1. Patent Search
  2. Details

INTEGRATED SYSTEM FOR BIOCATALYTICALLY PRODUCING AND RECOVERING AN ORGANIC SUBSTANCE

20220275323 · 2022-09-01

    Inventors

    Cpc classification

    International classification

    Abstract

    The invention relates to a method for recovering a biocatalytically produced organic substance from a reaction mixture, comprising—providing a reaction mixture, wherein the organic substance is produced using a biocatalyst, which reaction mixture comprises a substrate for the biocatalyst in a continuous aqueous phase, and wherein further a product recovery phase is present into which the organic substance migrates or onto which the organic substance absorbs or adsorbs; and—separating the product recovery phase comprising the produced substance from the aqueous phase and the biocatalyst. The invention further relates to a bioreactor system for biocatalytically producing a substance, comprising an apparatus, said apparatus comprising a reaction compartment (11) situated in a lower part of the apparatus and a separator compartment (9).

    Claims

    1. A method for recovering a biocatalytically produced organic substance from a reaction mixture, comprising providing a reaction mixture, wherein the substance is produced using a biocatalyst, which reaction mixture comprises an aqueous phase, in which the biocatalyst is preferably dispersed or dissolved, which aqueous phase comprises a substrate for the biocatalyst, and wherein further droplets or bubbles of a fluid product recovery phase or particles of a solid product recovery phase are dispersed in the continuous aqueous phase, into which droplets or bubbles produced substance migrates respectively into or onto which particles of solid product recovery phase produced organic substance adheres; and separating the product recovery phase comprising the produced substance from the aqueous phase and the biocatalyst; wherein the production of the organic substance and the separation of the product recovery phase are carried out in an apparatus comprising a reaction section, containing the reaction mixture wherein the substance is produced, and a separation section wherein the product recovery phase comprising the produced substance is separated from the aqueous phase, wherein the method comprises a simultaneous production and separation stage, wherein at least during said simultaneous stage, substrate and/or product recovery phase is fed into the reaction section continuously or intermittently, flow conditions in the reaction section are turbulent flow conditions, reaction mixture—of which mixture the product recovery phase comprises the produced substance—is fed continuously or intermittently from the reaction section into the separation section, which fed reaction mixture enters said separation section under essentially laminar flow conditions, in which separation section the product recovery phase is separated from the aqueous phase, under essentially laminar flow conditions or intermittently alternating between laminar flow conditions and no-flow conditions, and product recovery phase, comprising the biocatalytically produced substance, is recovered continuously or intermittently from the separation section of the apparatus.

    2. Method according to claim 1, wherein the migration rate, the adsorption rate or the absorption rate of the produced substance into/onto the product recovery phase in the reaction section is at least during said simultaneous production and separation stage maintained at about the same rate as the rate at which produced substance (as part of the separated product recovery phase) is separated from the aqueous phase in the separation section and/or wherein the migration rate, adsorption rate or absorption rate of the produced substance into/onto the product recovery phase in the reaction section is maintained at about the same rate as the rate at which the substance is produced.

    3. Method according to claim 1, wherein the Sauter mean diameter (D[3,2]), of the droplets, bubbles respectively particles of the dispersed product recovery phase in the reaction section is maintained within the range of about 10 to about 250 μm, preferably within the range of about 10 to about 150 μm, more preferably within the range of about 20 to about 80 μm.

    4. (canceled)

    5. (canceled)

    6. Method according to claim 3, wherein the Sauter mean diameter (D[3,2]), of the droplets, bubbles respectively particles of the dispersed product recovery phase in the reaction section is maintained within the range preferably in the range of about 150 to about 250 μm.

    7. Method according to claim 1, wherein at least during said simultaneous production and separation stage the substrate and the product recovery phase are fed into the reaction section at a rate at which the production rate of the substance and the migration rate of the produced substance into the product recovery phase are about the same.

    8. Method according to claim 1, wherein the concentration of the produced substance in the aqueous phase is maintained at a value at which the biocatalyst activity is essentially non-inhibited by the presence of the organic substance, at least during the simultaneous stage and/or wherein the concentration of biocatalyst-inhibiting contaminants originating from the substrate is maintained at a value at which the biocatalyst activity is essentially non-inhibited by the presence of the organic substance, at least during the simultaneous stage.

    9. Method according to claim 1, wherein the product recovery phase is an organic liquid for which the produced substance has a higher affinity than for the aqueous phase and which liquid phase is fed into the reaction section at least during the simultaneous stage, which product recovery phase preferably is a phase for which the produced substance has a partitioning coefficient (i.e. the ratio of the equilibrium concentration of the produced substance in the product recovery phase to the equilibrium concentration of the produced substance in the aqueous phase) of at least 3, more preferably of 5-1000, in particular of 7-250, more in particular of 10-100.

    10. (canceled)

    11. Method according to claim 1, wherein at least during the simultaneous stage a gas is fed in a lower part of the reaction section, which gas may be a product recovery phase, and which gas generates or contributes to an upwards motion of the reaction mixture, whereby reaction mixture flows into a riser situated between the reaction section and the separation section providing a transport channel between said sections and from the riser into the separation section, in which method preferably the product recovery phase is a liquid phase having a lower density than the aqueous phase, wherein downstream of the riser the reaction mixture comprising the product recovery phase, enriched with produced substance, is separated from the gas that generated or contributed to the upwards motion, wherein said reaction mixture separated from said gas is separated in the separation section into an upper layer comprising the product recovery phase enriched with product substance and a lower layer comprising the aqueous phase, including biocatalyst, wherein product recovery phase enriched with produced substance is recovered from said upper layer, and wherein aqueous phase, including biocatalyst, from said lower layer is returned to the reaction section.

    12. (canceled)

    13. Method according to claim 1, wherein the reaction mixture is separated in the separation section into an upper layer comprising the product recovery phase enriched with produced substance and a lower layer comprising the aqueous phase, including if present, biocatalyst dispersed or dissolved in the aqueous phase, wherein product recovery phase enriched with produced substance is recovered from said upper layer, and wherein aqueous phase—including, if present, biocatalyst dispersed or dissolved in the aqueous phase—from said lower layer is returned to the reaction section, and wherein a gas is fed, e.g. sparged, into the reaction mixture in the separation section at a position below the interface between the upper and the lower layer.

