV-vis spectroscopy provided insight into the reversibility of a change in the ligand sphere of the [Pd] catalyst while the combined isomerization–methoxycarbonylation reaction of methyl oleate toward a,w-diester proceeded. This result agrees with the findings from activity measurements.[20] Hence, a potential mechanism may be suggested for the combined transesterification, isomerization, and methoxycarbonylation of plant oils, cf. Scheme 3. A thorough thermodynamic analysis provided detailed insight into the elementary steps. Some intermediate steps were
Scheme 2. Principle routes of the value-added chain from renewable resources (plant oils) that are directly converted into platform chemicals (a,w-diester and glycerol), via oleochemical building blocks such as the corresponding a,w-diacid, the a,w-diol, and the a,w-diamine, toward the desired polymers using existing technologies for polymerization.
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2081 – 2088
2083
CHEMSUSCHEM CONCEPTS
www.chemsuschem.org methoxycarbonylation reaction, such as o-(C6H4)(CH2PPh2)2, which was inactive, o-(C6H4)(CH2PtBu2) (CH2PPh2).[4, 10] Surprisingly, the latter (nonsymmetric) ligand was found to be as active as o-(C6H4)(CH2PtBu2)2 but much more robust against decomposition. From the DFT study, it may be assumed that the tBu moieties bound to the phosphorus atom of the ligand provide charge transfer, necessary for the catalytic reaction. On the other hand, tBu moieties are less robust than Ph moieties. However, Ph groups do not provide enough steric hindrance to fully isomerize the C=C double bond toward the terminal position. With respect to the costs of the catalyst, the question arises whether a ligand that is symmetric but contains a mix of tBu and Ph moieties, for example, o-(C6H4)(CH2PtBuR)2, where R denotes Ph or Ad, meets the properties of o-(C6H4)(CH2PtBu2)(CH2PPh2).
Scheme 3. Catalytic cycle of the synthesis of a,w-diesters from methyl oleate (n + m = L = 14) or methyl erucate (n + m = L = 18) present in triglycerides of the plant oil. Elementary steps are described in the text.
examined in greater accuracy; for example, the isomerization (1) of the C=C double bond proceeds in both directions with low barriers (2); however, a solvent effect was found to favor the formation of an n-alkyl (3). According to the consecutive addition of CO and methanol, as interpreted from the experimental results, CO insertion was not found to be an elementary step that is reversible; it rather undergoes a quasi-equilibrated elementary step (4). Nevertheless, the rate-determining step is the methanolysis, which proceeds favorably via intramolecular attack of methanol (5). This elementary step might be attended by PdP bond cleavage, which may lead to precipitation of Pd0. A microkinetic model was developed to calculate the apparent reaction barrier for the formation of 1,19diester from sunflower oil. This result was found to agree well with the experimental findings, and may thus be taken as indirect corroboration of the proposed reaction mechanism.[20] Additionally, the same data predicted this reaction to also occur when using other feedstocks and acids. To verify this hypothesis, different plant oils (e.g. sunflower oil, rapeseed oil, soybean oil, castor oil, olive oil, peanut oil) and acids were studied. Experimental results showed that Pd/o-(C6H4)(CH2PtBu2)2 is a very selective catalyst for the synthesis of a,wdiesters. However, the a,w-diesters are always obtained from oleic or erucic acid moieties present in the triglyceride of the plant oil. Other fatty acid moieties, independent of whether they are saturated or polyunsaturated, or even OH-branched monounsaturated ricinoleic acid moieties do not influence the reaction.[18] Surprisingly, these simple facts were not perceived before the reaction mechanism was elucidated; although recently a,w-diesters were already found to be obtained from other plant oils as well.[16] Finally, with the use of (technical) H2SO4 as co-catalyst, the catalytic system was found to open new routes for industrial applications.[18] Besides palladium-based organometallic catalysts containing o-(C6H4)(CH2PtBu2)2 as ligand, other ligands were tested for the 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2.3 Other a,w-functionalized monomers—Oleochemical building blocks Other a,w-functionalized derivatives of the a,w-diesters that are presumably of interest for polymerization can not be obtained directly from plant oils. Such derivatives include highly purified a,w-diacids, a,w-diols, and a,w-diamines. However, we synthesized them starting at the a,w-diesters. For example, highly purified a,w-diacids were obtained from a,w-diesters by alkaline hydrolysis and one subsequent re-crystallization (purity > 99 %).