Arno
Behr
*,
Jens
Eilting
,
Ken
Irawadi
,
Julia
Leschinski
and
Falk
Lindner
Department of Biochemical and Chemical Engineering, Technical Chemistry A, Dortmund University, D-44227 Dortmund, Germany
First published on 12th November 2007
Although glycerol has been a well-known renewable chemical for centuries, its commercial relevance has increased considerably in the last few years because of its rising inevitable formation as a by-product of biodiesel production. The present review gives a broad overview on the chemistry of glycerol starting from the classic esters and oligomers to new products like glycerol carbonate, telomers, branched alkyl ethers, propanediols and epoxides. In particular, the novel possibilities to control the numerous addition, reduction and oxidation reactions via heterogeneous, homogeneous and biocatalysis will be presented. A benchmark will be given to determine the products which will have the best chances of entering the market and which processes are currently most developed.
From left: Arno Behr, Jens Eilting, Julia Leschinski, Falk Lindner and Ken Irawadi | Arno Behr studied chemistry at Aachen University, Germany and did his PhD in 1979. After 10 years’ work in industry he assumed a position at the Chair of Technical Chemistry, University of Dortmund, Germany, in 1996. One of his research interests is the chemistry of renewable resources. Ken Irawadi earned his BSc in Agriculture Technology from Bogor Agricultural University, Indonesia and his MSc in Process Engineering at the TU Hamburg, Germany. He has been a lecturer and a researcher at Bogor Agricultural University. Since October 2006 he has been a PhD student under the supervision of Prof. Dr Arno Behr at the University of Dortmund, Germany. Julia Leschinski studied chemical engineering at the University of Dortmund, Germany. She was awared her Dipl.-Ing. in 2005 and has started her PhD studies in Dortmund at the division of Chemical Process Development under the supervision of Prof. Dr Arno Behr. Jens Eilting studied chemical engineering at the University of Dortmund, Germany. He got his Dipl.-Ing. in September 2005. In November 2005 he started his PhD studies in Dortmund at the division of Chemical Process Development under the supervision of Prof. Dr Arno Behr. Falk Lindner studied chemical engineering at the University of Dortmund, Germany. He received his diploma in 2005 and works now as a research assistant at the division of Chemical Process Development/Technical Chemistry A under the supervision of Prof. Dr Arno Behr. |
In 1866 the first technical usage of glycerol was the production of the trinitrate of glycerol, the so-called nitro glycerol. Dynamite is formed when nitro glycerol is adsorbed on diatomaceous earth, which can be used as an explosive, for instance, for the construction of tunnels, channels or railroad routes. Just for the construction of the Panama channel (started in 1879) 30000 tons of dynamite were applied. At the end of the nineteenth century the processing of natural oils and fats and hence the production of glycerol increased continuously.
Nowadays the production of glycerol from fats and oils is carried out by saponification yielding glycerol and soaps (Scheme 1, Y = ONa), by hydrolysis yielding glycerol and fatty acids (Scheme 1, Y = OH) or by transesterification with methanol yielding glycerol and fatty acid methyl esters (Scheme 1, Y = OCH3).
Scheme 1 Synthesis of glycerol from fats and oils. |
The commercial production of fats and oils has increased rapidly during the last century, especially in the last three decades. Whereas in 1970 the total production of fats and oils was only in the range of 40 million tons (Mt), this value climbed to about 144 Mt in 2005. A further considerable rise to about 200 Mt can be foreseen for the year 2015. However, it must be stressed that only a small part of this material is converted within the chemical industry: about 81% of the worldwide oil production is used for the food industry, 14% in the chemical industry and 7% for feed. Today more than 20 Mt annum–1 of fats and oils are processed chemically, yielding large amounts of glycerol.
A new trend has overlapped with this increasing production of natural oils and fats: for several years, renewable oils have also been used for energy applications. The so-called “biodiesel” has been implemented as automotive fuels in the US and in Europe. Biodiesel is a popular term for the fatty acid methyl esters formed according to Scheme 1, for instance from the transesterification of rapeseed oil with methanol. Biodiesel production in the EU was estimated to be about 6 Mt in 2006 and is forecasted to increase to about 12 Mt in 2010.
These trends influence remarkably the glycerol market. Before starting biodiesel production, the European glycerol market was in the range of 250000 to 400000 t annum–1. Now, additional glycerol stemming from biodiesel production is flooding the market. One ton of biodiesel yields about 110 kg of crude glycerol or about 100 kg of pure glycerol. Therefore, by 2010 about 1.2 Mt annum–1 of additional glycerol will enter the European market. Of course this influences substantially the glycerol price; while the glycerol price moved in the range of 1000 to 1300 € t–1 between 2000 and 2003, from 2004 to 2006 it decreased to 500 to 700 € t–1. Besides the high quality glycerol, there is a large quantity of impure glycerol on the market available for very low prices.
Obviously the question arises how this additional glycerol can be used wisely. There is already a great number of common glycerol applications; for instance, glycerol is used in pharmaceuticals, cosmetics (hair and skin care), soaps and toothpastes. Also the direct use as sweetener in candies and cakes and as a wetting agent in tobacco is widespread. Some chemical functionalisation of glycerol is also carried out in industry, especially the synthesis of some esters, polyethers and alkyd resins. There seems to be little scope to extend the uses of glycerol in these areas. So the question arises whether there are any additional fields of application, where glycerol itself or some of its derivatives can find new handlings in the chemical industry.1 The present article will provide a broad review regarding the areas of opportunity where it would be possible to capitalize on the surplus chemical glycerol. Well known chemistry, as for instance the formation of esters or acetals, and especially catalytic routes to new derivatives, will be considered and discussed in detail. In the following, we will first examine the production of esters, ethers, acetals and ketals of glycerol. Subsequently, the conversion of glycerol to diols and epoxides will be treated, followed by a review of the different possibilities to oxidise and to dehydrate glycerol to aldehydes, ketones or carboxylic acids. We will conclude by discussing other feasible reactions, like for instance the conversion of glycerol to synthesis gas.