    14. Method according to claim 1, wherein the produced organic substance is selected from the group consisting of hydrocarbons, in particular monoterpenes, sesquiterpenes, aromatic hydrocarbons; isoprenoids (terpenoids); organic acids, in particular C5-C24 fatty acids, more in particular C12-C20 fatty acids; alcohols, in particular alcohols having at least 4 carbon atoms; ketones, in particular ketones having at least 5 carbon atoms; aldehydes in particular aldehydes having at least 5 carbon atoms; cyclic carboxylic esters, in particular lactones; non-cyclic esters, in particular non-cyclic esters having at least 5 carbon atoms; lipids in particular glycerides; amines; amino acids; and peptides.

    15. Method according to claim 1, wherein the biocatalyst comprises a living organism and the produced substance is secreted into the aqueous phase or wherein the biocatalyst comprises an isolated enzyme or combination of isolated enzymes dispersed or dissolved in the aqueous phase or wherein the biocatalysts is an isolated enzyme or combination of isolated enzymes immobilized on one or more support materials dispersed in the aqueous phase.

    16. Method according to claim 1, wherein the biocatalyst comprises a micro-organism selected from the group of bacteria, archaea and fungi, preferably selected from the genera Pseudomonas, Gluconobacter, Rhodobacter, Clostridium, Escherichia, Paracoccus, Methanococcus, Methanobacterium, Methanocaldococcus, Methanosarcina, Aspergillus, Penicillium, Saccharomyces, Kluyveromyces, Pichia, Candida, Hansenula, Bacillus, Corynebacterium, Blakeslea, Phaffia (Xanthophyllomyces), Yarrowia, Schizosaccharomyces, Zygosaccharomyces more preferably from the group of Corynebacterium glutamicum, Escherichia coli, Bacillus subtilis, Bacillus methanolicus, Pseudomonas aeruginosa, Pseudomonas putida, Rhodobacter capsulatus, Rhodobacter sphaeroides, Paracoccus carotinifaciens, Paracoccus zeaxanthinifaciens, Saccharomyces cerevisiae, Saccharomyces pastorianus, Schizosaccharomyces pombe, Aspergillus nidulans, Aspergillus niger, Aspergillus oryzae, Blakeslea trispora, Penicillium chrysogenum, Phaffia rhodozyma (Xanthophyllomyces dendrorhous), Pichia pastoris, Yarrowia lipolytica, in particular from the group of Escherichia coli, Pseudomonas aeruginosa, Pseudomonas putida and Saccharomyces cerevisiae

    17. Method according to claim 1, wherein the product recovery phase is a hydrophobic liquid, preferably a hydrophobic liquid comprising or consisting of one or more hydrophobic liquids selected from the group alkanes, e.g. dodecane, and triglycerides, e.g. vegetable oils or wherein a gas is used as a product recovery phase for a gaseous or volatile produced substance.

    18. Bioreactor system for biocatalytically producing a substance, comprising an apparatus, said apparatus comprising a reaction compartment situated in a lower part of the apparatus and a separator compartment, a riser defining a channel at, near or directly above the top of the reaction compartment adapted to allow fluid from the reaction compartment to flow upward, a downcomer defining a channel between the outlet side of the riser and the inlet side of the separator compartment, adapted to allow (non-gaseous) fluid leaving the riser to flow downward into the separator compartment, the reaction compartment comprising an agitator, preferably a stirrer, a feed inlet for a substrate for use in the production of the substance, an inlet for a product recovery phase, which is preferably positioned in a middle part or bottom part of the reaction compartment, more preferably closer to the bottom of the reaction compartment than the feed inlet for substrate, an inlet for a gas phase, preferably a sparger which is typically positioned closer to the bottom of the reaction compartment than the feed inlet for substrate, which is typically positioned closer to the bottom of the reaction compartment than the inlet for the product recovery phase, the separator compartment comprising an outlet for product recovery phase, typically positioned closer to the top of the separator compartment than the outlet end of the downcomer, a recycle provision for recycling biocatalyst (typically fluid containing biocatalyst)) taken from a part of the separator compartment below the inlet for the gas phase to the separation compartment, and the apparatus having a headspace provided with an outlet for gas phase introduced into the apparatus via the inlets for gas phase.

    19. Bioreactor system according to claim 18, wherein the reaction compartment is positioned below the separator compartment and both compartments are on opposite sides of a partition and the recycle provision comprises one or more openings in the partition, which partition preferably is tilted and wherein one or more openings are present at or near the lowest point of the partition.

    20. Bioreactor system according to claim 18, wherein the separator compartment comprises an inlet for a gas phase, preferably a micro-bubble sparger which is typically positioned closer to the bottom of the separator compartment than the outlet end of the downcomer.

    21. Bioreactor system according to claim 18, further comprising a bioreactor vessel having an outlet for a fluid (comprising substrate and optionally biocatalyst and/or produced organic substance) that is connected via a fluid channel with the feed inlet of the reaction compartment of said apparatus.

    22. Process for isolating a biocatalytically produced organic substance from a liquid product recovery phase, comprising a first extraction, wherein the organic substance is extracted from an aqueous reaction medium—wherein the organic substance has been produced in the presence of biocatalyst catalyzing the production of the organic substance—into the liquid product recovery phase, which liquid product recovery phase forms a separate phase when contacted with the reaction medium, which liquid product recovery phase has a boiling point of at least 100° C., preferably of 150-300° C.; a separation of the liquid product recovery phase, containing the organic substance from the reaction medium; a second extraction, which is called a back-extraction, wherein the produced organic substance is extracted from the product recovery phase into a back-extraction liquid, different from the product recovery phase, having a boiling point of less than 100° C.; a separation of said extraction liquid containing the produced organic substance from the product recovery phase; and an isolation of the organic substance from said extraction liquid.

    23. Process according to claim 22, wherein the liquid product recovery phase and the back-extraction liquid are organic liquids, in which process preferably the liquid product recovery phase is an organic liquid selected from the group consisting of fatty acids or esters thereof, waxes, primary alcohols, oils (e.g. vegetable oils) and alkanes having a boiling point of 100° C. or more and/or wherein the back-extraction liquid is selected from the group consisting of alcohols (mono-, di or polyol), ethers and esters having a boiling point of less than 100° C.

    24. (canceled)

    25. Process according to claim 22, wherein the liquid product recovery phase has an aqueous solubility of less than 85 g/l, preferably of 0 to 1 g/L.

    26. Process according to claim 22, wherein the product recovery phase and the back-extraction liquid are brought in contact with each other at a temperature and in a ratio at which they form separate phases and wherein the produced organic substance is extracted from the product recovery phase into the back-extraction liquid, which product recovery phase is preferably essentially insoluble in the back-extraction liquid and/or which back-extraction liquid is preferably essentially insoluble in the product recovery phase; or wherein the product recovery phase and the back-extraction liquid are brought in contact with each other at a temperature and in a ratio at which they form a mixed phase, after which the temperature is changed to induce phase separation, whereby a back-extraction liquid phase is obtained enriched in the biocatalytically produced organic substance.

