METHOD FOR SEPARATING BIOMASS FROM A SOLUTION COMPRISING BIOMASS AND AT LEAST ONE AROMA COMPOUND

Abstract

The present invention relates to a method for separating biomass from a solution comprising bi-omass and at least one aroma compound. comprising providing the solution comprising bio-mass and aroma compounds. lowering the pH value of the solution below 7 by adding at least one acid to the solution comprising biomass and the at least one aroma compound. adding an adsorbing agent to the solution comprising biomass and aroma compounds. and carrying out first membrane filtration so as to separate the biomass from the solution comprising the at least one aroma compound.

Claims

1.-15. (canceled)

16. A method for separating biomass from a solution comprising biomass and at least one aroma compound, comprising the following steps in this order: (1) providing the solution comprising biomass and one or more aroma compound(s); (2) setting the pH value of the solution below 7.0 if needed, preferably by adding at least one acid to the solution comprising biomass and the at least one aroma compound; (3) adding at least one adsorbing agent to the solution comprising biomass and aroma compounds; and (4) carrying out a first membrane filtration so as to separate the biomass from the solution comprising at least one aroma compound wherein the at least one aroma compound has one glycosidic bond or no glycosidic bonds and is not a protein.

17. The method according to claim 16, wherein the pH value of the solution is lowered to a pH value in the range of 3.0 to 5.5.

18. The method according to claim 16, wherein said at least one acid is an acid selected from the group consisting of H.sub.2SO.sub.4, H.sub.3PO.sub.4, HCl, HNO.sub.3 and CH.sub.3CO.sub.2H.

19. The method according to claim 16, wherein said adsorbing agent is added in an amount in the range of 0.3% to 3% by weight.

20. The method according to claim 16, wherein said adsorbing agent is added as a powder having a particle size distribution with a diameter d50 in the range of 2 μm to 25 μm.

21. The method according to claim 16, wherein said first membrane filtration is carried out as cross-flow microfiltration or cross-flow ultrafiltration.

22. The method according to claim 21, wherein said cross-flow microfiltration or cross-flow ultrafiltration includes a cross-flow speed in the range of 0.5 m/s to 6.0 m/s.

23. The method according to claim 21, wherein said cross-flow speed is equal to or below 3 m/s.

24. The method according to claim 16, wherein said first membrane filtration is carried out at a temperature of the solution in the range of 8° C. to 55° C.

25. The method according to claim 16, wherein said first membrane filtration is carried out by means of a ceramic microfiltration or ultrafiltration membrane having a pore size in the range of 20 nm to 800 nm, or wherein said first membrane filtration is carried out by means of a polymeric microfiltration membrane or polymeric ultrafiltration membrane having a cut-off in the range of 10 kDa to 200 nm.

26. The method according to claim 16, further comprising carrying out a second membrane filtration with the solution comprising aroma compounds obtained by the first membrane filtration, and optionally followed by a reverse osmosis.

27. The method according to claim 26, wherein said second membrane filtration is carried out at a temperature of the solution being in the range of 5° C. to 15° C.

28. The method according to claim 16, wherein said at least one aroma compound comprises at least one polar aroma compound selected from the group consisting of Ambrox, Ambrox-1,4-diol, furaneol, benzoic acid, phenylethanol, raspberry ketone, pyrazines, sclareol, vanillin, vanillyl alcohol and vanilla glycoside.

29. A processing unit comprising i) a vessel filled with a solution containing biomass, at least one adsorbing agent and at least one aroma compound, wherein the at least one aroma compound has one glycosidic bond or no glycosidic bonds, and is not a protein, and wherein the pH value of the solution is below 7.0; ii.) a first filtration membrane which is a microfiltration or an ultrafiltration membrane; iii.) means to carry out a first membrane filtration across said first filtration membrane, which is a microfiltration or ultra-filtration, to generate a permeate containing the bulk of the aroma compounds, wherein the processing unit further comprises means to have the solution at a temperature in the range of 8° C. to 55° C. during the first membrane filtration; and iv.) means to separate the permeate of the first membrane filtration from the solution as described in i). above; v.) means to transport said permeate of the first membrane filtration to a second filtration membrane, vi.) means to adjust the temperature of the permeate to a temperature below 20° C., vii.) a second filtration membrane, viii.) means to carry out a second membrane filtration at a temperature below 20° C., and ix.) means to keep separate the permeate of the second membrane filtration from the permeate of the first membrane filtration; and optionally x) means to carry out a reverse osmosis treatment of the permeate of the second membrane filtration; wherein the surfaces of the parts of the processing unit that are in contact with the solution or any of the permeates are tolerant to pH values as low as pH 3.5 and optionally are made of material suitable for the production of food grade material.