Barrault et al.2 describe a relatively selective synthesis of monoglycerides in a “one pot” reaction by converting unprotected glycerol with different methyl esters which can vary in their chain length between 11 up to 17 carbon atoms. A guanidine catalyst is used as a homogeneous catalyst or as an immobilised catalyst fixed on crosslinked polystyrene (Scheme 2). The transesterification experiments with stoichiometric amounts of glycerol and methyl esters were performed at a temperature of 110 °C and with a relatively high catalyst concentration of 12.5 mol%. To achieve high yields between 96% and 100% a reaction time of 8 h was necessary when using the heterogeneous catalyst. With the homogeneous guanidine catalyst the reaction time increases from 2 to 3.5 h. The selectivity to monoesters reaches 65%, the selectivity to diesters 34% and the selectivity to triesters at most 5%.
Scheme 2 Selective catalytic syntheses of monoglycerides (PS = crosslinked polystyrene). |
The results are completely different when a heterogeneous catalyst is used. Here the selectivity to monoesters decreases from 62% for shorter methyl esters to 47% for longer methyl esters. For di- and triesters the situation is reversed: With increasing chain length the selectivity also increases. This steering effect may be caused by the hydrophobicity of the esters which is rising with a longer alkyl chain. Due to the polystyrene chain the heterogeneous catalyst also has a high hydrophobicity, so that the interaction between the catalyst and the methyl ester is more intensive. In the end this leads to a higher selectivity to the di- and triesters.
The influence of solvents has also been investigated: The reaction rate is very low in solvents in which the methyl esters are only sparingly miscible. In solvents where good miscibility is achieved the reaction rate is as high as without a solvent.
Barrault et al. have also investigated basic oxides as possible catalysts for the synthesis of glycerides.3 All experiments have been performed at 220 °C, with an equimolar glycerol/methyl stearate mixture and a catalyst concentration of 2.7 wt.%. Experiments were performed with ZnO, MgO, CeO2 and La2O3 (Table 1).
When La2O3 is used as catalyst the formation of the unwanted by-product acrolein is observed. Small conversions of glycerol lead to a high selectivity towards monoglycerides whereas at higher glycerol conversions the selectivity decreases dramatically. The highest conversion of glycerol could be achieved applying the catalyst with the lowest surface area.
Chang and Wu investigated the lipase-catalysed synthesis of triglycerides of phenylalkanoic acids.4 These esters act as pharmaceuticals which convert slowly into acids and glycerol inside the human body leading to a long residence time and a high concentration level of the active ingredient. The esterification to triglycerides could be performed in yields of 70% when the lipase Novozym 435 was used at a temperature of 60 °C. A solvent-free system leads to the best yields of triglycerides. Non-polar solvents only give very small yields whereas with increasing solvent polarity the yield to the products increases, too. By using 4-phenylbutyric acid an optimum temperature of 65 °C was found whereas higher temperatures lead to lower yields presumably because of deactivation of the enzymes. The water removal is a further very important aspect: Via removing the water—especially by the use of molecular sieves—the equilibrium of the reaction can be shifted towards the products. The use of saturated salt solutions or the removal of the water by pervaporation was not as effective.
Pérez-Pariente et al. investigated the synthesis of monoglycerides by converting glycerol with fatty acids using functionalised mesoporous materials as catalysts.5 The esterification of glycerol with oleic acid was performed with an equimolar ratio of the starting compounds at a temperature of 100 °C, a catalyst concentration of 5 wt.% and a reaction time of 24 h. The influence of the catalyst synthesis procedure, the effects of hydrophobicity and organosulfonic functional groups and the impact of the catalyst structure were studied in detail. The catalyst synthesised by the so-called co-condensation method achieved a glycerol conversion of 24%, while the catalyst built by the post synthesis only led to conversions of 11%, each with a selectivity to the monoglycerides of about 69%. Hydrophobic catalysts lead to higher conversions of glycerol and to a higher selectivity of monoglycerides (75–95%). To create the acid sites of the catalyst two different sulfonation procedures were tested: The method with SO3 proved to be more effective than the method with chlorosulfonic acid, because a higher acid content could be obtained. Concerning the catalyst structure the MCM-41 catalyst proved to be the best one, better than SBA-15, SBA-12 and SBA-2. MCM-41 led to the highest selectivity to monoglycerides (77%).
Corma et al.6 investigated Lewis basic hydrotalcites as catalysts for the conversion of methyl oleate with glycerol to esters which are valuable as surfactants and emulsifiers. In detail they compared Lewis- and Brønsted-basic materials. At a reaction temperature of 200 °C the Lewis-basic hydrotalcite led only to a yield of monoglycerides of about 60% while solid Brønsted-basic catalysts reached 80%. Water has an important influence on the catalytic performance of the hydrotalcites, because higher water amounts can block the catalytically active sites on the surface.
The conversion of glycerol with acetic acid leads to the glycerol mono-, di- and triacetins, which can be purchased for instance from Cognis.7 Bremus et al.8 invented a process to produce the triacetin continuously: In a first step the glycerol is converted only partly with acetic acid. In a second unit this reaction mixture is mixed with acetic acid anhydride to achieve a nearly complete conversion to triacetin. As a final step the triacetin is separated and purified by distillation. The non-toxic triacetin can be used for instance as a textile auxiliary.9
Roice et al. published a paper about the use of glycerol dimethacrylate as monomer. This diester can be used for the synthesis of copolymers with new interesting properties.10
Scheme 3 Syntheses of glycerol carbonate. |
Glycerol carbonate and its esters are very interesting derivatives of glycerol, because they can be used as solvents for many applications, for instance in colours, varnishes, glues, cosmetics and pharmaceuticals.11 The use as extraction medium or as monomer for the synthesis of new types of polymers may also become of interest.