    27. (canceled)

    28. Process according to claim 22, wherein the product is isolated from the back-extraction liquid by distillation, or wherein the product isolated by precipitation, in particular by crystallization, e.g. by reducing the temperature to a temperature at which the organic substance solidifies, whilst the back-extraction liquid remains liquid.

    29. Process according to claim 22, wherein the reaction medium is a reaction mixture, wherein the organic substance is produced using a biocatalyst, which reaction mixture comprises an aqueous phase, in which the biocatalyst is preferably dispersed or dissolved, which aqueous phase comprises a substrate for the biocatalyst, and wherein further droplets of the liquid product recovery phase are dispersed in the continuous aqueous phase, into which droplets produced substance migrates; and separating the product recovery phase comprising the produced substance from the aqueous phase and the biocatalyst; wherein the production of the organic substance and the separation of the product recovery phase are carried out in an apparatus comprising a reaction section, containing the reaction mixture wherein the substance is produced, and a separation section wherein the product recovery phase comprising the produced substance is separated from the aqueous phase.

    30. Process according to claim 29, wherein the process comprises a simultaneous production and separation stage, and wherein at least during said simultaneous stage substrate and/or product recovery phase is fed into the reaction section continuously or intermittently, flow conditions in the reaction section are turbulent flow conditions, reaction mixture—of which mixture the product recovery phase comprises the produced substance—is fed continuously or intermittently from the reaction section into the separation section, which fed reaction mixture enters said separation section under essentially laminar flow conditions, in which separation section the product recovery phase is separated from the aqueous phase, under essentially laminar flow conditions or intermittently alternating between laminar flow conditions and no-flow conditions, and product recovery phase, comprising the biocatalytically produced substance, is recovered continuously or intermittently from the separation section of the apparatus.

    31. (canceled)

    Description

    EXAMPLES

    Example 1: Anaerobic Production of ABE (acetone-butanol-ethanol)

    [0149] ABE (acetone-butanol-ethanol) was catalytically produced by Clostridia beijerinckii (NCIMB 8052) using technical oleyl alcohol as the product recovery phase (solvent). The microorganism strain was kindly provided by Food and Biobased Research, Wageningen University and Research Centre. The strain was obtained from the NCIMB collection.

    [0150] The following medium was prepared for the experiment:

    TABLE-US-00001 TABLE 1 Medium prepared for the experiment Solution/medium Total amount Preculture 1 medium (CM2) 0.3 kg Preculture 2 medium (CM2) 3.5 kg Batch medium (CM2) 70 kg Feed medium (glucose, NH.sub.4acetate) 75 kg Yeast extract stock 0.5 kg Iron sulfate solution (1000X)* 0.1 kg Oleyl alcohol (technical) 200 kg Base (2M KOH) 2 kg Antifoam 204 1 kg
    The CM2 medium has the following composition:

    TABLE-US-00002 TABLE 2 CM2 medium composition CM2 medium components Final concentration K.sub.2HPO.sub.4•3H.sub.2O 0.8 g/L KH.sub.2PO.sub.4 1 g/L NH.sub.4acetate 2.9 g/L p-aminobenzoic acid (PABA) 0.1 g/L MgSO.sub.4•7H.sub.2O 1 g/L FeSO.sub.4•7H.sub.2O solution (1000X) 1 mL Antifoam 204 7 mL Sterile additions Glucose 60 g/L Yeast extract 2.5 g/L

    [0151] The medium was autoclaved at 121° C. for 20 min except for glucose and yeast extract (YE Duchefa Biochemie prod. Y1333.0500) solutions. These solutions were autoclaved separately at 110° C. for 20 min and at 121° C. for 20 min respectively and added sterilely afterwards. The pH of the CM2 medium was adjusted to 6.3. The iron solution was prepared separately as 1000 concentrated solution dissolved in 0.1 M HCl and then, 1 mL of the iron stock solution was added to the rest of the CM2 medium. For medium preparations, it can be also referred to Diallo, M., Simons, A. D., van der Wal, H., Collas, F., Houweling-Tan, B., Kengen, S. W., & López-Contreras, A. M. (2019). L-Rhamnose metabolism in Clostridium beijerinckii strain DSM 6423. Appl. Environ. Microbiol., 85(5), e02656-18.

    [0152] The feed medium consisted of 330 g/L glucose and 1 g/L NH.sub.4acetate that was autoclaved at 110° C. for 20 min and sparged with N.sub.2 in order to remove the oxygen. The product recovery phase oleyl alcohol (Chempri, Raamsdonkveer) was not sterilized. The solvent was colored with Oil Red O at a concentration of 75 mg/kg.sub.solvent.

    [0153] The 6 anaerobic jars for preculture 1 had each a butylrubber on top and an aluminum cap. The jars were filled with 5 mL demi water and flushed with N.sub.2 gas for 10 min. Afterwards the jars were filled with 50 mL CM2 medium. Just before inoculation, the jars were flushed with N.sub.2 for 10 min again.

    [0154] 6 vials of 1 mL frozen culture stocks were heat shocked for 75 s at 100° C. in a water bath and immediately cooled for 1 min under a running tap. The heat shocked cells were then transferred to the sterile anaerobic jars with 50 mL CM2 medium. A bike valve equipped with a 0.2 μm sterile filter was placed into the rubber cap of the anaerobic jar to release the gases formed during the incubation. Preculture 1 was incubated at 37° C. and 0 rpm for 24 h.

    [0155] After incubation of preculture 1, 180 mL of preculture 1 was added to 3.32 L CM2 medium in a 3.5 L bioreactor. The anaerobic bioreactor was heated to maintain a temperature of 37° C., flushed with N.sub.2 gas and incubated for 24 h without any stirring or pH control.

    [0156] The integrated bioreactor was prepared as followed: the pH probe was polarized and calibrated offline (2-points calibration with buffer solutions of pH 4.0 and 7.0). An empty sterilization was performed at 125° C. for 30 min. The batch medium (without yeast extract and glucose) was prepared and pumped into the reactor. A full sterilization of the reactor is subsequently performed at 121° C., 20 minutes. Only after full sterilization and cooling of the reactor, the batch sterile solutions (yeast extract and glucose) are pumped into the reactor. Nitrogen gas flowed into the reactor for at least 30 min to make the medium anaerobic. The fermentation controllers were set in the control unit eZ-control. If the pH drops below 5.2 after 24 h after inoculation, the pH is controlled with base. After the batch phase, the stirring speed is increased from 0 rpm to 300 rpm.

