Bio-based chemicals from biorefining: protein conversion and utilisation
E.L. Scott, M.E. Bruins and J.P.M. Sanders, Wageningen University, The Netherlands
Abstract:
The depletion of fossil feedstocks, fluctuating oil prices and the ecological problems associated with CO2 emissions are forcing the development of alternative resources for energy, transportation fuels and chemicals: the replacement of fossil resources with biomass. The conversion of crude oil products utilises hydrocarbons and conversion to (functional) chemicals with the aid of co-reagents, such as ammonia, and various process steps. Conversely, proteins and amino acids, found in biomass, contain functionality. It is therefore attractive to exploit this to reduce the use, and preparation of, co-reagents as well as eliminating various process steps. This chapter describes how biorefineries can add value to protein containing rest streams by using amino acids as economically and ecologically interesting feedstocks and that, by taking advantage of the chemical structure in biomass rest streams, a more efficient application can be developed other than solely utilising it for the production of fuels or electricity.
Key words
amino acids; protein; rest streams; separation
22.1 Introduction
Due to ecological concerns regarding the use of fossil resources and greenhouse gas emissions, as well as issues relating to energy security, fluctuating oil prices and the volume oil reserves, governments around the globe are looking for alternative methods to produce bulk energy (heat and electricity), transportation fuels and chemicals using alternative resources. For the production of energy there are a number of options being explored and implemented such as the use of wind, solar and hydro/geo-thermal generation. In these cases, energy can be generated without the use of carbon-containing resources. However, in the case of mass produced transportation fuels and chemicals, which rely on carbon (and other) functionality, the alternatives for production require the use of an alternative carbon source to oil. In this case, the use of biomass is the only method for their production.
Why should amino acids be considered as interesting chemical feedstocks in the first place? The petrochemical industry utilises hydrocarbon feedstocks, e.g. ethylene and propylene, together with co-reagents and conversion steps to incorporate functionality to produce a large variety of chemicals and materials. The production of ethylene and propylene from biomass has been widely reported in the literature. Generally this involves fermentation, isolation of the desired molecule and subsequent (chemical) conversion to the desired hydrocarbon. Using this approach, (partially) bio-based alternatives to fossil-derived chemicals could be generated. However, it still requires the use of reagents, conversion processes and energy (often supplied in the form of natural gas). Thus significant fossil resources are still required to produce more functionalised products. If the overall aim is to develop processes to produce industrial products from biomass which reduce fossil input, then also the process for the conversion should be considered in the whole approach and not just the feedstock.
Amino acids contain functionalities, such as –NH2 and –COOH, which are similar to an array of functionalised industrial chemicals. In addition to this, they can be readily converted to other functionalities such as nitriles. Therefore it is attractive to exploit this in order to bypass the use and preparation of co-reagents such as ammonia, as well as eliminating various process steps. Thus potentially greater savings on the fossil energy required may be achieved (Scott et al., 2007). Allied to this, it is already known that lower raw material and capital investment costs are incurred in the production of non-funtionalised chemicals compared to more functionalised compounds (Lange, 2001). Thus more efficient use of the functionality of amino acids may be beneficial in the production of functionalised chemicals.
22.2 Protein and amino acid sources derived from biofuel production
During the production of bioethanol and biodiesel, rest streams are generated. In the case of biodiesel, press-cake and glycerol are formed. The use of the glycerol side stream has been successfully developed to produce traditional chemicals. Solvay produces epichlorohydrin from glycerol using the Epicerol™ process (Solvay Chemicals, 2006). Other bulk chemical products that are being developed from glycerol rest streams include conversion to methanol. The consortium Bio-Methanol Chemie Nederland (BMCN) uses glycerol to produce synthesis gas which is then converted to methanol (Chemie Magazine, 2006). As well as this, Archer Daniels Midland (ADM) are exploring the use of carbohydrates and/or glycerol as a feedstock for the production of propylene and ethylene glycols (Chemical Week, 2005). The press-cake rest stream formed has received less attention in its application. In general, press-cakes are utilised as animal feed due to the significant protein content (as well as containing carbohydrates and lignin). However, not all amino acids in the protein are required for animal nutrition. It is conceivable that non-essential amino acids are removed (after isolation and hydrolysis of the protein) and used for other applications. A similar argument can be used for dried distillers grains and solubles (DDGS), the rest stream of bioethanol production. Wheat DDGS contains 40% crude protein (Nuez Ortín and Yu, 2009). It is considered that cost-effective isolation of amino acids from wheat DDGS will improve the economic value of the bioethanol by-product and will aid the economic viability of the whole bioethanol production chain (Bals et al., 2009). At present, DDGS is recognised as nutritious animal feed, but due to its protein content, it would also serve as a raw material for amino acid production. In this case, certain amino acids may become an attractive option as feedstocks for the production of chemicals. Gluten, a major protein of wheat, is high in glutamic acid that is an excellent starting material for the synthesis of bulk chemicals via γ-aminobutyric acid (GABA) (Lammens et al., 2009). Protein extraction from DDGS is not yet employed successfully at any scale.