The direct synthesis of glycerol carbonate by the conversion of glycerol with carbon dioxide (route B) is therefore an interesting target of scientific research groups. Aresta et al.12 have investigated this reaction using di(n-butyl)tin dimethoxide, di(n-butyl)tin oxide and tin dimethoxide as catalysts. Under the investigated reaction conditions (50 bar CO2, 180 °C) tin dimethoxide led only to traces of glycerol carbonate. However, di(n-butyl)tin dimethoxide led to a conversion of glycerol of maximum 7% depending on the reaction conditions. No solvent was used and molecular sieves were added in order to remove water from the gas phase. Using di(n-butyl)tin oxide as catalyst conversions of glycerol up to 2% were reported.
Similar tin catalysts were investigated successfully by Ballivet–Tkatchenko in the direct synthesis of dimethyl carbonate working in supercritical carbon dioxide.13 The reaction of glycerol and carbon dioxide has still to be further optimised to become a technical route to glycerol carbonate.
The catalytic conversion of glycerol with urea to glycerol carbonate has been investigated by Yoo and Mouloungui.14Scheme 4 depicts the formal chemical equation.
Heterogeneous zinc catalysts were used to perform the reaction. At temperatures between 140–150 °C and a pressure of 40 mbar equimolar amounts of glycerol and urea react to glycerol carbonate. Table 2 shows the main results of the reaction optimisations.
Catalyst | Reaction time/h | Molar yield (%) |
---|---|---|
Zn(CH3C6H4-SO3)2 | 1 | 81 |
Zn(CH3C6H4-SO3)2 | 1.25 | 85 |
ZnSO4·H2O | 2 | 83 |
ZnSO4 (pre-treated for 3 h at 450 °C) | 2 | 86 |
In particular, the catalyst Zn(CH3C6H4-SO3)2 leads to yields of more than 80% towards glycerol carbonate in a relative short reaction time of about 1 h. Similar high yields can be achieved by other catalysts, but only at longer reaction times.
Glycerol carbonate can split off carbon dioxide yielding glycidol.15 For this reaction (Scheme 5) zeolithe A proved to be a very active catalyst. At a temperature of 180 °C and a pressure of 35 mbar a yield of 86% and a purity of 99% could be attained.
Scheme 5 Conversion of glycerol carbonate to glycidol. |
Rokicki et al. investigated the use of glycerol carbonate as monomer to produce hyperbranched aliphatic polyethers (polyglycerols).16
• The oligomers of glycerol in which the glycerol units are linked together via ether bondings to linear or cyclic compounds. When a higher number of glycerol molecules is linked together these compounds are also called polyglycerols. Both the oligoglycerols as well as their esters have already found important technical applications.
• The alkyl ethers of glycerol, for instance the glycerol tertiary butyl ethers.
• The alkenyl ethers of glycerol, for instance the telomers of glycerol with butadiene.
Scheme 6 Synthesis of oligoglycerols by intermolecular dehydration of glycerol units. |
The oligoglycerols are gaining more and more interest as products used in cosmetics, food-additives or lubricants.17 Short overviews about the synthesis of glycerol oligomers from di- to pentaglycerol have been published by Rollin et al.18 and Pérez-Pariente et al.19 Generally oligoglycerols are produced using basic homogeneous catalysis, but lately increased attention has been paid towards heterogeneously catalysed processes.20 Despite a lower activity heterogeneous catalysts reveal many advantages: Firstly, the separation of the catalyst does not cause any problems because the catalyst is a solid which is insoluble in the reaction medium. Secondly, a higher selectivity has been observed when heterogeneous catalysts such as zeolites, ion exchangers or mesoporous molecular sieves of the type MCM-41 are used.17,21,22 However, it should be emphasised that so far the conversion of glycerol using heterogeneous catalysts is by far lower than using homogeneous catalysts.
A typical comparison is given in Table 3: The homogeneously catalysed oligomerisation of glycerol with the catalyst Na2CO3 leads to a conversion of 80% after a reaction time of 8 h, while with the heterogeneous catalyst Cs-MCM-41 the conversion reached 95% but only after 24 h. Both reactions were carried out at 260 °C under ambient pressure. The amount of catalyst used was 2 wt.%. The homogeneous catalyst yields a very broad product distribution from 31% diglycerol to 24% higher oligoglycerols. In contrast most heterogeneous catalysts yield relatively high amounts of diglycerol, for instance 59% with Cs-MCM-41, and only very low amounts of higher oligomers.17 Ion exchangers, for instance Amberlyst 16 and Amberlyst 31, are also very selective catalysts for the specific formation of the short chain oligomers and only a low reaction temperature of 140 °C is needed. The amount of catalyst used in these reactions was in this case 5 wt.%. However, if these acidic catalysts are used, significant amounts of acrolein are formed.
Catalyst | Na2CO3 | Cs–MCM–41 | Amberlyst 16 | Amberlyst 31 |
---|---|---|---|---|
Conversion (%) | 80 (8 h) | 95 (24 h) | 35–40 | 35–40 |
Selectivity (%) | ||||
Diglycerol | 31 | 59 | 85 | 75 |
Triglycerol | 28 | 30 | 15 | 25 |
Tetraglycerol | 17 | 11 | <1 | <1 |
Higher oligomers | 24 | <1 | <1 | <1 |
Moreover homogeneous catalysts lead to a different product distribution in contrast to heterogeneous catalysts: Homogeneous catalysts produce mostly linear oligomers whereas heterogeneous catalysts result predominantly in branched oligomers.17
Akzo Nobel reported a saponite catalyst for the synthesis of glycerol oligomers.23 The catalyst is a Mg-rich clay which leads to a glycerol conversion of 56% after 40 h of reaction. In this case the selectivity to diglycerol is 30%.