    [0157] The stirring speed was suitable to maintain turbulent flow conditions in the fermentation reactor and D[3,2] at a value in the range of about 10 to about 100 μm.

    TABLE-US-00003 TABLE 3 Operation conditions of the experiment during the batch phase and the continuous phase Parameter Value Unit Temperature 37 ° C. Vessel pressure  0.2 Barg N.sub.2 flow 25-35 L. min.sup.−1 pH 6.3 to 5.2 — pH control Only base after 24 h — Stirring Batch phase: 0 rpm Continuous phase: 250-300 Anti-Foam cycle Foam: on contact Yes

    [0158] The batch phase ended 17 h after inoculation when the glucose concentration was close to 0 g/L and the continuous feed of glucose was started at 0.25 kg/h.

    [0159] The feed rate was adjusted during the continuous phase as following:

    TABLE-US-00004 TABLE 4 Adjustments of the feed rate during the continuous phase Time after inoculation [h] Feed rate [kg/h] 23.0 0.4 24.8 1.2 28.6 1.0

    [0160] The feed rate was adjusted because the glucose drop after the batch was almost to 0 g/L. The glucose concentration was increased over 20 g/L. The feed rate was then decreased stepwise to maintain a minimum glucose concentration of 20 g/L.

    [0161] Moreover, the continuous solvent addition was set to 0.6 kg/h after the end of the batch phase until 24 h after inoculation and set to 1.3 kg/h till the end of the fermentation. The harvest was started 22 h after inoculation. The recirculation loop was opened to the maximum since the beginning of the fermentation. Flow conditions in the separator compartment were kept non-turbulent (laminar flow)

    [0162] Results and process details are shown in FIGS. 20-27.

    [0163] FIG. 20 shows the development of pH during the fermentation.

    [0164] FIG. 21 shows the optical density (OD) values of the batch and continuous phase.

    [0165] FIG. 22 shows the measured recirculation flow during the batch and continuous phase

    [0166] FIG. 23 shows the N.sub.2 gas flow in and liquid level measurement of the separation compartment.

    [0167] FIG. 24 shows the average feed rate during the continuous phase after the batch ended (black line)

    [0168] FIG. 25 shows solvent (recovery phase) addition during the continuous phase.

    [0169] FIG. 26 shows the organic fraction of the harvest (recovered phase)

    [0170] FIG. 27 shows the butanol concentration in the aqueous phase (caqu) and in the organic phase (corg). The butanol concentration increased during the batch phase to approximately 6 g/L. With the start of the recovery phase addition, a lot of the butanol was extracted into the recovery phase. The productivity of the microorganism decreased after the batch phase which is why the butanol concentration is decreasing during the continuous phase.

    Example 2: Recovery of an Aerobically Produced Hydrocarbon

    [0171] A hydrocarbon (the sesquiterpene humulene) was fermentatively produced by a genetically modified E. coli, using castor oil as product recovery phase in an integrated bioreactor system (apparatus) as schematically shown in FIG. 1, having a volume of 100 L. Such E. coli and the use thereof in the fermentative production as such are generally known in the art. See also Semra Alemdar et al eng. Life Sci 2017, 17, 900-907 (https://onlinelibrary.wiley.com/doi/full/10.1002/elsc.201700043).

    [0172] LB medium (Luria Bertani medium) was autoclaved and glucose was added sterile with a final concentration of 10 g/L. The preculture was inoculated with 1.2 mL cells from cryogenic frozen culture stocks in 500 mL LB medium in a flat bottom flaks with baffles. 5 flasks were prepared and the preculture was incubated for 8 h at 30° C. and 250 rpm.

    [0173] Before the LB medium was filled into the 100 L integrated bioreactor system, the pH and DO (dissolved oxygen) probes were calibrated. Afterwards an empty sterilization of the reactor system was performed for 30 min at 125° C. and the LB medium was filled into the fermentation compartment of the reactor system. A full sterilization of the reactor was subsequently performed for the sterilization of the LB medium (121° C., 20 minutes). Only after full sterilization and cooling of the reactor, the batch sterile solutions were pumped into the reactor. The fermentation controllers: pressure, stirring, air flow, temperature, foam and pH should be started in the eZ-control unit (control system of the reactor) with indicated fermentation set points (Table 4). Afterwards, the DO calibration (2-point calibration) can be performed.

    [0174] The fermentation compartment of the bioreactor system was supplied with aqueous fermentation broth, comprising substrate predominantly at carbon limiting conditions) and the E. coli. This supply was based on standard procedure, with the proviso that additionally product recovery phase (30 to 40 g/kg castor oil) was dispersed in the aqueous medium.

    TABLE-US-00005 TABLE 5 Initial set points of the batch phase Parameter Value Unit Vessel pressure 0.3 Barg Air flow 47 L/min 1 vvm (in the ferm. compart.) DO (fed-batch) >30 % Stirring Batch: 700 rpm Fed-batch: 400-700 rpm Anti-Foam controlled by sensor on contact — Cycle time: 3 S Dead time: 15 S

    [0175] During the batch phase the cells grew exponentially until the initial amount of glycerol is consumed. The oxygen saturation started to drop from 100% after about 1 hour. Just before the end of the batch, the DO decreased close to 0% shortly and increased abruptly to 100% at the end of the phase. At the end of the batch phase, the pH increased due to consumption of secondary metabolites, therefore pumping of acid was required. The foam was controlled by the foam sensor according to the defined cycle.

    [0176] The fed-batch phase began when all substrate of the batch was consumed and oxygen saturation increased up to 100% starting with an exponential feed phase (see also FIG. 9 for details on accumulated feed addition throughout the fermentation). During this phase, the separation compartment also gradually became filled with fermentation broth. During this phase, the cells grew at an exponential rate but still under carbon limiting conditions. The next feed phase was at a constant rate, to avoid further oxygen limitations and to match a desired biomass concentration set point in the reactor. When the feed rate reached the desired set point during the exponential feed phase (initial 200 g/h), the software transited automatically to the constant feed phase regime and maintained the current feed set point (450 g/h). Both the fermentation and separation compartment were full of fermentation broth from the beginning of the fermentation. Therefore, the recirculation loop was open (maximum) since the beginning of the fermentation to ensure biomass recirculation, enough nutrients supply and to prevent oxygen limitation. The level control loop—which was control based on the height of broth in the separation compartment—was started after exponential feed. The height set point was defined as the current measured value which was expected between 700 and 800 mm. From that point onwards, the bleed rate (broth drawn from the recirculation loop) should match the feed rate, so that the level in the reactor was maintained.