European guidelines aim for 10% of all transportation fuels to be derived from biomass by 2020 (EU, 2009). This will therefore result in rest streams containing proteins (20–40 wt% of the dry matter). Combined with the estimated worldwide biofuel production (IEA, 2011), this will lead to a production of ca. 100 million tons of protein per year corresponding to ca. 5 wt% (ca. 5 million tons) of each amino acid. By 2050, the IEA predicts an even higher production of biofuels and therefore substantial generation of protein-containing rest streams. Thus considerable amounts of each individual amino acid could be available as a feedstock for the production of bulk chemicals without competing with food and feed markets.
22.2.1 Using rest streams derived from biofuel production
Gasoline alternatives/replacements have also been developed from starch and sugar sources. Particularly in South America, the use of sugarcane to produce ethanol has been exploited as has the use of carbohydrate from corn and grain (US and EU). The production of so-called first generation bioethanol has attracted a lot of attention arising from issues surrounding ‘food versus fuel’. The reports ‘World agriculture: towards 2015/2030’ (FAO, 2003) and ‘International Assessment of Agricultural Science and Technology for Development’ (UNEP, 2008), describe sufficient food (production) for a growing global population and that new agricultural technology increases food production, but that in developing countries many will remain hungry and environmental issues attributed to agriculture still exist. The World Bank and the International Monetary Fund (IMF) have also commented on some of the many causes of rising food prices, including bad weather, high oil prices, increased demand for meat and dairy products in some Asian countries as well as the ambitions in the West to use biofuels derived from grain to reduce oil consumption.
This has led to developments to produce second generation fuels focusing on using potentially abundant and inexpensive lignocellulosic agricultural waste streams from primary agricultural production as a feedstock which is not food competitive. The use of biorefineries, where a crop may be separated into various fractions, for food and other applications such as animal feed and raw materials for biofuels and chemicals will play a key role. As well as this, other technical challenges include obtaining fermentable sugars from (hemi-)cellulose and the conversion of other sugars, e.g. xylose to ethanol.
Rudolph Diesel demonstrated the diesel engine using peanut oil as a fuel at the World Exhibition of 1900. Today, the production of biodiesel relies on the use of crops such as soya, rape and palm which are pressed (or extracted), releasing oils (and generating a rest press-cake) which are then transesterified yielding fatty acid esters (usually methyl) and glycerol. Debate as to land use and destruction of the rainforest is a particular issue. For example, the growth of soya in South America has raised concern about the destruction of the rainforest for its production.
Rest streams from biofuel production could lead to opportunities to obtain large volumes of protein. Theoretically, the volumes would allow sufficient amino acids to be available as feedstocks for bulk functionalised chemicals. However, one should also take into account the complex issue regarding ethical, social and ecological issues surrounding biomass with respect to competition with food production and prices, land use and carbon debt. Strategies that address these suitably are maybe the most rate-determining step for approval and success in biofuel production and use of the co-produced rest streams in non-food/feed technical applications.
22.3 Protein isolation, hydrolysis and isolation of amino acid chemical feedstocks
Proteins need to be isolated and subsequently hydrolysed before their amino acid content can be utilised as a feedstock for chemicals. Several techniques have been developed for the isolation of protein from biomass residues, which are based on extraction using aqueous acid, base or ethanol (Fig. 22.1). It was observed that solutions combined with protease or cellulase aided solubilisation (Dale et al., 2009; Bals et al., 2009; Cookman and Glatz, 2009; Wolf and Lawton, 1997).
