Short title: Ester Formation

DESCRIPTION

FIELD OF THE INVENTION

The present invention is in the field of a process for producing an ester, such as a sustainable ester, from an aqueous solution comprising a carbohydrate, batch wise or continuously, wherein use of raw material is limited and if possible re-used. The present invention is in the field of green technology.

BACKGROUND OF THE INVENTION

Dimethyl carbonate (DMC) is a liquid organic compound with the formula OC(OCH ₃ ) ₂ . As it is colourless and flammable it has to be handled with care. It can be used as a methylating agent, but many of these reactions require a high pressure, such as higher than about 1000 kPa . Further, despite its much lower toxicity and its biodegradability it is not used very much as it is a relatively weak methylating agent compared to e.g. typically used reagents. It is noted that as a

consequence production and application of dimethyl carbonate is limited in volume. Use is considered to be limited to gas- liquid interaction.

Many compounds, such as those originating from natural sources such as sugar and cellulose, may be used as a starting point for forming further molecules. These compounds may be considered as chemical building blocks. Such compounds are a potential "carbon-neutral" feedstock for fuels and chemicals. An example is hydroxymethylfurfural (HMF) , or 5- (Hydroxymethyl ) furfural . Recently, an enzyme from Cupriavidus basilensis (HMF14) was used to convert HMF to FDCA using molecular oxygen. Also Pseudomonas putida can convert HMF to FDCA. Such bioconversion is typically effected in water, at ambient temperature and pressure.

HMF is typically produced from sugars, such as through dehydration of fructose. HMF is also naturally generated in sugar-containing food during heat-treatments and it is formed in the Maillard reaction. HMF can be converted to 2,5- dimethylfuran (DMF) , which is a liquid biofuel. HMF can be converted into 2 , 5-furandicarboxylic acid (FDCA) by oxidation. It is noted that dehydration processes using

hydroxymethylfurfural (HMF) as intermediate are generally non- selective, not cost effective and there is no industrially viable oxidation technology that can operate in concert with the necessary dehydration processes.

Several dicarboxylic acids are of industrial relevance, and so are their esters. FDCA and its esters are valuable sources in production of polyesters. FDCA can be used as an important renewable resource e.g. to substitute terephthalic acid (PTA) and to form polymers. There are various further derivatives of FDCA. FDCA undergoes reactions typical for carboxylic acids. FDCA has also been applied in pharmacology. Other dicarboxylic acids are considered important platform chemicals such as fumaric, malic, itaconic and succinic acid. Due to their chemical functionality, succinic acid can be further transformed into different chemical derivatives and into polymers. Succinic acid undergoes reactions typical for carboxylic acids. The uses of succinic acid are broad in the chemical industry. Many microorganisms are able to convert carbohydrates to succinic acid. Recently, engineered

Corynebacterium glutamicum strains, for example, have been used to produce succinic acid in high yields and

concentrations .

Succinic acid and its esters, e.g. dimethyl succinate, can be used in the production of valuable polyesters such as poly (butylene succinate) (PBS). PBS is a fully biodegradable polymer with good thermal stability and processing ability. Other poly (homo- and copolyesters ) can be produced by

condensation of succinic acid diesters with diols.

Applications and market for those polymers have been

increasing in the latest years.

The market size for sustainable esters is still relatively small (about 0.1 million tonnes) at a selling price range of a few dollar per kilogram. Higher production cost of these esters compared with e.g. petrochemical-based esters has hampered the commercialization, therefore there is a need to reduce production cost of esters, to increase ester

biosynthesis capability, to broaden a utilizable substrate range, and to produce novel polymers.

The production of several carboxylates has been hampered by its recovery and purification. Conventional methods, e.g. precipitation, consume substantial amounts of chemicals and produce stoichiometric waste, therefore are considered polluting processes. Alternatives have been described

elsewhere. Various patents recite a process in which a carboxylic acid (namely lactic and succinic acid) or a carboxylate is extracted by an amine-based extractant, or adsorbed by a basic sorbent. Regeneration of the auxiliary phase is performed using ammonia or a low molecular weight amine ( trimethylamine) which forms an ammonium carboxylate salt that can be decomposed thermally if required, yielding a desired carboxylic acid. The extraction/sorption stage is greatly hindered by the solution pH and the basicity of the ion exchanger. In the case of succinic acid, at neutral pH a mixture of the (carboxylic) acid and its monovalent and divalent salts is obtained, which is unwanted. The process needs improvement in order to enhance e.g. stability of the ammonium carboxylate salt and to reduce the amount of sorbed salt by e.g. carbon dioxide acidification. So e.g. succinic acid recovery at neutral pH is not provided and further at the best a mixture of salts is produced.

Some other (succinic acid) purification processes are energy demanding, can create corrosion problems, and have mass and heat transfer problems.

Production of esters from carboxylates will require its purification and acidification preceding ester formation by conventional esterification reaction.

In some publications concepts of acidification and salt-to- acid conversion are misconceived, such as that ion-exchange based process does not require acidification and effectively acidification takes place by exchanged hydrogen ions from the resin .

Further a process for production of especially diethyl succinate from succinate salts and ethanol using sulphuric acid for acidification is recited. The process intends to avoid acid purification prior to an esterification step. The advantages are limited to that aspect and to the purity of the obtained ester. On the down side mineral acid is consumed and waste salts are produced. The process can not be integrated further .