[13] The catalytic hydrogenation of nonactivated esters is challenging. Alternatively, esters have been reduced using stoichiometric quantities of metal hydrides, such as LiAlH4, but this has been found to be ineffective for the synthesis of long-chain hydrocarbon a,w-diols.[4] However, a few systems that catalyze the hydrogenation of non-activated esters have been reported.[25–27] Surprisingly, the ruthenium hydride complex [(PNN)RuH(CO)], based on the pincer ligand PNN (2-(di-tert-butylphosphinomethyl)-6-(diethylaminomethyl)-pyridine),[27] was found to catalyze the reduction of a,w-diester towards its corresponding a,w-diol.[13] In actual fact, this catalyst operates already under mild conditions, such as # = 115 8C and pH2 10 bar, and we obtained a yield of the a,w-diol of 98 % using 0.5 mol % catalyst. More sophisticated is the synthesis of linear but also primary a,w-diamines. Inspired by the selective synthesis of primary amines directly from alcohols and ammonia,[28] we found that this reaction also proceeds from a,w-diol toward a,w-diamine. This reaction was catalyzed at rather mild conditions using the dearomatized acidine-based pincer complex [RuHCl(A-iPr-PNP) (CO)], where PNP denotes 4,6-(di-iso-propylphosphinomethyl)acridine, developed by Milstein et al.[28] At already 140 8C, we obtained a yield of primary a,w-diamines of 68 %, using only 0.25 mol % catalyst, an excess of ammonia, and 2-methyl-2-butanol as solvent.[13, 29] Within this context, the a,w-diester occupies a pivotal role as a platform chemical, with its derivative monomers, such as linear aliphalic a,w-diacids, a,w-diols, and a,w-diamines, acting as oleochemical building blocks for the polymer industry, cf. Scheme 2.[13] ChemSusChem 2014, 7, 2081 – 2088
2084
CHEMSUSCHEM CONCEPTS 2.4 Polymer synthesis The presented oleochemical building blocks as well as the identified platform chemical a,w-diester were found to have similar properties as other polymerizable monomers of shorter segments.[4, 30] Thus, for polymerizing these novel monomers existing technologies may be used. Quinzler et al. synthesized polyesters made up of alternating a,w-diesters and a,w-diols that contain 19 or even 23 (CH2) segments. The polycondesation of stoichiometric quantities of a,w-diester and a,w-diol was reported being catalyzed using Ti(OBu)4 at a temperature range of 110 8C–150 8C and at a pressure of 0.01 mbar. After a reaction time of 17 h and cooling, a quantitative yield was obtained.[4] Polyamides, such as PAx.19, where x E {6, 9, 10, 12, 19},[5] were synthesized using melt polycondensation on a pilot plant scale, where equimolar amounts of a,w-diamines and a,w-diacids were employed. Typical reaction conditions were # = 240– 260 8C, p = 18–26 bar, and in some cases an admixture of 50 % distilled water and were kept invariant for about 2 h.[30] At first glance, we found that the melting temperature decreases with increasing length of the monomers. Additionally, thermal and mechanical properties are very close to those of the established PA12. A unique selling point of these novel high-performance polymers—especially PA6.19 and PA10.19—is their very high translucence compared to that of PA12, which is beneficial for many applications, but also improved thermostability and high impact strength. This trend could be confirmed for PAx.23, where x E {6, 9, 10}, as well.[30] Polyurethanes are formed by reacting a polyol with an isocyanate. The length of the linear hydrocarbon chain of the polyol is expected to mainly contribute to the polyurethanes’ properties, such as elasticity and good hydrolysis resistance. The reaction of long-chain a,w-diacids [e.g., (CH2)n,with n = 19 or 23] and short-chain a,w-diols (e.g. diethyleneglycole) results in polyesterpolyols which, when further treated with diisocyanites using polycondensation, yield polyurethanes. However, since highly purified a,w-diols of a chain length of either 19 or 23 (CH2) units are available, these monomers may directly react with isocyanates to form polyurethanes. Since the 1,19-diol has a high melting point, and the synthesis of polyurethanes re-
www.chemsuschem.org quires the reactants being in liquid form, the polyurethane formation is accelerated so that the processing time for forming a certain compound is limited. However, the reaction behavior was controlled best at # = 100–110 8C when using hexamethylene diisocyanate and an organo-metal catalyst.[30] It was found that the hydrophobic domain of the polyurethanes based on C19 monomers is more beneficial for hydrolysis resistance (even at 70 8C) than the one of C12 and shorter monomers.