Important derivatives of the oligoglycerols are the oligoglycerol esters. Starting from the oligomers described above, monoesters can be produced via esterification with carboxylic acids or transesterification with carboxylic acid methyl esters (Scheme 7). Di- or triesters of glycerol oligomers can be synthesised in a comparable way.
Scheme 7 Synthesis of monoesters of diglycerol. |
Plasman et al. discuss the applicability of these polyglycerol esters,24 especially as antifogging and antistatic agents or as lubricants in the food industry. Only 0.5% of diglycerol monooleate are sufficient to achieve an antifogging effect in food packages. Polyglycerol esters are biodegradable and exhibit a high thermal stability as well as a colourless appearance even after thermal treatment in air at 250 °C.
Barrault et al. compared MgO and Cs-MCM-41 as heterogeneous catalysts for the selective synthesis of glycerol esters and diglycerol esters.17Table 4 shows the results of the conversion of glycerol with methyl dodecanoate at 240 °C. At these conditions the glycerol is esterified to its mono-, di- and triesters and oligomerised to diglycerol followed by further esterification. Whereas MgO leads to a high selectivity of glycerol esters (∑ 89%) the mesoporous catalyst Cs-MCM-41 leads to higher amounts of diglycerol esters (∑ 54%). The concentration of catalyst was 14 wt.% referring to glycerol and 6 wt.%. referring to the ester.
Hao described polyglycerol esters which are used as stabilisation agents in water-in-gasoline microemulsions.25 The addition of small amounts of water to gasoline reduces both NOx emissions as well as soot formation. The water can be dissolved in gasoline by adding a mixture of a polyglycerol ester, a surfactant, and an alcohol thus yielding a microemulsion. The best system proved to be diglycerol diisostearate combined with sodium oleate in a ratio of 6 : 4, dissolved in 1-butanol. The highest water content in the microemulsion could be reached at a temperature of 16 °C.
Jérôme et al.26 reported on a phosphazene catalyst for the synthesis of the mono- and diesters of glycerol derivatives. Using this catalyst, regioselective syntheses can be performed.
The last group of glycerol ethers to be presented are the hyperbranched polyglycerols. Scheme 8 depicts the general molecular structure of these compounds.
Scheme 8 Hyperbranched polyglycerols. |
However, until now hyperbranched polyglycerols cannot be synthesised directly from glycerol as feedstock. The synthesis of such molecules, for instance via reaction of trimethylolpropane and glycidol, has been described in detail by Mülhaupt et al.27,28 Mecking et al. reported the incorporation of Pd-nanoclusters into polyglycerols to yield catalysts which are active in the hydrogenation of cyclohexene.29 These palladium/polyglycerol catalysts can easily be recycled and show a constant activity. Haag et al. presented polyglycerols able to solubilise hydrophobic drugs.30
The structure of the polyglycerols and polyglycerol esters were investigated by using small-angle neutron scattering (SANS).31
Further investigations have been performed by Dworak et al.32 They investigated the polymerisation of glycidol using Lewis and protonic acids. Moreover, they elucidated the structure of the polymer by NMR-measurements and suggested a reaction mechanism.
In 2001 a company was founded which deals with hyperbranched polymers commercially.33
Glycidol has also been mentioned for the synthesis of modified epoxy resins. Those resins are produced by Evonik.34
Furthermore, hyperbranched polymers have been described which have been obtained from glycerol carbonate, which has been synthesised previously from glycerol and dimethyl carbonate.16
Glycerol as a building block for polymers has also been previously described. The Konkuk University in Seoul has made a patent application in which glycerol and fatty acids are essentials for polyurethaneamide adhesives.35
• The Williamson synthesis of sodium-glycerolate with an alkyl halide.
• The addition of an active alkene, especially a branched alkene, to glycerol.
• The condensation reaction between glycerol and an aliphatic alcohol under elimination of water.
Aubry et al. reported on the synthesis of glycerol 1-monoethers starting from 1,2-isopropylidene glycerol, potassium hydroxide and different bromoalkanes.36 As these monoethers have the characteristics of solvents as well as of surfactants they are entitled “green solvo-surfactants”, which should be considered for the replacement of ethylene or propylene glycol ethers. However, their synthesis only proceeds with time-consuming protecting group chemistry and stoichiometric amounts of alkali bromides are formed as by-products.
The two other routes of ether synthesis will be demonstrated with the example of the glycerol tertiary-butyl ethers, the GTBE. They can be synthesised via the reaction of glycerol with isobutene or with tertiary butanol (Scheme 9).