Also production of alkyl esters from several solid metal carboxylates e.g. via conventional esterification using carbonic acid is recited. Reaction conditions however are harsh (P = 20-60 bar and T = 175 °C) . Disadvantages are a relative low yield, consumption of methanol, formation of by ^¬ products and low selectivity, amongst others.

Some organic processes relate to phase 'transfer catalysis

(PTC) . In a regular application, catalytic amounts of phase transfer agents are used which in principle facilitate interphase transfer of species, making reactions between reagents in two immiscible phases possible. PTC can be used in both liquid-liquid and solid-liquid systems. PTC typically involves a series of equilibrium and mass-transfer steps, beside main reactions. Conversion efficiencies are typically limited, e.g. due to low activity of catalysts. In general, the recovery of chemicals (catalysts) is at least difficult. Typically PTC is combined with other enhancement techniques to overcome disadvantages, albeit these are not explored in detail yet, despite promising chances (see e.g. S.D. Nalk and L.K. Doraiswamy, AIChE Journal, March 1998, Vol. 44, No. 3 pp.612-646) .

Some further articles describe the application of PTC to the production of esters. Moreover a process for production of methyl esters from carboxylic acids or salts using dimethyl carbonate in the presence of a strong base homogeneous

catalyst ( 1 , 8-Diazabicycloundec-7-ene known as DBU, 1,4- diazabicyclo [2.2.2] octane known as DABCO, 1,5,7- Triazabicyclo [ 4.4.0 ] dec-5-ene knows as TBD or 4- Dimethylaminopyridine known as DMAP) under microwave

irradiation is described. Despite a focus on fine chemicals, it covers a wide range of carboxylates . Furthermore, it shows feasibility of using a base as a catalyst for the alkylation reaction. However chemicals are not recovered, nor can

biobased processes be integrated.

In a similar context, it is recited the use of anion exchange resins for O-alkylation of carboxylate ions. Anions (substrates) tested were alkyl and aryl mono carboxylates. As alkylating agents, strong alkyl- and aryl- halides (mostly bromides) were used. It recites a purely synthetic strategy, without much further details. As a disadvantage the process is not cost effective and not green. Incidentally W098/15519 A2 recites a process for the recovery of purified lactic acid values from an aqueous feed solution containing lactic acid, lactic acid salt, or mixtures thereof, comprising: bringing the feed solution into contact with a substantially immiscible anion exchanger to form a substantially water-immiscible phase comprising an anion exchanger-lactic acid adduct; effecting a condensation

reaction in the substantially water-immiscible phase between a carboxylic moiety of the lactic acid adduct and a moiety selected from a hydroxyl moiety and a primary or secondary amine moiety to respectively form a lactic acid ester or amine product; and separating the formed lactic acid product from the anion exchanger. This document is considered background art, in a neighbouring field. The process further suffers from at least a few of the disadvantages mentioned, including superfluous use of chemicals.

The present invention therefore relates to a process for obtaining a reaction product, suitable for use as a bio-based chemical building block, which overcomes one or more of the above disadvantages, such as minimizing chemical consumption, using renewable material, providing a highly pure compound, being non-toxic, at low (energy) costs, without jeopardizing functionality and advantages.

SUMMARY OF THE INVENTION

The present invention relates in a first aspect to a green process minimizing use of chemicals and energy for producing a carboxylate ester, such as a biobased ester, from a renewable resource according to claim 1. Therein the dialkyl carbonate and specifically DMC is considered as a green methylating agent.

Therein consumption of chemicals such as inorganic acids and bases, and associated production of inorganic waste salt are prevented, contrary to conventional methods that produce up to 1 kg salt per kg ester.

The present process can start with a bio-based production (fermentative or enzymatic) of a carboxylate. Such is carried out while controlling the pH at neutral

values (4-10) using e.g. an inorganic base. Thereafter the carboxylate may be captured using an anion exchange sorption step which may liberate the inorganic base above. In an example a bicarbonate or carbonate base is also an inorganic carbon source for a microorganism. As a key step therein the (sorbed) carboxylate is transformed in an ester derivative with a dialkyl carbonate, such as DMC, and

preferably with simultaneous regeneration of an anion

exchanger, for example into the (bi ) carbonate form. Optionally an ester produced can be purified with ease and refined by e.g. conventional means.

It is noted that use of DMC in synthesis of fine chemicals is known to some extent. Therein it is purely used as a reagent and no integration with (product) recovery occurs. In addition, when DMC is used as methylating agent in organic chemistry applications carbon dioxide and methanol are

released and are not used in the (subsequent) process. In the present process, on the contrary, such inorganic carbon will remain attached to a support such as a resin and is (re-) used in e.g. the fermentation as inorganic carbon source. In the example the inorganic carbon may be the counterion of the resin. Even further, a support such as an anion exchange resin has not been used previously as a catalyst. Most often strong organic bases are used (DBU, TBD) which are expensive and difficult to recover after reaction.