    [0177] The continuous addition of the product recovery phase, castor oil, to the fermentation compartment of the reactor was started at 37 hours after inoculation.

    [0178] Product recovery phase, containing the humulene was formed as a top layer in the separation compartment; this was removed intermittently from the separation compartment, without using gas injection into the separation compartment. Flow conditions were laminar inside the separation compartment. The bottom layer (aqueous phase comprising E coli) was partly recycled and partly bled.

    [0179] FIGS. 2-10 provide information about the flow rates and results of the experiment.

    [0180] FIG. 2 shows dissolved oxygen (DO) profile during the fermentation. The grey vertical lines show subsequent starts of changes in the process parameters. From left to right (second from the top to bottom in the legend:

    [0181] Start Exp. Feed=start exponential feed phase

    [0182] Start Const. Feed=constant feed phase

    [0183] Cont. Oil add.=start of the continuous addition of castor oil, decrease feed rate

    [0184] FIG. 3 shows the product concentration in organic and aqueous phase. The product concentration is highest in the organic phase. So the transfer from the aqueous phase to the organic phase was successful and the organic phase was enriched with product compared to the aqueous phase.

    [0185] The concentration of humulene in the aqueous phase increased only slightly during the fermentation while the concentration in the organic phase remained constant. A semi steady state was achieved during the fermentation, sufficient to keep the concentration of the humulene at a non-inhibiting level.

    [0186] FIG. 4 shows the amount of product recovery phase (solvent for the product) added to the reaction section and recovered from the separation section (harvest). The partitioning coefficient depends on the concentration in both phases. The partitioning does not necessarily scale linear with the concentrations. That is why the aqueous product concentration can increase slightly while the organic concentration remains constant. The harvest of organic phase from the separation section occurred intermittently while the solvent addition was continuously.

    [0187] FIG. 5 shows the stirring speed during the fermentation. The stirring speed was suitable to maintain turbulent flow conditions in the fermentation reactor and D[3,2] at a value in the range of about 10 to about 100 μm.

    [0188] FIG. 6 shows the cell dry weight (CDW) during the fermentation in the reaction medium inside fermentation compartment and in the organic phase formed by the castor oil plus product (cream, top layer in the separation compartment) from separation section. The cell biomass was mainly in the reaction section and bleed. The biomass was not found or in a very small concentration found in the cream in the separation section. So the harvested organic phase was considered to be free of cells which is beneficial for further DSP. Moreover, the cells were not lost with the harvest but recycled to the reaction section (fermentation compartment).

    [0189] FIG. 7 shows the fraction of the organic phase in the different compartments (the content of organic phase in the reaction compartment was 1-4%, which was at a desired level).

    [0190] FIG. 8 shows the recovery rate expressed as cream recovery rate from the separation section.

    [0191] FIG. 9 shows the total amount of aqueous solution of substrate added during the continuous phase.

    [0192] FIG. 10 shows the air flow set point and recirculation flow during the fermentation.

    Example 3: Determination of the Product Recovery Particle Size

    [0193] Santalene was produced by a genetically modified E. coli using dodecane as product recovery phase under turbulent flow conditions in the fermentation compartment.

    [0194] The fermentation compartment was a CSTR (volume of 7 l). The pH and DO probe were calibrated before the fermentation. At the start sterile batch medium was pumped into the sterile fermentation compartment. The batch medium composition was as in Example 2. The batch medium had a volume of 3 L. Glycerol was the carbon source (substrate). The product recovery phase, dodecane, was added 24 h after inoculation of the reactor, after which it was intermittently withdrawn, see also FIG. 12 for the dodecane balance over time. The glycerol feed was started when the DO value reached >15% after a drop to almost 0%.

    Droplet Size

    [0195] This method is generally applicable to determine D[3,2] in a method according to the invention, in particular for liquid recovery phase.

    [0196] During the fermentation, images of the product recovery droplets were recorded by a SOPAT probe (SOPAT Gmbh; https://SOPAT.de) which consisted of a probe that could be placed into the fermentation vessel coupled to a computer system. To provide sufficient lighting for the pictures a back lighting was applied. The image analysis software was capable of dealing with high oil fractions and the presence of cells in the fermentation broth.

    [0197] Droplet measurements by SOPAT were conducted at the end of a fermentation run in a non-sterile way with no overpressure on the system. The SOPAT probe was put through the headplate into an open position of a push-valve. Both mirror and lens of the SOPAT were covered with a thin layer of Rain-X Rain repellent.

    [0198] To clean the lens, the probe had to be removed from the reactor and placed back again. Cleaning of the probe was performed with 70% ethanol wipes and rinsed afterwards with MilliQ water.

    [0199] For in situ image acquisition three hours online measurements were taken. Every 3 minutes, 30 pictures were acquired every five minutes, starting at 5 min and ending at 175 min. After the experiment, the accompanying particle detection software was used to measure the size of the droplets in the pictures provided by SOPAT Gmbh, see also Maaß, S., Rojahn, J., Hänsch, R., Kraume, M., (Computers & Chemical Engineering. 45. 27-37. 10.1016/j.compchemeng.2012.05.014). “A MATLAB® based image recognition algorithm has been implemented to automatically count and measure particles in multiphase systems.

    [0200] A given image series is pre-filtered to minimize misleading information. The subsequent particle recognition consists of three steps: Pattern recognition by correlating the pre-filtered images with search patterns, pre-selection of plausible drops and the classification of these plausible drops by examining corresponding edges individually.”

    [0201] During a fermentation, air bubbles are present in the mixture interfering with the image analysis. These false positives were eliminated by setting a maximum particle size of 200 μm which was validated by manual removal of the air bubbles. The detected droplets were converted to size distributions and values for the Sauter mean diameter. The number of pictures was not sufficient to have more than 1000 droplets per data point as you statistically would have needed.

    [0202] The Sauter mean diameter D[3,2] (a.k.a. d.sub.32) is calculated with (d.sub.32=Σd.sub.i.sup.3/Σd.sub.i.sup.2).

    [0203] Results of the droplet size measurement are shown in the following Table.