Scheme 9 Synthesis of GTBEvia reaction of glycerol with isobutene or tertiary-butanol (m-GTBE = monoethers; h-GTBE = higher ethers). |
The conversion of glycerol with isobutene yields the mono-, di- and tri-tertiary butyl ethers of glycerol. Due to the two remaining hydroxy groups the monoethers (m-GTBE) are still soluble in polar solvents, whereas a mixture of the di- and triethers, the so-called “higher ethers” (h-GTBE), are soluble in unpolar media, for instance in hydrocarbons. This property enables the application of the higher GTBE as fuel additives. Because of the glycerol content this additive can be assigned to the renewables and may help to meet the biocomponent target of the European Union. h-GTBE can be used as a diesel fuel additive, which reduces remarkably the emissions of particulate matter.37,38 However, because of the branched alkyl substituents the higher GTBE can also be used as potential components for the use in gasoline (“octane-boosters”) substituting the methyl tertiary butyl ether MTBE which is under discussion because of some environmental disadvantages.39,40
The catalytic reaction of glycerol with isobutene was first investigated in 1992 by Behr and Lohr at the Henkel Company.41 They discovered some very active homogeneous and heterogeneous catalysts, for instance p-toluene sulfonic acid, acidic ion exchangers like Amberlyst 15 and several synthetic zeolites. In 1996 Behr and Obendorf42,43 continued these investigations and developed a technical process for the production of the higher GTBE. With p-toluene sulfonic acid (p-TSA) as the best catalyst the etherification of glycerol with isobutene was performed in a stirred batch reactor at 90 °C with a molar ratio of isobutene to glycerol of 2 : 1. At the beginning of the reaction the system consists of two immiscible phases, the polar glycerol and the unpolar isobutene phase. As the reaction proceeds, the mono-, di- and triethers are formed and after 40 min the reaction system becomes single-phased. After a total reaction time of about 3 h the reaction equilibrium is achieved yielding a mixture of about 45 wt.% diethers, 30 wt.% monoethers and 5 wt.% of unconverted glycerol.
Based on these laboratory investigations a continuous production process of the higher ethers was designed (Fig. 1), which enables the complete recycling of the homogeneous catalyst p-TSA and of the unwanted monoethers. The process consists of a reactor cascade operated at 20–30 bar followed by an extraction at the same pressure conditions. Isobutene is fed directly to the reactor, whereas glycerol is fed to the extractor separating the catalyst and the monoethers from the reaction mixture. The polar glycerol/monoether-phase (R1) is run back to the reactor, whereas the unpolar phase is conducted to a flasher were the unreacted isobutene is stripped and than also recycled to the reaction unit. The bottom product of the flash passes a vacuum rectification column obtaining pure higher ethers overhead. The bottom product of the rectification which contains some traces of m-GTBE and the catalyst p-TSA is passed back to the reactor (R2). This process seems to be more economic than other processes proposed in some patents.44–50
Fig. 1 Flow scheme for the technical process for the production of higher GTBE (R1 = recycle of glycerol, p-TSA and m-GTBE; R2 = recycle of traces of m-GTBE and p-TSA). |
Karinen and Krause have investigated the same reaction system using Amberlyst 35WET as a heterogeneous catalyst.51 At a ratio of isobutene and glycerol of 2 : 1 the highest selectivity towards ethers of about 92% can be reached whereas a molar ratio of 4 : 1 leads to a selectivity of about 80%. 80 °C is described to be the best reaction temperature. High ratios of isobutene to glycerol lead to small fractions of monoethers, whereas small ratios lead to large amounts of monoethers, which, however, cannot be used as octane-boosters.43
In 2003 Klepáčová et al. investigated the reaction of glycerol with tert-butanol in the presence of heterogeneous catalysts52 and compared Amberlyst 15 with the zeolites H-Y and H-Beta. By using Amberlyst 15 as catalyst the highest glycerol conversion is achieved at 90 °C after 3 h. A catalyst concentration between 5 and 8 wt.% and an equimolar ratio between tert-butanol and glycerol gives optimum results.
In 2005 Klepáčová et al. published new investigations53 where they compared experiments with isobutene as well as with tert-butanol. If the same temperature and catalyst is used isobutene always leads to higher conversions of glycerol than tert-butanol. This can be explained by the formation of the by-product water, which deactivates the catalyst. At 60 °C with the catalyst Amberlyst 35 in the dry form the conversion of glycerol with isobutene is almost quantitatively, whereas the conversion of glycerol with tert-butanol is only 86%.
Like other ethers of glycerol they can be used as emulsifiers or surfactants due to their amphiphilic nature after hydrogenation of the remaining double bonds. Moreover, glycerol monoethers in particular have some special advantages regarding e.g. toxicity and biodegradability.57
The telomerisation is catalysed homogeneously at temperatures around 80 °C by metal complexes of palladium or nickel containing suitable ligands, e.g. phosphines, leading to telomer selectivities > 95%.55,56,58,59 Because of the homogeneous catalyst, recycling of metal and ligands is necessary to make the reaction efficient. Some recycling concepts are already described in the literature.58,60
They can be added up to 10 vol% of the fuel used61,62 and one main advantage is—comparable to the GTBE—the reduction of particle emissions.63 A second application of glycerol acetals is the use as scent or flavour, e.g. the acetalisation of phenylacetaldehyd or vanillin with glycerol leading to hyacinth or vanilla flavours.64,65Glycerol acetals can also be used as the basis for surfactants.66
Acetals or ketals can be formed by reaction of glycerol with various aldehydes or ketones yielding five- or six-membered cycles (Scheme 11).
In the acetalisation reaction, the five- and six-ring compounds are usually formed in a ratio of 50 : 50. Some efforts have been made to change and control this ratio by varying the reaction parameters such as temperature, aldehyde or ketone to glycerol ratio and the choice of solvent. One example is the reaction of glycerol with acetone in dichloromethane at about 40 °C leading to a ratio of the five-membered to the six-membered ring as 99 : 1. In the case where the reaction is carried out with formaldehyde under the same conditions only a ratio of 22 : 78 is found.