Ester derivatives of target carboxylates have not been previously synthetized by alkylation using dialkyl carbonate; especially esters of succinate and FDCA have not been

synthesized before using DMC. Mild conditions, such as neutral pH, relatively low temperature and low pressure would be beneficial and are provided by the present process. In

addition, the present integrated process and regeneration of e.g. a (strong) anion exchange resin is not known. The present level of integration comprising e.g. an anion exchanger as a core of a downstream process is novel. In the present direct downstream catalysis chemicals and auxiliary phases are seized to a maximum, e.g. an ion exchanger is used as a sorbent and catalyst and the alkylating agent as

reagent and regenerant . Furthermore, the way of supplying (re ^¬ using) required inorganic carbon in bio-based production, e.g. fermentation, by integrating both bio-based production and sorption processes is also novel. It is noted that to obtain methyl esters, methanol is obviously cheaper than dimethyl carbonate. However, when using methanol, water needs to be evaporated from the carboxylate salt, and the carboxylate needs to be acidified with an inorganic acid, producing stoichiometric amounts of waste salts. This is not the case for e.g. dimethyl carbonate.

The present process is more efficient and largely

integrated. It uses less materials, energy and equipment and it produces less or no waste (salts) .

After thorough scientific research, in particular relating to recovery of carboxylates, such as succinate and 2,5- furandicarboxylate, and synthesis of diesters thereof,

inventors have arrived at the present process. Equilibrium studies (e.g. Ion exchange isotherms) and column dynamics (sorption breakthrough profiles) have

been carried out for the above mentioned carboxylates using several ion exchange resins. As a result, static and dynamic sorption capacities as well as selectivities and affinities have been determined. With respect to the synthesis of

diesters by alkylation, experiments were carried out to establish a basic understanding about the reaction

stoichiometry since the present reaction has not been reported before to the knowledge of inventors. Also influence of water and temperature on reaction kinetics was quantified. In

addition, resin stability was tested after several sorption- reaction cycles. Also an impact of e.g. resin functional group (basicity) on the present integrated process was studied.

Thereby the present invention provides a solution to one or more of the above mentioned problems .

Advantages of the present description are detailed

throughout the description.

DETAILED DESCRIPTION OF THE INVENTION

The present invention relates in a first aspect to a green process minimizing use of chemicals and energy for producing a carboxylate ester, such as a biobased ester, from a renewable resource comprising the steps:

providing a water soluble carboxylate,

providing a transfer agent,

removing water, if present, and

providing a di-alkyl carbonate, reacting the carboxylate, transfer agent and di-alkyl carbonate thereby forming a di-alkyl ester of the carboxylate, and a carbonate, substantially in absence of water. In an example the process relates to sorbing the carboxylate on the transfer agent, thus releasing a carbonate base, removing water to a required (minimal) extent, if present, reacting the sorbed carboxylate with a di-alkyl carbonate thereby forming an alkyl ester of the carboxylate and regenerating the

transfer agent to a carbonate form.

In the present process the carboxylate is water soluble, as e.g. it is produced in an aqueous solution, such as a broth, it is e.g. transferred in an aqueous solution, and typically an acidic form of the carboxylate can only be obtained at relative low pH, such as pH = 2, which is

cumbersome. In an aspect the water soluble carboxylate allows for a high degree of integration of process steps.

The present transfer agent performs a catalytic action, and may perform further actions, such as capturing and

releasing reagents, such as by sorption and desorption. It preferably comprises an exchangeable counterion, as in e.g. an anion exchanger. In an example the transfer agent is capable of sorbing and acts as a catalyst. The present transfer agent can be regenerated in the present process. Specifically in an example the present transfer agent is capable of supporting (the process of) liquid-liquid and solid-liquid transfer catalysis. The transfer agent is typically provided in an auxiliary phase. Therewith the transfer agent assists an envisaged reaction.

Preferably the present process relates to liquid-liquid and solid-liquid phase transfer catalysis, such as with an alkyl carbonate organic phase and a carboxylate-laden solid support, such as an anion exchanger.

The present di-alkyl carbonate may have two alkyls being the same or being different.

The present reaction provides optimal results in

(substantial) absence of water. Thereto water, if present, is removed to a large extent, such as 95% or more, preferably 99% or more, even more preferably 99.5% or more, such as 99.9% or more. Thereto a solid support, such as an exchanger, is dried, such as with an alcohol, such as with methanol produced in the present process.

The present process provides e.g. high yields, minimised waste production, advanced integration, at low (energy) costs.

In an example of the present process the transfer agent is a positively charged transfer agent. In an example it

preferably comprises a nitrogen atom, such as a quaternary nitrogen atom, a guanidine, a guanidinium, a quaternary onium, N, -N-dimethylformamide , N-methyl-2-pyrrolidone, and tetra methyl urea. It may also relate to an anion exchange resin, such as Dowex MSA. The transfer agent preferably is supported on a solid or diluted in a suited liquid phase, such as on a porous material, a polymer structure, silica, functionalized material, organic insoluble aliphatic or aromatic hydrocarbon, long-chain (>C ₈ ) alcohol, and combinations thereof. It has been found that these transfer agents perform best e.g. in terms of yields, waste, etc.

In an example of the present process the carboxylate is produced by biotransformation, such as microbial or enzymatic transformation, or a combination thereof, such as by

conversion by a microorganism from a broth, such as with C. glutamicum or P. putida, preferably conversion in an aqueous medium, wherein a pH of the medium is from 4-10, preferably from 5-9, more preferably from 6-8. Therewith high yields are obtained. Further a high degree of integration with such transformation may be obtained.