22.3.1 Protein isolation from residues of vegetable oil and biofuel production
Oilseeds can be processed in two distinct ways. Traditionally, press-cakes from, for example, rapeseed and sunflower are co-produced with their respective oils in a crushing process. Oil is also often recovered from the seeds by extraction using solvents, e.g. hexane. To eliminate the use of organic solvents, other extraction media such as supercritical fluid extraction or aqueous extraction processing are also used. Thermal or enzyme treatment can be used to enhance oil availability during (aqueous) extraction processes. The method of oil production affects the quality and properties of the remaining protein fraction, which is currently used as animal feed. After oil removal, protein extraction can occur with methods, such as acid, alkaline or protease (enzyme) assisted extraction. This generally yields proteins (with acid or alkaline extraction) or a mixture primarily containing peptides and amino acids (with enzyme-assisted extraction). Alternatively, proteins can be extracted prior to oil extraction to minimise denaturation if intact protein is the main desired product.
Soya protein extraction is a well-developed process. Processing soya affects the solubility and biological activity of the proteins. The native proteins in principle are water soluble. In the process of extracting the oil and subsequent heat processing, the solubility of the protein in water is reduced. The extent of denaturation and reduction in the solubility are related to the intensity of heat treatment(s) that the bean is exposed to. Processing soya beans involves drying, cracking, dehulling and rolling into flakes. The flakes are either milled to yield full fat soya flour or extracted with hexane to remove the oil. Any residual hexane is stripped off using steam. The extracted soya bean flakes are processed into standard meal for animal feed or into special protein products. Generally, the protein content is ca. 45–50 wt%. To obtain a higher protein content, the removal of components other than oil, such as sugars and minerals, is required. The protein can then be dissolved and isolated by iso-electric precipitation. The isolated protein can be hydrolysed to produce amino acids.
A promising technology for protein extraction is based on the ammonia fibre expansion (AFEX) pretreatment technology (Dale, 1981). AFEX involves treating the biomass in an ammonia solution (0.65–3.5 MPa, 70–150°C, 5–15 min). This enables ammonia to permeate in the fibrous structure and solubilise the protein. With a rapid drop in pressure, the cellulosic structure is disrupted. Using this technology, protein is extracted and simultaneously cellulose is processed to be more susceptible to cellulase mediated hydrolysis to fermentable sugars. In a two-step protein extraction with AFEX, ca. 80% of protein present could be extracted (Dale et al., 2009).
22.3.2 Protein hydrolysis
Once extracted, proteins need to be hydrolysed to generate amino acids. In the food industry, it is general practice to hydrolyse vegetable proteins using 6M HCl at elevated temperatures and prolonged periods of time. Major drawback of this method to produce amino acids as chemical feedstocks is the considerable amount of base required to neutralise the hydrolysate leading to high concentrations of inorganic salts. In addition to this, the harsh acid hydrolysis leads to hydrolysis of (hemi-)cellulose to sugars and furans which are dissolved in the amino acid mixture. Both salts and soluble organic molecules need to be separated from the amino acid mixture as it will have a detrimental effect on the efficiency of the downstream processing (Scott et al., 2010). In addition, a number of amino acids are destroyed during the hydrolysis process leading to potential losses.
An alternative protein hydrolysis method, which does not involve formation of large quantities of salts and use elevated temperatures, is based on the use of proteases. Using protease cocktails, hydrolysis yields of up to 90% can be achieved (Hill and Schmidt, 1962). On the downside, use of proteases for protein hydrolysis incurs high enzyme costs. To enable the reuse of proteases, and reduce costs, immobilisation might prove an option. One economically attractive method could be the formation of cross-linked enzyme aggregates (CLEAs) and would also protect against autolysis (Zhu and Pittman, 2003; Sangeetha and Abraham, 2008).
A recent development is the application of supercritical water in the simultaneous extraction and hydrolysis of proteins from biomass residues (Esteban et al., 2010; Kang and Chun, 2004; Sereewatthanawut et al., 2008; Zhu et al., 2010). Under supercritical conditions, water exhibits a lower dielectric constant and a higher ion product of water at ambient conditions. This results in a higher H3O+ and OH− ion concentration, meaning both acid and base catalysed reactions can be performed without addition of acid or base. However, reported hydrolysis yields are limited and addition of low concentrations of acid or base is still required. Major drawbacks in the use of supercritical water are its corrosiveness (which will affect the reactor investment costs) and the high energy input required for its formation.
22.3.3 Isolation of desired amino acids
The hydrolysis of bio-derived proteins results in a complex mixture of 20 amino acids. Some of these have the potential to be converted into chemicals, while others, such as methionine, may be more useful as animal feed (Leuchtenberger et al., 2005). To make the best use of protein, or amino acid rest streams, it is necessary to separate them in order to make them applicable for their different applications.