To perform an acetalisation, a catalyst is needed, which is mostly a solid acid catalyst. For example Amberlyst 15 catalyses the acetalisation of glycerol with various aldehydes or ketones e.g.butanal61 or acrolein67 at temperatures below 70 °C leading to high selectivities. Also Amberlyst 36 is active in the acetalisation of glycerol, e.g. with formaldehyde at T = 50–110 °C with a yield of 77%. The glycerol formal can be used as disinfectant or solvent for medical or cosmetic applications.68,69 Furthermore, various zeolites catalyse the acetalisation reaction.64 To modify the solubility and other physical properties, glycerol acetals can be varied by a further reaction of the remaining hydroxy group, which is shown in an example in Scheme 12.70
Scheme 12 Modification of the glycerol formal by etherification of the hydroxy group. |
Another standard way to change the polarity of glycerol acetals is the etherification of the remaining hydroxy group with tertiary alkenes.71
A further way to change the properties of given acetals or ketals is the transacetalisation by which e.g. long chained molecules (Scheme 13) can be obtained.72,73 Overall very polar (Scheme 12) as well as very unpolar acetals and ketals (Scheme 13) can be produced.
Scheme 13 Acetalisation of glycerol with octadecanal to 2-heptadecyl-4-hydroxymethyl-1,3-dioxolan. |
While 1,3-propanediol is mainly used as a starting material for polymers, 1,2-propanediol is often used without further modification e.g. as an additive in nutrition products, solvent for colourings and flavours, wetting agent in tobaccos, additive in cosmetics, as component of break or hydraulic fluids, lubricants or anti-freezing agent. Furthermore, it can be used as a starting material for solvents, emulsifiers and plasticisers.
Scheme 14 Comparison of the reaction routes to 1,2-propanediol starting from propene or glycerol. |
In particular, when 1,2-propanediol or its derivatives are applied in food, cosmetics or pharmaceutical products, the use of fossil raw materials is less favourable to the consumer acceptance, which leads to the demand of a renewable feedstock such as glycerol or sugars. Agribusiness companies such as Archer Daniels Midland (ADM) have developed processes based on these starting materials.76Glycerol can be converted to 1,2-propanediol using heterogeneous, homogeneous or biocatalysts, which are described in the following section.
Another approach is the use of Raney-nickel as hydrogenation catalyst. As in the case of Ru as a catalyst, reaction conditions are relatively mild.85 The use of rhenium as catalyst metal in a multimetallic catalyst leads to conversions up to 80% with selectivities > 30% at temperatures > 120 °C and 80 bar hydrogen pressure.86 Besides the influence of metal and support also the choice of solvent has a strong influence on the performance of the catalyst system, especially on the selectivity towards 1,2- or 1,3-propanediol.87
Not only the catalyst systems and solvents have been improved, but also completely new processes have been developed. One example will be described in section 5.2a.
Scheme 15 Different routes to 1,3-propanediol starting from ethene, propene or glycerol. |
Problems in the conventional processes are the high pressure applied in the hydroformylation and hydrogenation steps as well as the use of aromatic solvents in the first and loss of acrolein due to extraction processes in the second example. The yields are around 80% in the first and about 40% in the second process, so besides the demand of renewable sources like glycerol, there is also a huge interest in improving yields and overall selectivity of the processes applied. Therefore, the reaction from glycerol to 1,3-propanediolvia heterogeneous, homogeneous or biocatalytic processes may become an attractive alternative.
Scheme 16 Stepwise conversion of glycerol to propanediols. |
Scheme 17 Alternative process from glycerol to 1,3-propanediol using protection groups. |
Besides glycerol, other polyols, like starch or glucose, can be used in the production of 1,3-propanediol by fermentation with the bacteria mentioned above. One example of a technology switch from petrochemicals to renewable sources is the fibre SORONA made by DuPont. Genencor International and DuPont have developed in conjunction a process for producing 1,3-propanediol by fermentation with recombinant Escherichia coli. The genome with the information about the conversion of glycerol to 1,3-propanediol has been isolated from a donor organism and transferred to E. coli. This modified organism is 500 times more effective than the original one.114,116–120 The process can convert glycerol but was afterwards switched to the cheaper glucose, now a fundamental technique for DuPont to produce 1,3-propanediol. In the first plant in Loudon, Tennessee, 500 t year–1 of 1,3-propanediol can be produced starting from glucose obtained from corn.121–124 According to DuPont the process saves 40% of the energy consumed by its petrochemical counterpart.
Scheme 18 Conversion of glycerol to epichlorohydrine. |
Another epoxide with the glycerol structure is the glycidol whose synthesis has already been described in Scheme 5 in section 2.2.
In the following section (section 7.1) some valuable glycerol oxidation products are presented together with their main application fields. Secondly, the most important catalytic oxidation processes of glycerol are discussed (section 7.2). In the last section (section 7.3) the dehydration of glycerol to acrolein is presented.