In an example of the present process the carboxylate is a mono-, di- or tri- carboxylate, preferably selected from acetate, acrylate, adipate, benzoate, butyrate, FDCA,

fumarate, 3-hydroxybutyrate , 3-hydroxypropionate , itaconate, lactate, malate, pimelate, propionate, succinate, and

combinations thereof. As is e.g. shown by example 4 any

combination of carboxylates may be used. Such indicates the present process is relatively robust and can be used over a broad scope of carboxylates and mixtures thereof. Such is particularly relevant if the one or more carboxylates are produced by biotransformation. It is noted that the above is also applicable to a broad scope of di-alkyl carbonates, and combinations thereof. From a practical point of view an actual scope of di-alkyl carbonates may be limited, in order to obtain a limited set of products (di-alkyl esters of carboxylates ) .

In an example, using DMC, methyl acetate, methyl acrylate, dimethyl adipate, dimethyl benzoate, methyl butyrate, dimethyl FDCA, dimethyl fumarate, methyl 3-hydroxybutyrate, methyl 3- hydroxypropionate, dimethyl itaconate, D- or L-methyl lactate, dimethyl malate, dimethyl pimelate, methyl propionate,

dimethyl succinate, and combinations thereof are formed. Also using diethyl carbonate, ethyl acetate, ethyl acrylate,

diethyl adipate, diethyl benzoate, ethyl butyrate, diethyl FDCA, diethyl fumarate, ethyl 3-hydroxybutyrate, ethyl 3- hydroxypropionate, diethyl itaconate, D- or L-ethyl lactate, diethyl malate, diethyl pimelate, ethyl propionate, diethyl succinate, and combinations thereof are formed. The above esters relate to readily available carboxylates, which can be formed relatively easy with the carbonates, providing the advantages of the invention.

In an example of the present process the ester is a mono-, di- or tri ester, preferably a diester in its final form. The ester preferably comprises one, two or three alkyl groups as indicated below, such as methyl, ethyl, propyl and

combinations thereof.

In an example of the present process the di-alkyl

carbonate comprises an alkyl selected from the group of C1-C12, preferably C1-C6, such as methyl, ethyl, propyl, butyl, iso- propyl, pentyl and hexyl, C1-C12 cyclic carbonates, preferably ethylene, propylene and butylene, and combinations thereof.

The di-alkyl carbonate may comprise cyclic compounds with linked alkyl chains such as ethylene carbonate (the carbonate of ethylene glycol) . Especially the smaller alkyls perform well, such as methyl and ethyl.

In an example of the present process the water soluble carboxylate comprises one or more of Na ⁺ , NH4 ⁺ , K ⁺ , Mg2 ⁺ , Ca2 ⁺ , and combinations thereof.

In an example of the present process the carboxylate is sorbed such as by an anion exchanger from an aqueous solution, i.e. fermentation broth with or without previous purification. In a general example, the carboxylate is sorbed in a fixed bed operation after protein and cell removal performed by

conventional means such, as filtration operations. In a particular example, the carboxylate is sorbed in expanded bed mode without the need of protein or cell removal. In a

particular instance of such example, both processes, i.e.

biotransformation and sorption, are integrated and occur simultaneously.

In an example of the present process the water soluble carboxylate is sorbed such as by an anion exchanger, wherein the anion exchanger comprises the transfer agent, wherein the transfer agent comprises a (bi ) carbonate species, thereby forming an aqueous (bi ) carbonate salt, and a complex of the carboxylate and transfer agent,

wherein the (bi ) carbonate salt solution is preferably recovered to be used in the production of the carboxylate. An advantage thereof is reduction of waste.

In an example of the present process the resin is dried, such as by desiccation, by heating, or by washing using a suitable drying solvent, preferably an alcohol, such as ethanol or methanol, and thereafter the di-alkyl carbonate is provided to the resin in order to form the ester, preferably provided as a liquid,

wherein the carbonate is preferably provided to the transfer agent.

In an example of the present process the ester is formed at a temperature of 20-140 °C, preferably 25-120 °C, at a pressure of 80-900 kPa, such as at 90-500 kPa, preferably at or near ambient conditions. Such will e.g. depend on whether the reaction is carried out in a Liquid-Solid or Gas-Solid mode. L-S modes typically require pressures above 101 kPa to keep e.g. DMC in a liquid state. Such temperatures and

pressures are relatively mild.

In an example of the present process the ester is purified without much trouble, such as by distillation,

crystallisation, or combination thereof.

In an example of the present process the carboxylate is purified.

In an example of the present process the reacting is carried out as a solid-liquid, gas-solid or liquid-liquid reaction, preferably as a solid-liquid reaction.

In a second aspect the present invention relates to a carboxylate ester obtainable by a process according to the invention, preferably wherein the carboxylate is produced by biotransformation. It is noted that such an ester will most likely comprise small amounts of side products, such as DNA of microorganisms. As such the present carboxylate ester will distinguish itself from e.g. purely chemically synthesized esters .

The one or more of the above examples and embodiments may be combined, falling within the scope of the invention.