    TABLE-US-00006 TABLE 6 Results of the measurement of the droplet size average Particles/Folder 2220 total Particle number 2220 (2220) scale factor [μm/Pixel] 1 Single Output Values Parameter mean Meaning dg [μm] 31.35 (Geometric Mean) d1, 0 [μm] 45.14 (Arithmetic Mean) d2, 0 [μm] 54.86 (Number-Surface Mean) d2, 1 [μm] 66.67 (Length-Surface Mean) d3, 2 [μm] 74.06 (Sauter Mean) d4, 3 [μm] 76.89 (De Brouckere Mean)

    Relation of Droplet Size and Recovery Rate

    [0204] The droplet size determines how fast the droplets can rise in the separation section and thus determine also the recovery rate. The rising velocity v.sub.c of the droplets can be calculated with Stokes law:

    [00001] v c = ( ρ oil - ρ water ) .Math. g .Math. d d 2 18 .Math. μ

    with the droplet diameter d.sub.d and viscosity μ and density ρ, see also FIG. 11.

    TABLE-US-00007 TABLE 7 Assumptions for the theoretical recovery rate in a 100 L pilot system Parameter Value Unit Density dodecane 750 kg/m.sup.3 Density water 1000 kg/m.sup.3 Gravitational constant g 9.81 m/s.sup.2 Liquid viscosity water 0.001 Pa .Math. s Liquid viscosity dodecane 0.001344 Pa .Math. s Area separation 0.025 m.sup.2 Pressure, temp. (1 bar, 298K)

    Example 4: Production and Recovery of Humulene in a Bioreactor System, Comprising an Additional Fermentor Upstream of the Apparatus with Fermentor Compartment and Separator Compartment

    [0205] The microorganism, product and solvent were as for Example 2: genetically modified E. coli, humulene, castor oil.

    [0206] The setup of the experiment was as followed: An apparatus according to the invention (having a bottom fermentation compartment and a top separation compartment, in connection via a centrally positioned riser surrounded by a downcomer and an internal recycle (not shown in graph) was connected to a big 1 m.sup.3 fermenter (called 10R10) so that liquid could be exchanged between the big fermenter and the reaction section of the integrated reactor system, as schematically shown in FIG. 13. The experiment was divided into two stages—batch and continuous stage.

    [0207] The fermentation started with the batch phase in the 1 m.sup.3 reactor. The operational conditions are shown in the following Table.

    TABLE-US-00008 TABLE 8 Operational conditions in the 1 m.sup.3 fermenter Parameter Value Unit Vessel pressure 0.4 barg Air flow 1 vvm DO (constant feed) >30 % Stirring DO Cascade (stirring only) rpm Min 100 Antifoam Yes: on contact

    [0208] During the batch phase the cells grew exponentially until the initial amount of glycerol was consumed. The oxygen saturation began to drop from 100% after about 1 hour. Just before the end of the batch, the DO decreased close to 0% shortly and then increased abruptly to 100%. This happened about 12 to 15 hours after the inoculation. The continuous phase began when all substrate was consumed and oxygen saturation increased up to 100%. The feed addition was activated and the start constant feed rate was 0.93 kg/h. The cascade DO control was set to min. 24% and controlled with the stirrer speed. There was no overpressure or feed rate or air flow rate adjustment.

    [0209] After the end of the batch phase, the integrated reactor/separator was filled with broth of the 1 m.sup.3 reactor. The feed addition (350 g/h) and the addition of product recovery phase (oil/solvent addition) at a fixed rate for 0.47 kg/h for 42 hours was started (see also FIG. 19). The operational conditions for the integrated reactor is shown in the following Table.

    TABLE-US-00009 TABLE 9 Operational conditions of the integrated reactor (100 L) during the continuous phase Parameter Value Unit Vessel pressure 0.5 barg Air flow (bottom) (MAX) 25 L/min 0.5 vvm (in the ferm. compart.) DO >30    % Stirring Fed-batch: 400-700 rpm Anti-Foam cycle Foam: on contact Yes Always controlled by sensor Cycle time: 3 s Dead time: 15 s

    [0210] The recirculation loop between fermentation compartment and separation compartment was opened to the maximum once the integrated reactor was filled with broth.

    [0211] The results of the integrated reactor are shown in FIGS. 14-20.

    [0212] FIG. 14 shows feed added to the integrated reactor during the continuous phase

    [0213] FIG. 15 shows the DO profile in the fermentation compartment of the integrated apparatus

    [0214] FIG. 16 shows the development of the cell dry weight (CDW) in the 1 m.sup.3 fermenter and the integrated reactor. The CDW concentration was the same in the 1 m.sup.3 fermenter and the integrated reactor system (reaction system). There was no concentration gradient and there was a homogeneous exchange of liquid between the two reactors.

    [0215] FIG. 17 shows the product concentration in the aqueous phase (1 m.sup.3 fermenter, reaction section—of the integrated reactor system) and the organic phase (in the separation section, i.e. harvest section, of the integrated reactor system) during the continuous phase. The product was enriched in the organic phase and successfully transported to the separation section of the integrated reactor system.

    [0216] FIG. 18 shows the recirculation flow from the separation compartment to the reaction compartment of the 100 L apparatus during the continuous phase.

    [0217] FIG. 19 shows the amount of organic phase (castor oil, ‘solvent’) added to and organic phase harvested from the integrated reactor system during the continuous phase. After the filling of the integrated reactor/separator and dosing of the solvent, the apparatus was able to separate the phases quickly without a long lag phase. The total amount of solvent was more than twice as much as compared to Example 2. So the fermentation capacity was increased by connecting the integrated reactor system to the 1 m.sup.3 fermenter while the area for phase separation remained the same. It was proven that the given size of the separation section can cope with bigger amounts of fermentation capacity.

    Example 5: Soybean Oil as Recovery Phase

    [0218] This experiment was performed similarly as Example 3 except that the product recovery phase was only added to the 100 L integrated reactor/separator apparatus. At the end of the fermentation, high solvent addition rates and the respective recovery rates were investigated. The solvent for this test was soybean oil. This lead to the higher recovery rate measured: 4.4 L/h or 172.7 L/m2/h.

    Example 6: Effect of Recovery Phase

    [0219] An experiment was performed similar to Example 2 but with an alkane (dodecane) as solvent, instead of a triglyceride oil (castor oil). Using dodecane as recovery phase was effective but resulted in lower recovery rates, see FIGS. 28 and 29.

    [0220] FIG. 28 shows a high recovery rate when castor oil was used. The product containing phase was a cream with very high organic fraction [0.77-0.83]. FIG. 29 shows that the recovery rate with dodecane was lower. The net rate difference (kg/h) is almost a factor of 5. The droplets distribution of dodecane is significantly smaller, among others due to surface tension and viscosity difference compared to castor oil. The product containing (dodecane) phase was a cream also containing fractions of clear oil with a fraction of organics of 0.6-0.72.