The oxidation of the primary hydroxy groups of glycerol produces glyceraldehyde, an intermediate in the carbohydrate metabolism. Glyceraldehyde is also used as a standard by which chiral molecules of the D- or L-series are compared. Further oxidation of glyceraldehyde produces carboxylic acids, like glyceric acid, tartronic acid and mesoxalic acid (Scheme 19). These acids are mainly converted into various market products, e.g.polymers or biodegradable emulsifiers. In addition to these applications, glyceric acid can be used for treatment of skin disorders,132,133 or as an anionic monomer of packaging material for exothermic and volatile agents.134 In its ester form, glyceric acid can act together with a quaternary ammonium salt as an effective and biodegradable fabric softener.135Tartronic acid can be used as a potentiating agent or adjuvant to increase the blood absorption of a tetracycline antibiotic.136 Moreover, it can be used to scavenge dissolved oxygen in alkaline water.137
Platinum metal catalysts are more selective for the oxidation of primary than for secondary alcohols. The main product of glycerol oxidation using platinum supported on active charcoal is glyceric acid (selectivity 55% at 90% conversion) with a small amount of DHA (selectivity 12%).139 The selectivity towards the secondary alcohol can be increased by association of the platinum with p-electron metals, especially lead and bismuth.140,141 These metal promoters can be added to the platinum-metals by co-impregnation, impregnation or redox surface reaction.142 In batch reactions the aerobic glycerol oxidation with Pt–Bi catalysts, prepared by means of the co-impregnation method, has a selectivity of 37% towards DHA at a 75% conversion of glycerol.139 A still better result of the Pt–Bi catalysts could be achieved in catalyticglycerol oxidation by using a fixed-bed reactor, producing DHA with 80% selectivity at 40% conversion).143,144
The utilizations of palladium catalysts for catalyticglycerol oxidations have been initiated by the group of Hutchings and the group of Kimura.145,146Palladium catalysts at basic conditions have a high selectivity towards glyceric acid (77% at 90% conversion), whereas the selectivities to the other reaction products (DHA, tartronic and oxalic acids) are always lower than 10%.139 However, palladium shows a weaker catalytic activity compared to platinum for the oxidation of the secondary hydroxy group of tartronic acid to produce mesoxalic acid.147
Despite its successful application in aerobic glycerol oxidation processes, platinum suffers strongly from oxygen poisoning at high oxygen partial pressures. Therefore, this system can only be utilised at low oxygen partial pressures or by using air at atmospheric pressure, in order to limit the oxygen dissolution.148 On the other hand, gold has a better resistance to oxygen poisoning compared to platinum.149Catalyticglycerol oxidation using gold catalysts have demonstrated a high selectivity to glyceric acid (more than 90% selectivity at 90% conversion), depending on the catalysts preparation methods used.148
To evaluate the effect of preparation methods on glycerol oxidation, gold catalysts were prepared by different methods, e.g. by sol–gel, incipient wetness and impregnation. In the sol–gel method, the effect of protection agents, for instance polyvinyl alcohol (PVA), sodium citrate and tetrakis(hydroxypropyl)phosphonium chloride (THPC) has been investigated, and it was shown that the gold catalysts prepared by the sol–gel method have a higher activity than the catalysts prepared by the two other methods.148,150
The increase in activity can also be caused by smaller catalyst particles. For gold it has been shown that the activity is highly dependent upon the particle size.150,151 If THPC or PVA are used to stabilise the colloid gold solutions, particles smaller than 10 nm were formed.148,150 However, if glyceric acid is the desired main product, the sol–gel method (Table 5, entries 3 and 4) is not the suitable method, since the selectivities are lower compared to the impregnation and incipient wetness methods (Table 5, entries 1 and 2).148
Entry | Preparation method | Glyceric acid selectivity at | |
---|---|---|---|
50% conversion | 90% conversion | ||
1 | Impregnation | 71 | 70 |
2 | Incipient wetness | 80 | 78 |
3 | Sol–gel (PVA) | 47 | 35 |
4 | Sol–gel (THPC) | 52 | 45 |
5 | Sol–gel (sodium citrate) | 75 | 75 |
Basic reaction conditions are essential for the gold catalysis, since the initial step of the oxidation and the dehydrogenation of glycerol cannot proceed in the absence of a base. In the presence of a base, the H+ is readily abstracted from one of the primary hydroxy groups to form glyceric acid.145,152
To improve the rate and selectivity of glycerol oxidation processes bimetallic catalysts were used. The comparisons between monometallic Au, Pd, or Pt and bimetallic Au–Pd153 or Au–Pt154,155catalysts proved that in many cases bimetallic catalysts have higher activity and resistance to deactivation (Table 6).141,142,153
Catalyst | Conversion (%) | Selectivity (%) | TOF/h–1a | |||
---|---|---|---|---|---|---|
Glyceric acid | Glycolic acid | Oxalic acid | Tartronic acid | |||
a Calculation of TOF/h–1 after 0.25 h of reaction. TOF members were calculated on the basis of total loading of metals. b ns = not specified. | ||||||
Pt/C | 50 | 42 | 31 | 8 | 6 | 532 |
Pd/C | 50 | 81 | 3 | 0 | 14 | 1151 |
100 | 21 | 27 | 12 | 39 | ns | |
Au/C | 50 | 65 | 12 | 10 | 9 | 1090 |
100 | 45 | 24 | 19 | 10 | ns | |
Au–Pt/C | 50 | 72 | 18 | 1 | 8 | 1987 |
100 | 31 | 35 | 5 | 28 | ns | |
Au–Pd/C | 50 | 77 | 5 | 0 | 18 | 1775 |
100 | 49 | 25 | 2 | 25 | ns |
Monometallic Pd/C catalysts exhibit the highest activity with 50 °C, compared to the other metals. Under these conditions, Pd/C shows a better performance than Pt/C which cannot even reach 100% conversion because of the oxygen poisoning. However, if Pt is combined with Au, this bimetallic catalyst has the highest activity compared to all other catalyst systems. The bimetallic Au–Pd catalyst also provides an increase in activity.
The alloyed phase of the bimetallic Au–Pd system is much more active than the same system with segregated metals. The key to avoiding palladium segregation is by slowing down the reduction rate of the palladium salt. This can be achieved via two consecutive preparation steps: First the gold sol is immobilised on carbon using sodium boranate as reducing agent and then the palladium sol is precipitated on the Au/C surface using hydrogen as reducing agent. Applying NaBH4 in the second step would result the unwanted palladium segregation.156
One example of a three-metal catalyst in glycerol oxidation is Pd–Bi–Ce supported on charcoal. This catalyst has a multifunctional role, i.e. to oxidise glycerol to tartronic and mesoxalic acid and then to carry out the polymerisation of the mesoxalic acid into poly(ketomalonate) (Scheme 20).147 Furthermore, in the case of excessive oxidation, a decarboxylation process takes time and results in the conversion of poly(ketomalonate) to poly(oxymethylene) (Scheme 21).157 The poly(oxymethylene) is an engineering plastic which is also known as polyacetal, acetal resin, polytrioxane, polyformaldehyde, and paraformaldehyde.