EXAMPLES

The invention is further detailed by the accompanying figures, which are exemplary and explanatory of nature and are not limiting the scope of the invention. To the person skilled in the art it may be clear that many variants, being obvious or not, may be conceivable falling within the scope of protection, defined by the present claims.

Example 1: Conversion of aqueous carboxylate, such as succinate- or FDCA- dianion into its respective dimethyl esters without isolating succinate or FDCA

It is observed that microbial or enzymatic production of carboxylic acids is usually more efficient at neutral pH than at low pH, as uncharged carboxylic acids may be toxic or inhibiting.

FDCA is so acidic that its production from . HMF will yield its corresponding anion unless the pH is 2 or lower. That pH is not feasible with typical microorganisms. Succinic acid has been produced from glucose at pH down to 3, but this is at the expense of the performance of the microorganisms used. In order to control the pH close to neutral values in state of the art processes, a neutralizing base is added during fermentation/enzymatic conversion, and carboxylate anions are obtained. Further, during downstream processing, an inorganic acid is added to recover the carboxylic acid. This leads to stoichiometric inorganic salt production, usually gypsum. In some cases such salt can be used, for example as fertilizer, and in some other cases bipolar electro-dialysis is used to split the salt into acids and bases that can be reused. However, these methods have limitations and drawbacks. Consequently, recovery of succinate or FDCA requires acidification and crystallization, and yields stoichiometric amounts of waste inorganic salt. This hampers the economic and ecologic sustainability of a such processes.

For the above mentioned dicarboxylic acids, the present process can be combined with fermentation and/or biotransformation at neutral pH (5-9) and it avoids acidification and crystallization. As an advantage it generates an inorganic base to be used in controlling e.g. fermentation pH.

The present process provides an advantageous bio-based ester production process from succinate or FDCA produced from renewable resources. Such seguence of operations is schematically presented as a flow diagram in Fig. 1.

Referring to Fig. 1, renewable resources, such as glucose or HMF, are converted by a specific biological agent to for example succinate or FDCA disodium salts (Na ₂ A) in a bioreactor 101, cells can be removed in 102 and recycled back to 101. In certain embodiments cell removal operation 102 can be avoided, such scenario is included into the proposed invention. The disodium salt is sent to a column operation 103 packed with a suitable anion exchanger in a (bi ) carbonate form, thus releasing (bi ) carbonate base as the carboxylate sorption takes place. Succinate or FDCA anions remain bound to the anion exchanger. The generated base is recycled to the bioreactor 101 for pH control. In the case of succinate production, certain bacteria may also consume part of the carbonate since inorganic carbon is reguired, therefore minimizing chemical consumption. Previous to the ester production, water is removed from the carboxylate- laden anion exchanger (Q ₂ A) in 104. Ester production is carried out in 105 by O-alkylation using dimethyl carbonate (DMC) , a cheap green solvent and reagent, in liquid or gas state. The reaction directly yields the diester of the respective carboxylate (DMA) , i.e. dimethyl succinate or dimethyl 2 , 5-furandicarboxylate , without any monomethyl ester. As ester formation takes place, regeneration of the anion exchanger to a (bi ) carbonate form is achieved. In certain reaction conditions, methanol can be produced as a reaction by-product.

DMC may need to be removed in 106 from the regenerated anion exchanger if reaction is carried out in liquid-solid conditions. The regenerated anion exchanger ( Q ₂ C0 ₃ , QHC0 ₃ ) can be reused in 103 in a subsequent sorption cycle. Operations 103, 104, 105 and 106 can be performed in the same process unit and therefore such operation mode is within the scope of this invention. The raw DMA stream, composed mainly of diester, methanol and DMC, from unit 105 is purified easily by conventional means, e.g. distillation, in 106 and pure diester is obtained. For the case of FDCA, .. dimethyl 2,5- furandicarboxylate is a preferred monomer for esterification with ethylene glycol towards PET-compatible polymers .

In a broader process embodiment, DMA can be polymerized in 108. Methanol produced as by-product can be recovered from the purification unit 107 and the polymerization process 108 and recycled to dimethyl carbonate synthesis 109.

Example 2: Conversion of succinic acid.

Succinic acid sorption from aqueous solutions

In the described process scheme, sorption is carried out using carbonate or bicarbonate counter-ions. The anion exchange reactions are:

Q ₂ C0 ₃ + Na ₂ Succ < >Q ₂ Succ + Na ₂ C0 ₃

2QHC0 ₃ + Na ₂ Succ^→Q ₂ Succ + 2NaHC0 ₃

Therein Q is a transfer agent and Q species are supported in the resin. The equilibrium characteristics of these reactions are not known and they have not been described in literature. Thus, succinate sorption isotherms were determined for both possible counter-ions.

It is noted that reaction characteristics of alkylation of e.g. sorbed carboxylates using dialkyl carbonates in general and DMC in particular have been topic of present studies .

Experimental

In a column operation, 80 mL of Dowex MSA resin (Dow chemical, strong anion exchange resin, macroporous structure, type I, CI ^" form) was transformed to the required forms (OH ^" , HC0 ₃ ^~ and CO ₃ ^2" ) . The resin bed was washed with deionized water (10 bed volumes), then converted to the hydroxide form by eluting chloride ion with 10 bed volumes of 1M NaOH solution. Superficial velocities were maintained close to 1 m/h according to supplier specifications. Given the column geometry, a flow rate of 8 mL/min was required. After conversion to OH- form, the column was washed until a pH close to neutrality (7-7.5) . The column suffered a volume expansion of approximately 18%.