3. Discussion 3.1 Drop-in strategy—Synthesis route of high-performance polymers for industry Establishing an industrial process that transforms renewable resources into high-value chemicals may primarily focus on the potential value chains in the utilization of the feedstocks. In a more general sense, producing high-performance polymers from nature follows the drop-in strategy. The drop-in strategy is characterized by an existing value chain, utilizing existing infrastructure and a mature market.[17] Hence, existing technologies for the final polymerization process do not need to be changed, when platform chemicals (here a,w-diester and glycerol) and intermediates obtained from them “drop into” an existing process, illustrated in Figure 1 by interlocking the renewable production route (top) with the petroleum-based route (bottom). From the left to the right, the bottom tiles Figure 1 depict crude oil extraction, refinery, platform chemicals, monomers (petroleum-based building blocks), existing industrial polymerization process, and final products. By analogy, renewable resources are turned into higher-value chemicals along the value-adding chain using a catalyst; these chemicals are already basic units (platform chemicals) that after functionalization serve as initial feedstock for polymerization, cf. top tiles. Polymers are then synthesized from such monomers, using existing fossil fuel-based technologies. Additionally, the simple up-scalable direct synthesis process of a,w-diesters from plant oils is essential to make such highperformance polymers competitive for the global market.
Figure 1. Drop-in strategy for platform chemicals (here: a,w-diester and glycerol) into existing processes.
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2014, 7, 2081 – 2088
2085
CHEMSUSCHEM CONCEPTS
www.chemsuschem.org
3.2 Environmental aspects: Polymer, plate, or fuel tank? A sensitive issue is the ethical question of the competition of renewable bio-resources between “polymer, plate, and fuel tank”. In order to assess the severity of the consequences, objective criteria can be found, on the one hand, in the quantity of plant oil required to meet the demand of high-performance polymers, and, on the other hand, in the fact that the desired a,w-diesters are exclusively made from nonbranched, monounsaturated alkyl chains present in plant oils. High-oleic sunflower oil contains only 4 % of polyunsaturated moieties, known as w3 and w6 fatty acid derivatives. However, this special breeding of sunflowers does not contain any w3 fatty acid moieties at all. Such plant oils are less suitable for human consumption than those containing greater quantities of polyunsaturated moieties. In the European Union, Article 13 Health Claims says that a nutrition product is a source of w3 fatty acids, when it contains at least 0.3 g alpha-linolenic acid per 100 g. In contrast, rapeseed oil contains more than 35 % polyunsaturated moieties.[18] From this perspective, there might not be the assumed ethical conflicts. This again is affirmed by the costs of such high-performance polymers. Another important issue arises with the question of how much of the cultivated area may be used for sunflower breeding and which location suits best in, for example, Europe to maximize the annual harvest. This is of course dependent on conditions of the soil and the climate. In Germany, the crop yield of rapeseed oil is 1.7–1.8 m3 Ha1, while that of sunflower oil is only 0.7 m3 Ha1. With a view on Europe, the conditions for sunflowers to thrive and prosper are best in the south of France and in Spain. Additionally, edibles are high-quality products from agriculture but quite low in their appraisal, which is provoked politically in Germany. Increasing food prices would raise the price of all products that have an added value. However, politicians are currently pursuing the goal of blending and substituting fossil fuels by bioethanol and biodiesel, respectively. With a view to the role biofuel may play in the future energy supply,[31] the energy density stored in the chemical bonds of the resource should be considered. This is a function of the effective H/C ratio,[1] H=C ratio ¼
nðHÞ 2nðOÞ 3nðNÞ nðCÞ
ð1Þ