In the case of simple mixtures of amino acids which differ enough with respect to their isoelectric point and polarity, selective precipitation can be achieved based on pH adjustment or addition of an organic solvent, or both, in combination with cooling and concentration (Garcia, 1999). Ion-exchange chromatography is more effective in the separation of complex mixtures of amino acids. However, for large-scale applications, high and unfeasible costs are anticipated due to low throughput and the need for regeneration (Scott et al., 2010). Separation of amino acids via chromatography is still only used for analytical purposes.
A promising technology to achieve amino acid separation is electrodialysis (ED), which removes ions from a dilute solution through ion-exchange membranes to another concentrate solution by applying an electrical potential as the driving force. In principle, it could separate the protein hydrolysate into acidic, basic and neutral amino acid streams according to their charge behaviour (Hara, 1963; Sandeaux et al., 1998; Krol, 1997). Further ED separation within these streams is challenging due to the similarity of their isoelectric points. To overcome this, the charge behaviour can be specifically modified. Amino acid modification may be achieved by removing groups, such as the carboxylic acid functionality. This may be achieved using bio-catalysis. If the modification directly leads to an end product, both chemical production and separation may be achieved in a sustainable route.
An example is the two basic amino acids arginine and lysine which can be obtained by either precipitation or via ED. Both of them are interesting feedstocks for the production of diamines. Lysine decarboxylase can be used to specifically convert lysine to 1,5-pentanediamine (PDA) in the presence of arginine to produce products with different charge, thus allowing isolation of products by subsequent electrodialysis (Teng et al., 2011). In a similar approach, the acidic acids glutamic and aspartic acid may be isolated by ED and glutamic acid specifically modified to pyroglutamic acid. Based on differences in solubility, the aspartic acid can be isolated and the pyroglutamic acid converted back to glutamic acid (Teng et al., 2012). It should be noted that the presence of salt during ED separation will limit efficiency. Therefore the method of protein hydrolysis will be important for a cost-effective separation process. In general, no large-scale technique for the separation of amino acids from complex mixtures exists and this will need to be developed to enable succesful conversion from biomass to amino acids to chemicals in biorefineries.
22.4 (Bio)chemical conversion of amino acids to platform and speciality chemicals
A review was published which describes a number of known reactions of amino acids to a variety of chemicals which included monomers and amines. While the reactions were known, and the chemical products are of industrial significance, most reactions were only ever reported as reactions/decomposition of amino acids in a nutritional context and not as a preparative method (Scott et al., 2007). Since then, investigation into the (bio)chemical transformation of a number of amino acids has been explored.
The formation of acrylamide from asparagine and aspartic acid has been reported but only in the context of the formation of toxic compounds during cooking (Taeymans et al., 2004). Recently, Könst et al. (2009) described the conversion of aspartic acid to β-alanine using L-aspartate α-decarboxylase. Although not disclosed by the authors, routes to both acylamide and acrylonitrile were suggested.
Glutamic acid is produced by the hydrolysis of the amide functionality of glutamine. Glutamic acid can undergo enzymatic decarboxylation, resulting in the formation of γ-aminobutyric acid (GABA). Werpy and Petersen report the use of glutamic acid, produced from fermentation, as a building block for a number of chemical products such as 5-amino-1-butanol and glutaric acid (Werpy and Petersen, 2004). More recently glutamic acid as a platform for a number of chemicals has been described, as shown in Fig. 22.2. A fed batch process using glutamic acid and immobilised glutamic acid α-decarboxylase has been described (Lammens et al., 2009). The resultant GABA can then undergo ring closure and methylation using methanol in the presence of a catalytic amount of a metal halide salt to form N-methylpyrrolidone (NMP) (Lammens et al., 2010). In further articles from the same author, it was also shown that this route shows a favourable techno-economic as well as a positive ecological footprint compared to the current petrochemical route (Lammens et al., 2012b). 2-Pyrrolidone, used for the production of N-vinylpyrrolidone, can also be readily formed by the cyclisation of GABA. Further, glutamic acid can undergo oxidative decarboxylation to cyanopropanoic acid followed by subsequent decarbonylation elimination reaction, resulting in the formation of acrylonitrile (Le Notre et al., 2011).