The resin in the OH- form was further converted to bicarbonate and carbonate forms. Elution using 1 M solutions of sodium bicarbonate and sodium carbonate were used. Bed expansions for these forms were 4.6 and 11.2%, respectively. The column was subsequently washed with deionized water until pH near to neutral.

Batch sorption experiments were performed in scintillation vials using a 1:10 (w:v) phase ratio for four hours. Neutral succinate solutions were used. Agitation and temperature were controlled to 120 rpm and 25°C. Succinate concentrations were determined by HPLC using a Waters HPLC system comprising a Bio-Rad Aminex HPX-87H column (7.8 x 300 mm) . Phosphoric acid was used as an eluent. Quantification was done by UV detection at 210 nm using an external standard .

Results

The succinate sorption equilibrium distribution dependence for the tested counter-ions is shown in Fig. 2 (Succinate loading (g/g dry resin) versus total succinate (g/1) . As expected, succinate sorption equilibrium is favoured by bicarbonate and carbonate counter-ions . The saturation capacities are 0.18 and 0.22 g/g dry resin, respectively. Affinities are greater as well. Thereby, a column sorption operation based on either of these exchanger counter-ions will perform better than chloride-based operations (mainly narrower mass transfer zone, improving operating column capacity) .

The ^' equilibrium pH corresponding to the sorption isotherms was determined for each counter-ion (Fig. 3, equilibrium pH versus total succinate (g/1)) . The observed trends were expected, although the pH values for carbonate indicate some dissociation to bicarbonate species. Study of reaction characteristics.

Succinate ester production via O-alkylation using DMC has not been described in previous literature. It is known that DMC has tuneable active centres which reactivity depends on the temperature, being 120°C a regular temperature for methylation. In the presence of water, the expected stoichiometry is as follows:

2CH ₃ OCOOCH ₃ + Q ₂ Succ + 2H ₂ 0 → 2QHC0 ₃ + H ₃ CSuccCH ₃ + 2CH ₃ OH

. CH ₃ OCOOCH ₃ + Q ₂ Succ → Q ₂ C0 ₃ + H ₃ CSuccCH ₃

CH ₃ OCOOCH ₃ + H ₂ 0 → +2CH ₃ OH + C0 ₂

In all reactions, dimethyl succinate (DMS) is formed from sorbed succinate anions. In the first two reactions, methanol is produced as by-product and can be recycled for the synthesis of DMC, therefore meeting most of the green chemistry requirements. Water in excess will promote DMC hydrolysis side-reaction.

Experimental

In a typical run, 1.5 g of oven-dried succinate loaded resin (0.24 g succinate/g total dry resin) was loaded on a catalyst addition device held within the autoclave reactor. In the vessel, 30 g of DMC were added. The vessel was then purged 10 times with nitrogen in order to achieve an inert atmosphere. The reactor was then heated up to 120°C and the temperature maintained constant throughout the reaction. Agitation was controlled at 750 rpm. Once the reactor has attained stable conditions, succinate-laden resin was released and O-alkylation reaction started. Sampling was performed regularly to establish reaction kinetics. After eight hours the reactor was cooled down to room temperature and vented, and final concentrations of DMS, methanol and water were determined. DMS identity was confirmed by GC-MS. DMS and methanol were quantified using a Agilent gas chromatography system comprising a HP-INNOWax PEG capillary column (60m x 0.25mm, 0.15 μιη film) . Detection was done using an FID detector, helium was used as a carrier gas. Identity of the compounds was confirmed by gas chromatography- quadrupole mass spectrometry. In order to discern in which form (type of counter-ion) the resin would be after reaction, the resin was eluted and the effluent pH checked. The reacted resin was washed three times with methanol and water and 2 mL of resin were eluted in a column using a solution of 20 g/L of sodium nitrate at 0.5 mL/min .

Results

During the reaction, an increase in the reactor pressure was noticed. This might indicate that low vapour pressure substances are generated during the reaction. It was noticed that after eight hours the reaction was reaching stable pressure.

Fig. 4 shows the DMS yield reaction (time (h) ) profile, defined as mole of DMC produced by mole of succinate sorbed. After 2 h, 0.87 mole DMS/mole succinate was produced. Final reaction yield achieved was 0.93 mole DMS/mole succinate.

The methanol produced exceeded (1.6 times higher) the stoichiometric amount expected. DMC hydrolysis might be occurring .

After reaction, the washed resin was eluted with sodium nitrate. During elution, the pH of the effluent was kept between 9.1 and 8.2. This might be a clear indication that the resin was mainly in the bicarbonate form after reaction.

Conclusion

Using DMC, sodium succinate was converted into 94% dimethyl succinate without further optimization. Sodium bicarbonate or carbonate formed is a useful co-product, particularly in the present process.

Example 3 : Conversion of relevant carboxylates

It has been found that several interesting carboxylates such as acetate, fumarate and malate, can be produced by biological transformations, mostly fermentation processes, from renewable resources. If produced in dedicated

fermentations at neutral pH, these carboxylates can be recovered and transformed to their ester derivatives while recycling the produced bicarbonate base as described in the present process.