For a direct comparison, assuming a copolymer such as PA19.19, made of alternating 1,19-diamine and 1,19-diacid monomers, the H/C ratios for the 1,19-diamine, 1,19-diacid, and PA19.19 are 36:19, 28:19 and 64:38 (= 32:19), respectively; while the H/C ratio of biodiesel is less than 34:19. The reason that biodiesel does not achieve 34:19 is due to polyunsaturated hydrocarbon chains present in the fuel. The quantity of energy stored in the chemical bonds does not differ significantly when comparing biodiesel with the sequence of polymers made of the same raw material. Hence, the question arises: why not utilize renewable resources for the production of higher value-added contaminant-free polymers, before reusing them for zero-emission power generation. To complete this 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 2. Comparison of the value-added chain starting at rapeseed and following the goal currently pursued by politicians, blending and substituting fossil fuels by bioethanol and biodiesel, respectively, with that of the proposed concept for the production of high-performance polymers from nature. The smallest environmental and ethical conflict may be assumed when using sunflowers as feedstock.
view of zero emission, methanol and CO, also required within this process, may be synthesized from syngas and by gasification of the residual biomass after harvesting the plant oil from the seeds, respectively. Figure 2 provides a detailed overview of this concept. Going beyond the CO2 liberated into our atmosphere, one has to distinguish between a small quantity of CO2 that is recycled forming C and O2 by the chlorophyll of the plants, and an excess quantity of CO2. It is widely accepted that the excess CO2 causes the serious environmental problem known as “global warming” or “greenhouse effect”, but there is an even worse environmental problem being considered: The excess CO2 in our atmosphere (that is no longer recycled by plants) is an admission of O2 depletion; the O2 covered in the excess CO2 was originally breathable and thus an increasing extremely serious environmental problem for mankind. On a metric scale, recalling the atomic weight of CO2 (44 u) and O2 (32 u), the daily loss of O2 equals 32/44 73 % of the excess CO2. Even though the global warming is generally considered, it seems that governments and industries do not face the fact that we need oxygen to breathe. Instead, CO2 capturing and storage in depositories underground is believed to solve the environmental problem, which in the author’s opinion only shifts the main challenge of this generation to one of the next generation. It is an important task—maybe a basic need for the survival of mankind—to find solutions to reduce the quantity of excess CO2. On a broader perspective, this is why resources need to be used in their most efficient way. The “greenness of chemical processes and the products”, as promoted by Jimnez-Gonzlez et al.,[32] shall be evaluated for the proposed strategy using different metrics for resources, solvents, renewability, and efficiency. Such metrics are the mass intensity, P
Im ¼
i mðmi Þ mð1,19-diesterÞ
ð2Þ
where m(mi ) is the mass of material mi ; the solvent intensity, P
Is ¼
i mðsi Þ mð1,19-diesterÞ
ð3Þ
ChemSusChem 2014, 7, 2081 – 2088
2086
CHEMSUSCHEM CONCEPTS
www.chemsuschem.org
where m(si ) is the mass of solvent si ; the renewables intensity, P
Ir ¼
i mðri Þ mð1,19-diesterÞ
ð4Þ
where m(ri ) is the mass of renewable ri ; and the atom economy, Mð1,19-diesterÞ P ae ¼ i MðRi Þ
ð5Þ
where M(Ri ) is the molar mass of the reactants Ri, respectively.[32] For a typical batch reaction according to direct synthesis reaction process of 1,19-diester (VMeOH = 180 mL, Vplant oil = 60 mL, ncatal = 2.4 mol %, nCH3SO3H = 9.6 mol %, pCO = 30 bar, # = 80 8C, and t = 32 h), calculations estimate Im 3.76 kg/ kg1,19-diester, Is 2.48 kg/ kg1,19-diester, Ir 3.65 kg/ kg1,19-diester, and ae 62 %, where CO and methanol are assumed to be synthesized from the residual biomass after harvesting the plant oil from the seeds. Another metric is the E-factor, which provides a measure of the quantity of waste produced per quantity of product.[33] Hence, the author recommends considering glycerol as a coproduct that does not need to be treated as waste. The direct synthesis process according to Scheme 1b was found to provide salt-free glycerol,[30] which again confirms the complete utilization of the feedstock. However; a life cycle assessment would provide an in-depth analysis, especially with a view to zero-emission CO and methanol synthesis from biomass and power generation.