Scheme 20 Structure of poly(ketomalonate). |
Scheme 21 Structure of poly(oxymethylene). |
Scheme 22 Production of acrolein and acrylic acid based on propene or glycerol. |
Additional to this petrochemical route, acrolein can also be obtained from glycerol by dehydration in the liquid or in the gas phase on heterogeneous catalysts. Typical catalysts are for instance zeolites, Nafion composites, aluminas, phosphor- or silicotungstic acids or acid salts which should all have an Hammet acidity of less than + 2 leading to a yield of greater 70% at T = 250–350 °C.164–166 A further approach is the direct reaction to acrylic acid, starting from glycerol to avoid separation problems (Scheme 22).167
Heterogeneous catalysis is not the only appropriate route for the dehydration of glycerol, but also homogeneous systems or even systems without catalyst can lead to acrolein. One approach is the reaction in near supercritical or supercritical water (T > 300 °C, p > 300 bar), using either no catalyst or different additives (Table 7).
Catalyst/additive | Yield (%) | Literature |
---|---|---|
None | 12 | 168 |
Sulfuric acid (1–5 mmol l–1) | 74 | 169,170 |
Zinc sulfate (0–790 ppm) | >40 | 171,172 |
Besides using chemical pathways, acrolein can also be produced by biocatalysis. Glycerol can be fermented to 3-hydroxypropanale.g. by Lactobacillusreuteri and subsequent thermal dehydration leads to acrolein.108 The catalysts and reaction systems used are the same as in the production of 1,3-propanediol with 3-hydroxypropanal as an intermediate step (section 5.2).
Scheme 23 Production of synthesis gas from triglycerides viaglycerol. |
The synthesis gas can also be utilised in Fischer–Tropsch reactions with the great advantage that the CO/H2-mixture does not contain any oxygen impurities, making the reaction much easier to handle.
There are several processes to prepare synthesis gas from glycerol. These alternatives, the steam reforming, the aqueous phase reforming and the gasification in supercritical water, are discussed in the following sections.
Scheme 24 Main reactions in the steam reforming of glycerol. |
Typical catalysts are platinum, ruthenium or rhenium applied on a carrier like carbon, yttrium oxide or ceria washcoat. Some characteristic results are listed in Table 8. Platinum is a very effective steam reforming metal because it is already active at very mild conditions (400 °C) and yields great amounts of carbon monoxide.
Temperature/°C | Catalyst | S/Ca | C/Ob | Y H2 (%)c | S CO (%)d | Lit. |
---|---|---|---|---|---|---|
a Steam to carbon ratio. b Carbon to oxygen ratio in feed. c Yield H2 according to eqn (3). d Selectivity from carbon to CO. e Rhodium–ceria washcoat. All experiments were carried out at atmospheric pressure. | ||||||
1055 | RhCeWc e | 0 | 1.2 | 56 | 79 | 174 |
825 | RhCeWc e | 2 | 1.2 | 75 | 58 | 174 |
862 | RhCeWc e | 4.5 | 1.2 | 79 | 27 | 174 |
400 | Pt/C | 2.3 | 1 | — | 80 | 175 |
500–600 | Ru/Y2O3 | 3.3 | 1 | 82 | 20 | 173 |
The aqueous phase reforming metals such as platinum, palladium or Ni–Sn alloys show high selectivity for hydrogen production, whereas supported nickel catalysts tend to deactivation with time, which may be due to sintering of the metal particles. Also the support plays an important role in the selectivity of the APR process: Basic or neutral supports, e.g. alumina, favour the formation of hydrogen and carbon dioxide. Since the separation of H2 and CO2 is relatively easy, aqueous phase reforming can be used to produce pure hydrogen.176
The conversion of glycerol increases with temperatures from 20% (at 450 °C) up to 90% (at 700 °C); a further increase does not lead to higher conversions. The composition of the gas slightly changes with the temperature: at higher temperatures a larger amount of hydrogen can be obtained. Experimental results from capillary reactors showed that the pressure has no influence on the carbon efficiency and the product gas yield.177
Scheme 25 Compilation of the main glycerol derivatisation routes. |
• An important aim is the selective synthesis of glycerol monoesters with both homogeneous and heterogeneous catalysts. The direct catalytic reaction of glycerol and carbon dioxide to glycerol carbonate still remains a great scientific challenge.
• The formation of glycerol ethers has been carried out via different routes using alkyl halides, alkenes and dienes as reaction partners. The application of the ethers as fuel additives would consume a great amount of the glycerol surplus formed in the biodiesel production.
• Glycerol acetals and ketals can also be considered as potential fuel additives. A great advantage of these derivatives is the broad range of polarity which can be tuned in a few reaction steps.
• The production of the two propanediols on the basis of glycerol has already been started. In particular, 1,3-propanediol seems to have become a new platform chemical in polyester chemistry.
• The two-step reaction of glycerol to epichlorohydrine, which was developed and implemented by Solvay, is a typical example for the partial change from petrochemical to renewable starting materials.
• The oxidation and dehydration chemistry of glycerol is still in its infancy and further investigations are needed to find better catalysts and reaction performance. However, the possible target molecules of glycerol oxidation may become interesting building blocks in fine and polymer chemistry.
• Finally, the conversion from glycerol to synthesis gas or pure hydrogen, respectively, has to be mentioned which may become an alternative to the traditional synthesis routes based on raw fossil materials.
This journal is © The Royal Society of Chemistry 2008 |