    Example 7 butanol Production with Addition of oleyl alcohol as Recovery Phase During the Batch Stage

    [0221] The setup of the experiment was similar to Example 4 except for the initial batch medium and solvent addition. The initial batch medium was 30 L and 29 kg of liquid recovery phase (solvent; techn. oleyl alcohol) was added during the batch stage.

    [0222] FIG. 30 shows the butanol concentration in the organic phase in the fermentation compartment and in the separation compartment. The maximum concentration of butanol possible in this set up at steady state was about 18 g/L in the organic phase.

    [0223] Theoretical steady states for butanol production in a 100 L integrated reactor system according to the invention, as schematically shown in FIG. 1:

    TABLE-US-00010 TABLE 10 Assumptions made to estimate steady state points for a theoretical butanol production via fermentation. N.sub.X - biomass, q.sub.P - specific production rate, P - product. Assumptions Value Solvent partitioning m 4 Solvent partitioning m 10 Solvent partitioning m 15 Solvent addition [kg/h] 6.048 q.sub.P [kg.sub.P/kg.sub.X/h] 0.41 N.sub.X [kg] 2.5 V.sub.ferm [L] 100 c.sub.inhibition [g/L] 20

    [0224] FIG. 31 shows the effect of product inhibition on the microorganism's expected productivity and thus the effect of ISPR on the productivity. For the three different solvents shown (each with their own butanol partitioning (m=4,10,15), three different steady states can be calculated, indicated by the dotted circles. P—product, aqu—aqueous, m—partitioning of product in solvent and aqueous phase, in this example.

    Example 8: Recovery of vanillic acid

    [0225] Vanillin was catalytically converted into vanillic acid by S. cerevisiae CEN.PK.113-7D. Technical oleyl alcohol was used as the primary product recovery phase (solvent) during cultivation and it was coloured with 75 mg/kg.sub.solvent Oil Red O to visualize the solvent. The integrated bioreactor was prepared similar to example 1, except with fermentation air as aeration medium and the batch medium was SMD2 medium.

    [0226] A preculture was performed in 5 shake flasks with baffles with 100 mL YPD medium (10 g/L Bacto-yeast extract, 20 g/L Bacto-peptone, 20 g/L glucose; e.g. BD, Difco™& BBL™manual, 2.sup.nd edition https://legacy.bd.com/europe/regulatory/Assets/IFU/Difco_BBL/242820.pdf) each. Each flask was inoculated with 1.8 mL frozen stock culture and incubated at 30° C. at 175 rpm for 16 h. 115 mL of the preculture was transferred to the integrated bioreactor. The integrated bioreactor was filled with 70 kg of SMD2 medium (Tables 11-13).

    TABLE-US-00011 TABLE 11 Batch medium composition Batch medium SMD2 Final concentration (NH.sub.4).sub.2SO.sub.4 5 g/L KH.sub.2PO.sub.4 3 g/L MgSO.sub.4•7H.sub.2O 0.5 g/L Bacterial Yeast Extract 5 g/L Sterile additions Glucose monohydrate 30 g/L Trace elements (1000x stock) 1 mL/L Vitamin solution (100x stock) 10 mL/L

    TABLE-US-00012 TABLE 12 Composition of the trace element solution which was adjusted to pH 6.7 with 5M NaOH. The solution was sterilized at 110° C. for 20 min and stored in the fridge wrapped in aluminium foil. Trace element solution Concentration C.sub.6H.sub.6FeO.sub.7 (Fe(III)citrate) 100 g/L CoCl.sub.2•6 H.sub.2O 4 g/L MnCl.sub.2•4 H.sub.2O 23.5 g/L CuCl.sub.2•2 H.sub.2O 1.97 g/L H.sub.3BO.sub.3 5 g/L Na.sub.2MoO.sub.4•2 H.sub.2O 4 g/L Zn(CHCOO).sub.2•2 H.sub.2O 16 g/L EDTA 8.4 g/L

    TABLE-US-00013 TABLE 13 Composition of the vitamin solution which was filter sterilized. Vitamin solution Concentration D-Biotin 5 g/kg Ca D(+) pantothenate 100 g/kg Nicotinic acid 100 g/kg Myo-inositol 25 g/kg Thiamine chloride 100 g/kg Pyridoxol hydrochloride 100 g/kg p-Aminobenzoic acid (paba) 20 g/kg Riboflavin 50 g/kg

    TABLE-US-00014 TABLE 14 Composition of the feed medium Feed medium Final concentration Glucose monohydrate 120 g/L Yeast extract 5 /L

    [0227] The cultivation in the bioreactor was performed in 3 stages. The first stage was a 16 h batch fermentation during which no solvent, nor feed, nor vanillin solution was added. The second stage (16 h to 43 h) was started with the feed addition (composition in Table 14) at 500 g/h, a solvent addition at rate of 0.34 kg/h and vanillin solution (10 g/L) addition at 0.23 kg/h. The third stage began after 43 h when the vanillin solution was added at 0.7 kg/h, while solvent addition and feed addition remained the same as in the second stage.

    [0228] Samples of the fermentation broth (reaction section, bottom of reactor) and organic phase (separation section, top of reactor) were taken approximately every 2 h during the day and 3 samples during the night, starting at 16 h (directly after the end of the batch, stage 1).

    [0229] The organic phase (including product recovery phase and product) were withdrawn with the start of stage 2 until the end of the fermentation. The pump was set to a fixed rate so that all separated organic phase would be harvested. The organic phase recovery rate was on average 0.1 kg/h during stage 2 and 0.18 kg/h during stage 3.

    [0230] Main cultivation parameters can be found in Table 15:. The pH was controlled with 25 (v/v)% NH.sub.4OH solution. The pH control was activated when the pH dropped below 4.5 and sufficient NH.sub.4OH solution was added to maintain the pH at 4.5.