Conclusions This Concept provides a strategy synthesizing high-performance polymers from renewable resources using catalytic routes and suggests a process for industry. It seems that currently the wish of serving a continuously increasing economy dominates the choice in which concept being invested. A cheaper producible alternative catalyst may be based on a symmetric ligand, such as Pd/o-(C6H4)(CH2PtBuR)2, where R denoted Ph or Ad. However, there is a lack of the catalytic characterization of such systems. From an environmental point of view, producing high-performance polymers from nature allows the use of safer auxiliaries (e.g. less hazardous acid), catalysts and renewable feedstocks, and provides an energy efficient method for preventing waste material by almost completely utilizing the feedstock. In my opinion, this would contribute to increasing the yield of the desired product and reduce significantly the production costs and the CO2 emission. With respect to the Twelve Principles of Green Chemistry,[34] this concept is a technological leap forward for a sustainable green chemical industry. 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Acknowledgements The author is grateful for an interview with Martin Schulz (President of the European Parliament). Fruitful discussions with Bernhand Conzen (President of the Farming Community Heinsberg, Germany), Angela Kçckritz (LIKAT) and Franz-Erich Baumann (Evonik Ind.) are acknowledged. The TOC graphic was designed by Duncan J. Mowbray. Keywords: catalysis · plant oils · polymers · renewables · sustainable chemistry
[1] P. N. R. Vennestrøm, C. M. Osmundsen, C. H. Christensen, E. Taarning, Angew. Chem. Int. Ed. 2011, 50, 10502 – 10509; Angew. Chem. 2011, 123, 10686 – 10694. [2] J. J. Bozell, Science 2010, 329, 522 – 523. [3] D. J. Cole-Hamilton, Angew. Chem. Int. Ed. 2010, 49, 8564 – 8566; Angew. Chem. 2010, 122, 8744 – 8746. [4] D. Quinzler, S. Mecking, Angew. Chem. Int. Ed. 2010, 49, 4306 – 4308; Angew. Chem. 2010, 122, 4402 – 4404. [5] PAx.19: polyamide made of alternating monomer units, where x represents the chain length of the a,w-diamine and 19 the chain length of a,w-nonadecanediacid used for synthesis; PUR19: polyurethane made of a,w-nonadecanediol. [6] http://plasticeurope.de. [7] A. Behr, D. Obst, A. Westfechtel, Eur. J. Lipid Sci. Technol. 2005, 107, 213 – 219. [8] C. Jimnez-Rodriguez, G. R. Eastham, D. J. Cole-Hamilton, Inorg. Chem. Commun. 2005, 8, 878 – 881. [9] V. de La Fuente, M. Waugh, G. R. Eastham, J. A. Iggo, S. Castillon, C. Claver, Chem. Eur. J. 2010, 16, 6919 – 6932. [10] T. Fanjul, G. Eastham, N. Fey, A. Hamilton, A. G. Orpen, P. G. Pringle, M. Waugh, Organometallics 2010, 29, 2292 – 2305. [11] F.