Experimental

Individual solutions of 20 g/L of succinic, fumaric, malic and acetic acid were prepared and adjusted to neutral pH using sodium hydroxide. The strong anion exchange resin Dowex MSA in the bicarbonate for was used. Batch loading experiments were performed in scintillation vials using a 1:10 (w:v, resin : solution) phase ratio for four hours until equilibrium was reached. Agitation and temperature were controlled to 120 rpm and 25°C. The carboxylate

concentration of the initial solutions and at equilibrium were determined by HPLC as described in example 2. The amount of carboxylate sorbed to the resin was calculated from corresponding mass balance.

Once the resin was loaded, it was washed quantitatively three times with five volumes of deionized water and

subsequently three times with five volumes of methanol using a Millipore Steriflip 60 μτ nylon net filtration unit. The resin was dried in an oven at 72 °C for 24 h and transferred to a desiccator where vacuum (31 kPa) was applied during 4 h.

Dried resin (0.2 g) was put in a glass tube along with dimethyl carbonate as a solvent (5 g) . The tubes were flushed with nitrogen in order to remove air preventing undesired reactions . The reaction tubes were placed in a heating block at 100 °C for 10 h. Agitation was controlled at 500 rpm. After reaction, the concentration of methyl esters was determined by gas chromatography as described in example 2.

Results

Table 1 shows the obtained batch loading and ester yields for the selected carboxylates . It should be noticed that the batch loading for succinate was lower than the obtained capacity in column operation. Contrary to the column loading case, in batch sorption the bicarbonate counterion is not flushed out of the system leading to anion competition and therefore lower carboxylate loading. The equilibrium pH of the aqueous solutions was in all the cases between 8.1 and 9.0 as bicarbonate was released from the resin. Esters of succinate (control experiment), and acetate were produced quantitatively. Dimethyl fumarate was produced in good yield. Interestingly, malate was converted to dimethyl malate and also to dimethyl fumarate, via

dehydration at the alpha carbon of malate. Table 1: Resin loading and ester yield for relevant carboxylates

Carboxylate Resin loading Ester yield

g/g dry resin Mol ester/mol carboxylate

Succinate 0.16 1 , .00

Acetate 0.11 1 , .00

Fumarate 0.19 0 , .44

Malate 0.18 0 , .10 (0.33 ^a )

"Yield for dimethyl fumarate based on sorbed malate

Conclusion

This example shows the broad potential of the present process. Furthermore it shows that, when applied, the present process can recover relevant carboxylates from aqueous solutions and produce their respective esters in good yields.

Example 4 : Conversion of carboxylates from a

multicomponent mixture simulating a succinate fermentation broth

A typical succinate fermentation broth will contain other carboxylates as a homosuccinate fermentation is rarely encountered. It has been reported that a bioconversion of glucose to succinate using an engineered Corynebacterium glutamicum strain produces, among others, minor quantities of acetate, malate and fumarate. It is expected that these carboxylates will also be sorbed, to a certain extent, on the anion exchange resin and therefore transformed to their respective esters. After transformation, such esters are easily separated (from one and another and from a solvent) by distillation or crystallization to the required purity. Thus, multicomponent sorption capacities and ester formation experiments were performed to quantify the sorption and reaction characteristics of such system.

Experimental

A simulated fermentation broth containing 134 g/L of succinate, 1.2 g/L of acetate, 4.4 g/L of malate and 1.5 g/L of fumarate was prepared. The pH of the simulated broth was adjusted to pH 7.0 by addition of sodium hydroxide. For the sake of demonstration at laboratory scale, the solution was diluted to obtain a simulated broth of 10 g/L of succinate, during this process the other carboxylates were diluted accordingly .

Dynamic sorption experiments were carried out in an adjustable height Omnifit glass chromatography column (10 mm. internal diameter x 300 mm height) . The column was filled using a slurry packing technique and vibrated manually during settling to improve particle distribution. The

experimental set-up was completed by a Shimadzu LC-8A HPLC pump, a Waters 484 tunable UV absorbance detector set at 210 nm, online effluent pH and conductivity detectors and a

Pharmacia FRAC-200 fraction collector. A Dowex MSA strong anion exchange resin in the bicarbonate form was used and a 21 mL bed volume was packed (16 g wet resin equivalent to 5.7 g dry resin, 270 mm height) . After loading the column was washed with ten bed volumes of deionized water and loaded using the above mentioned simulated fermentation broth at 2 mL/min. Fractions were collected at the column outlet for carboxylate concentration analysis. Column

dynamic capacity was calculated using the breakthrough data by means of integral analysis. Concentration of carboxylates was determined by HPLC as described in Example 2.

Ester formation experiments were carried out in a

similar way as described in example 2. In this case, 30 g of dimethyl carbonate was added and 1 g of dry resin loaded in the previous step was held in the solid addition device until reaction temperature (100°C) was reached. Resin was released and reaction started. The reaction was carried out for 10 h. The esters formed were determined by gas

chromatography as described in example 2.

Results

Sorption studies were done using a diluted simulated fermentation broth. Although it was found that the succinate titers are low, the molar ratio between the different

components was kept, resulting in an experiment that is intended to reflect phenomenologically the real concentrated case in terms of component separation.