    TABLE-US-00015 TABLE 15 Cultivation parameters during batch, set point 1 and set point 2 Parameter Value Temperature 30° C. Vessel pressure 0.1 barg Liquid working volume ~70 L Air flow 30 L/min pH 6.3 to 4.5 pH control 25 (v/v)% NH.sub.4OH Stirring Batch (stage 1): 300 rpm Stage 2 and 3: 300-600 rpm (controlled to DO cascade 30%) Antifoam control Foam: on contact Solvent Oleyl alcohol

    [0231] Vanillin, vanillic acid and vanillic alcohol are potential inhibitors to microorganisms [Converti et al. Brazilian Journal of Microbiology (2010) 41: 519-530]. This is why the concentration of all of these components should be maintained at lower inhibiting concentrations in order to keep fermentation rates fast. Aqueous vanillin can already be toxic for microorganisms at a concentration of 0.5 g/L [Hansen et al Applied and Environmental Microbiology (May 2009), p2765-2774]. FIG. 33 shows that the aqueous vanillin concentration was below 0.5 g/L during the whole fermentation, while vanillic acid and vanillic alcohol were enriched in the organic phase (vanillin: square dark & square light, vanillic acid (van. acid): circle filled & circle pattern and vanillic alcohol (van. Alcohol: triangle filled & triangle pattern) in the organic and aqueous phase. The growth was not inhibited by the compounds present at the measured concentrations (FIG. 32). By harvesting continuously, the organic concentration of oleyl alcohol was maintained below 4%. This lower overall oil content (due to removal of oil in time) minimized the influence of the dispersed phase on the microorganism morphology. In this experiment, vanillic alcohol was formed as by-product.

    [0232] The product of interest is vanillic acid. Vanillic acid was formed by S. cerevisiae and accumulated in the broth until the steady state was reached (FIG. 34 showing the vanillic acid concentration in the fermentation broth; the vanillic acid concentration in aqueous phase was analysed for stage 2 and the end of stage 3.4). At the end of stage 2 (set point 1, 43 h), the yeast began to convert vanillin to vanillic acid with a productivity of 0.009 g/L/h which increased to 0.026 g/L/h when the vanillin addition to the fermenter was increased. The steady state concentration of vanillic acid in the aqueous fermentation broth was approximately 0.63 g/L.

    [0233] FIG. 35 shows the vanillic acid production rate (square), extraction rate (circle) and separation rate (triangle) over time. As can be seen in FIG. 35, the extraction rate equalled the production rate when all organic phase and aqueous phase concentrations remained constant (or steady state) which is from ˜65 hours onwards. The net organic phase extraction and production rate were around 0.002 kg/h vanillic acid.

    [0234] The overall product removal rate from the reactor at the end of the fermentation was >70% of the net production of product at that time. When taking into account limitation in analyses methods, hold-up changes during time due to system kinetics and mass balance gaps, the overall product removal rate is close to the net product rate, especially between 40-50 h. If desired, an increase in overall organic phase recovery can be achieved if emulsion stability is controlled more tightly. Stability, possibly increased by the 0.6 kg antifoam 98/007K (Basildon Chemicals) dosed, can be optimized further.

    [0235] The production rate was maintained high and the inhibition low by extracting the inhibiting compounds into the organic phase. The production rate was successfully increased during the third stage (set point 2, after 43 h).

    Example 9: Back-Extraction of vanillic acid

    [0236] Vanillic acid obtained from catalytical conversion within Example 8 was extracted from the product recovery phase (oleyl alcohol) into an auxiliary solvent (70 (v/v)% ethanol in water). The organic phase withdrawn from the separation section in the top of reactor, consisting mostly of oleyl alcohol, vanillic acid and vanillic alcohol, was centrifuged at 4000 rpm for 20 min to remove possible solid residuals. Afterwards, the organic phase was divided into 3 sets of duplicates, each with 3.5 mL organic phase. The sample sets were overlaid with 70% ethanol solution in the following amounts—2 mL (sample number 1, 2), 3 mL (sample number 3, 4) and 6 mL (sample number 5, 6)—and mixed at room temperature for 1 h. Oleyl alcohol and 70% ethanol formed two phases before and after mixing. In order to ensure clean liquid/liquid phase separation, the samples were centrifuged at 4000 rpm for 20 min after the mixing step. The samples were analysed according to their vanillic acid content in oleyl alcohol.

    [0237] FIG. 36 shows the successful back-extraction of vanillic acid from oleyl alcohol into 70% ethanol (auxiliary solvent). 70% ethanol has a lower boiling point than oleyl alcohol, which facilitates the downstream processing of the product towards the desired quality by solvent evaporation or (vacuum) distillation. The separation of vanillic acid and ethanol is less energy-intensive due to the overall stream size reduction and heat of evaporation of ethanol, which is noticeable lower then e.g. water. A remark here is that ethanol cannot be used directly in the fermentor as it is highly miscible with aqueous solutions. Therefore, oleyl alcohol was chosen as primary solvent for vanillic acid extraction from fermentation broth. Oleyl alcohol is not toxic towards the microorganism, while maintaining adequate aqueous product concentration and phase separation properties.

    [0238] Additionally, the back-extraction experiment confirms the benefit of enriching the product of interest in the auxiliary solvent (Table 16). Vanillic acid was enriched in ethanol compared to vanillic alcohol. Continuous extraction of vanillic acid in a reactor according to the invention (such as schematically shown in FIG. 1) by oleyl alcohol and back-extraction to low boiling solvent enabling further efficient downstream processing for product recovery.

    TABLE-US-00016 TABLE 16 Concentration of vanillic acid and vanillic alcohol before and after concentration demonstrating enrichment of vanillic acid in 70% ethanol Vanillic Vanillic acid in alcohol Vanillic Vanillic oleyl in oleyl acid in alcohol in alcohol alcohol ethanol ethanol Sample [g/L] [g/L] [g/L] [g/L] 1 1.15 0.25 3.62 0.82 2 1.31 0.30 3.59 0.78 3 1.17 0.30 2.42 0.50 4 1.39 0.30 2.13 0.48 5 0.55 0.19 1.60 0.32 6 0.47 0.17 1.55 0.31 Before 3.37 0.75 — — extraction

    [0239] Additionally, phase separation was tested for two different phase equilibria. Oleyl alcohol, including vanillic acid, was used as primary solvent (product recovery phase) in both experiments and either (i) 70% ethanol in water (aqueous ethanol) or (ii) hexylcinnamaldehyde were used as back-extraction solvent. 1.5 mL oleyl alcohol and 1.5 mL hexylcinnamaldehyde were mixed; 3.5 mL oleyl alcohol were mixed with 3 mL aqueous ethanol. The samples with hexylcinnamaldehyde were stored in the fridge overnight to identify the effect of temperature on phase separation. Two distinct phase equilibria were observed at the different temperatures. (i) was a liquid-liquid system of phases of aqueous ethanol and oleyl alcohol at room temperature. (ii) was a liquid-solid system of liquid oleyl alcohol and solid phase containing vanillic acid after cooling (6° C.). (i) and (ii) showed that a wide range of temperature is possible for phase separation strategies which can be used for the back-extraction of vanillic acid.