-M. Petrat, F.-E. Baumann, H. Hger, G. Walther, A. Martin, A. Kçckritz, DE 10 2010 002 809A1. [12] F.-M. Petrat, F.-E. Baumann, H. Hger, G. Walther, A. Martin, A. Kçckritz, WO 2011/110249A1. [13] G. Walther, J. Deutsch, A. Martin, F.-E. Baumann, D. Fridag, R. Franke, A. Kçckritz, ChemSusChem 2011, 4, 1052 – 1054. [14] G. Berndes, M. Hoogwijk, R. van den Broek, Biomass Bioenergy 2003, 25, 1 – 28. [15] F. D. Gunstone, Lipid Technol. 2008, 20, 48. [16] M. R. L. Furst, R. Le Goff, D. Quinzler, S. Mecking, C. H. Botting, D. J. Cole-Hamilton, Green Chem. 2012, 14, 472 – 477. [17] P. Roesle, C. J. Durr, H. M. Mçller, L. Cavallo, L. Caporaso, S. Mecking, J. Am. Chem. Soc. 2012, 134, 17696 – 17703. [18] G. Walther, A. Martin, A. Kçckritz, J. Am. Oil Chem. Soc. 2013, 90, 141 – 145. [19] M. R. L. Furst, T. Seidensticker, D. J. Cole-Hamilton, Green Chem. 2013, 15, 1218 – 1225. [20] G. Walther, L. R. Knçpke, J. Rabeah, M. P. Che˛cin´ski, H. Jiao, U. Bentrup, A. Brckner, A. Martin, A. Kçckritz, J. Catal. 2013, 297, 44 – 55. [21] A. Behr, A. J. Vorhold, K. A. Ostrowski, T. Seidensticker, Green Chem. 2014, 16, 982 – 1006. [22] J. A. Kenar, Lipid Technol. 2007, 19, 249 – 253. [23] G. Walther, A. Martin, A. Kçckritz, D. Selent, DE 10 2011 007 527A1. [24] D. Selent, W. Baumann, A. Bçrner, DE 10 3331 143 B4. [25] S. N. Ege, Organic Chemistry, D. C. Heath&Company, Lexington, 1989, p. 596. [26] Y. Pouilloux, F. Autin, J. Barrault, Catal. Today 2000, 63, 87 – 100. [27] J. Zhang, G. Leitus, Y. Ben-David, D. Milstein, Angew. Chem. Int. Ed. 2006, 45, 1113 – 1115; Angew. Chem. 2006, 118, 1131 – 1133. [28] C. Gunanathan, D. Milstein, Angew. Chem. Int. Ed. 2008, 47, 8661 – 8664; Angew. Chem. 2008, 120, 8789 – 8792. [29] F.-E. Baumann, A. Martin, M. Ullrich, M. Roos, P. Hannen, F.-M. Petrat, H. Hger, A. Kçckritz, G. Walther, J. Deutsch, EP 2 484 659A1.
ChemSusChem 2014, 7, 2081 – 2088
2087
CHEMSUSCHEM CONCEPTS [30] F.-E. Baumann, U. Fehrenbacher, T. Hoffmann, D. Fridag, A. Kçckritz, M. Kugler, F.-M. Petrat, B. Tbke, G. Walther, Final report: Neuartige Polymere auf Basis von Triglyceriden; Eigenverlag des Fraunhofer ICT, Germany, 2012. See www.fnr-server.de/ftp/pdf/berichte/22001308.pdf (accessed June 2014). [31] L. Caspeta, N. A. A. Buijs, J. Nielsen, Energy Environ. Sci. 2013, 6, 1077 – 1082. [32] C. Jimnez-Gonzlez, D. J. C. Constable, C. S. Ponder, Chem. Soc. Rev. 2012, 41, 1485 – 1498.
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org [33] R. A. Sheldon, Green Chem. 2007, 9, 1273 – 1283. [34] K. Sanderson, Nature 2011, 469, 18 – 20.
Received: May 3, 2014 Revised: May 22, 2014 Published online on July 22, 2014
ChemSusChem 2014, 7, 2081 – 2088
2088