As ion exchange is the main interaction mechanism

present, selectivity rules according to anion valence apply. Final column capacities are shown in table 2. All the

carboxylates were sorbed according to their concentration and selectivity coefficient. As a consequence, the column capacity towards succinate was reduced as compared to example 2. Although fumarate was present in a lower

concentration than malate, it was sorbed more effectively. These results show that a selectivity order was established as fumarate>succinate≥malate>acetate .

After loading using the simulated fermentation broth, the resin was used in alkylation experiments with dimethyl carbonate as solvent and alkylating agent. Table 2 also shows the obtained ester yields. Dimethyl succinate and methyl acetate were produced in excelent yields. Dimethyl fumarate was produced from sorbed malate and fumarate as described in example 3. Dimethyl malate was not detected.

Table 2: Sorption capacities for succinate and other expected carboxylate by-products and their respective ester yields .

Carboxylate Dynamic column Ester yield

capacity (mol ester/mol

(g carboxylate/g carboxylate )

dry resin)

Succinate 0.20 0.98

Malate 0.01 0.00

Fumarate 0.02 0.16 ^a

Acetate 4.0xl0 ^~4 1.00

aYield for dimethyl fumarate based on combined sorbed malate and fumarate

Conclusion

This example shows that given a simulated fermentation broth in which succinate was the major product, succinate was selectively recovered and effectively converted to dimethyl succinate. Other sorbed carboxylates were also alkylated to their respective esters.

Example 5: FDCA case comparison.

Anion exchange resins can be used in the carbonate or bicarbonate form. If FDCA is produced as sodium dianion, the following exchange will occur.

Q ₂ C0 ₃ + Na ₂ FDCA—* Q ₂ FDCA + Na ₂ C0 ₃

2QHCO , + Na ₂ FDCA —¥ Q ₂ FDCA + 2NaHC0 ₃

It has been found that other anions present in biotransformation broth will elute early, and may be recycled to e.g. the biotransformation with the liberated

Na ₂ CC>3. The carbonate is preferably the base for controlling the biotransformation pH.

Experiments with succinate have shown that succinate binding much stronger than carbonate or bicarbonate binding, thus allowing efficient anion exchange in a column mode.

So far, the resin drying step has been done by simple desiccation, though other options are open, such as use of an alcohol, such as the methanol produced.

In an example the dried, FDCA-loaded resin is in an example heated in the presence of dimethyl carbonate to obtain dimethyl FDCA:

2CH ₃ OCOOCH ₃ + Q ₂ FDCA + 2H ₂ 0 → 2QHC0 ₃ + H ₃ CFDCACH ₃ + 2CH ₃ OH

2CH ₃ OCOOCH ₃ + Q ₂ FDCA → Q ₂ C0 ₃ + H ₃ CFDCACH ₃

CH ₃ OCOOCH ₃ + H ₂ 0 → 2CH ₃ OH + C0 ₂

It has been found that the monomethyl ester cannot desorb from the resin because there are no free cations, and therefore it is not found in the dimethyl carbonate solution. Distillation of excess DMC is easy whereas distillation also allows removal of methanol (if hydrolysis occurs due to traces of water) and purification of dimethyl furandicarboxylate .

Fig. 5 presents a comparison of waste production according to known reaction stoichiometries . Therein Na ₂ CC>3, HMF, 0 ₂ , H ₂ S0 ₄ and glycol are added, whereas C0 ₂ , ag, Na ₂ SC>4, H ₂ 0 and PET are removed, in comparison to the present addition of HMF, 0 ₂ , CO and glycol, whereas C0 ₂ , H ₂ 0 and PET are removed.

In an example, per kg FDCA the proposed method does not produce any waste according to the reaction stoichiometry as currently known. In contrast, the default method that we want to replace produces 0.91 kg sodium sulphate waste per kg FDCA (Fig.5) . Moreover, the present process saves 0.20 $ on chemicals according to the overall process stoichiometry (Fig. 6) . This is considered substantial.

Fig. 6 presents a comparison of main chemicals costs according to known reaction stoichiometries . Therein Na ₂ C0 ₃ , HMF, 0 ₂ , H ₂ S0 ₄ and glycol are added, whereas C0 ₂ , aq, Na ₂ S0 ₄ , H ₂ 0 and PET are removed, in comparison to the present addition of HMF, 0 ₂ , CO and glycol, whereas C0 ₂ , H ₂ 0 and PET are removed.

Only for the differences between the processes, costs are indicated, in $/kg FDCA. Assuming that renewable CO can be used, it is assumed that the extra C0 ₂ produced in the proposed case does not lead to emission costs.

FIGURES

The invention although described in detailed · explanatory context may be best understood in conjunction with the accompanying figures.

Fig. 1 shows details of a schematic sequence of operations for dimethyl succinate production.

Fig. 2 presents succinate sorption equilibrium distribution dependence (isotherms) for the tested counter- ions .

Fig 3. shows equilibrium pH corresponding to the sorption isotherms for succinate determined for each counter-ion.

Fig. 4 shows the DMS yield reaction profile, defined as mole of DMC produced by mole of succinate sorbed.

Fig. 5 presents a comparison of waste production according to known reaction stoichiometries for FDCA.

Fig. 6 presents a comparison of main chemicals costs according to known reaction stoichiometries for FDCA.