The conversion of phenylalanine to cinnamic acid has been extensively described using phenylammonia lyase (Nijkamp et al., 2005; Ben-Bassat et al., 2005; Camm and Towers, 1973; Ogata et al., 1966). Subsequent decarboxylation of cinnamic acid results in the formation of styrene (Dahlig, 1955). An alternative method to produce styrene from cinnamic acid involves cross-metathesis (ethenolysis) (Sanders et al., 2010). Here, cinnamic acid undergoes a metathesis reaction with ethylene for the simulataneous production of styrene and acrylic acid, as shown in Fig. 22.3 (Spekreijse et al., 2012).
The synthesis of other aromatic compounds has been explored. Tryptophan has also been shown to undergo photocatalysis leading to intermediates that can be used in a route towards bio-based aniline (Hamdy et al., 2012; Sanders et al., 2009).
22.5 Alternative and novel feedstocks and production routes
22.5.1 Protein from leaves
Leaves are highly abundant, but have received little attention in their application. However, they contain many valuable components, such as protein, which can be used for industrial applications. Depending on species, protein content in leaf varies from 15% to 30%. Already in the 1960s, leaf protein was considered as a potential feed or food resource (Akeson and Stahmann, 1965; Gerloff et al., 1965), but its applications were limited due to a high proportion of insoluble protein and insufficient cost-efficient processing. Soluble protein varied from 15% to 60% of total protein, depending on species and processing methods (Kammes et al., 2011; Chiesa and Gnansounou, 2011). As well as this, high fibre content and other anti-nutritional factors, such as cyanide and tannins, reduce the viability of leaves as a protein source for humans, but it is used in feed, due to the amino acid composition. Other specific leaves such as sugarbeet, cassava and grass have also been evaluated as a feedstock (Lammens et al., 2012a).
22.5.2 Protein from algae
Microalgae has been promoted as a future source of transportation fuels. It has been estimated that production of biodiesel from algae to replace current European diesel requirements would lead to 300 Mton of algal protein (Wijffels and Barbosa, 2010).
22.5.3 Protein from potato starch production residues
During the processing of potatoes for starch extraction, the main waste stream is Protamylasse™, which mainly contains sugars, organic acids, proteins and free amino acids and currently has a limited use. Allied to this, it has been shown to be a promising substrate for the production of cyanophycin, a polymer consisting of a poly(aspartic acid) backbone with arginine side chains (Elbahloul et al., 2005; Mooibroek et al., 2007). Interestingly, this polymer is insoluble under physiological conditions, therefore offering the opportunity to isolate specific amino acids from a complex mixture and aid downstream processing. Cyanophycin has been demonstrated to lead to a number of chemicals (Könst et al., 2010, 2011a, 2011b).
The genes to produce cyanophycin can be incorporated into yeast for the simultaneous production of ethanol and cyanophycin in media containing sugars and amino acids. It could be conceived that such an approach incorporated into current ethanol production could lead to co-production of new value-added products.
22.6 Conclusion and future trends
There are a number of challenges that need to be tackled to avoid adverse effects of the use of biomass for non-food applications. Some ways to do this include:
• certification of biomass on sustainability criteria such as the influence of cultivation on the environment and social issues such as impact on food (production and market) and working conditions
• use of non-food crops on arid land which could not otherwise be cultivated for food; an example is the use of Jatropha curcas as a biofuel and non-food protein source
• use of (lignocellulosic) rest streams from primary agricultural production.
Improved biorefinery of agricultural materials will produce sufficient quality food and lead to suitable waste streams that can be used for other non-food technical and value-added applications. Activities and incentives to further develop the biorefinery concent are growing at a rapid pace.
Non-food application of amino acids from protein isolated from biomass required several specific technological solutions. Mild separation of proteins is vital where proteins are to be used in applications where functionality is important. This area has received some attention, but as yet cannot be carried out at low cost. The development of cost-effective techniques to isolate all single amino acids is a technology that is crucial for application of amino acids as chemical feedstocks. Some attempts to use specific conversions to aid separation have been studied and this could lead to more cost-effective in situ product formation and recovery. This is especially true during the isolation of specific molecules from dilute watery steams. Chromatography, while efficient, utilises significant chemicals for regeneration and the throughput for voluminous streams may be prohibitive. Concentration of product streams by distillation is both energy and cost intensive. To overcome this, an approach (using microorganisms) which concentrates molecules as insoluble materials allowing more straightforward isolation seems interesting.
In general, protein and amino acids can be important by-products in many biorefinery processes that are now aiming at producing other components like oil and sugar/ethanol. However, several other (groups) of components can be put to good use and thus create additional value for the overall chain. Examples of these components include fibres, (hemi)cellulose, phosphate, potassium (K +), organic acids, vitamins and lignin.