30. A method for reducing wear and tear on and/or energy consumption of membrane filtration equipment used in the separation of biomass from a solution comprising at least one aroma compound, wherein the method comprises these steps in the following order: i. providing the solution comprising biomass and aroma compound(s), ii. adjusting the pH value of the solution to a pH value in the range of 3.0 to 5.5, iii. adding one or more adsorbing agents to the solution comprising biomass and aroma compound(s); iv. optionally an incubation step sufficient for the one or more adsorbing agents to bind the color components in the solution, and v. carrying out first membrane filtration so as to separate the biomass from the solution comprising the comprising at least one aroma compound at cross-flow speeds of no more than 3 m/s. vi. optionally carrying out a second membrane filtration with the solution comprising aroma compounds obtained by the first membrane filtration.

Description

FIGURES

[0116] FIG. 1 shows a block diagram of a method for separating biomass from a solution comprising biomass and at least one aroma compound according to the present invention.

[0117] FIG. 2 displays in a block diagram the sequence of steps of the inventive methods when a separate incubation step S15 for the adsorbing agent and the three sub-steps S16/1 to S16/3 as explained above.

[0118] FIG. 3: Comparison of membrane flux as a function of DF and feed quality (with and without pre-treatment with 1.0% (w/w) active carbon). Circles represent without pre-treatment, squares with pre-treatment.

[0119] FIG. 4: Trends of UF membrane flux as a function of CF and feed quality (with and without pre-treatment with 1.0% (w/w) active carbon). Circles represent without pre-treatment, squares with pre-treatment.

[0120] FIG. 5: Comparison of membrane flux as a function of DF and feed quality (with and without pre-treatment with 1.5% w/w active carbon. Circles represent without pre-treatment, squares with pre-treatment.

[0121] FIG. 6: Comparison of UF membrane flux as a function of CF and feed quality (with and without pre-treatment). Circles represent without pre-treatment, squares with pre-treatment.

[0122] FIG. 7: Comparison of UF membrane flux as a function of relative permeate amount and feed AC concentration. Circles represent Trial 3A (0.7% (w/w) of AC), squares Trial 3B (1.5% (w/w) of AC)

EXAMPLES

[0123] The method according to the present invention will be described in further detail below. Whatsoever, the Examples shall not be construed as limiting the scope of the invention.

Generalized Example

[0124] A fermentation broth as a complex solution comprising biomass and at least one aroma compound is being prepared by standard methods. The pH value thereof is lowered to 4±0.1 by means of adding 10% sulphuric acid. Thereafter, a 30% suspension of active carbon Carbopal Gn-P-F (Donau Carbon GmbH, Gwinnerstrafle 27-33, 60388 Frankfurt am Main, Germany), which is food safe, is added and stirred for 20 min.

[0125] The thus prepared solution is supplied to the process apparatus, a semi-automatic MF lab unit from Sartorius AG, Otto-Brenner-Str. 20, 37079 Goettingen, Germany, modified for the purpose, and heated to 37° C. in a circulating manner with closed permeate. For separation purposes, the process apparatus includes a ceramic mono channel element (from Atech Innovations GmbH, Gladbeck, Germany) having an outer diameter of 10 mm, an inner diameter of 6 mm, a length of 1.2 m and a membrane made of Al.sub.2O.sub.3 having a pore size of 50 nm. As soon as the circulation of the solution is running and the solution comprises the target temperature of 37° C., the discharging of the permeate is started and the control of the trans membrane pressure is activated.

[0126] After terminating of the first membrane filtration, the process apparatus is stopped, the concentrate is disposed, and the process apparatus is being cleaned. Cleaning is carried out by means of 0.5% to 1% NaOH at a temperature of 50° C. to 80° C., wherein the NaOH is subsequently removed by purging.

[0127] In one embodiment, the first membrane filtration of the inventive methods includes three steps as will be explained in further detail below. The first step includes a first diafiltration having a factor of 0.5 (amount of diafiltration water=starting amount of fermentation broth x diafiltration factor). During diafiltration, the amount of water added is identical to the amount of permeate discharged. The first step is a continuing step and the volume in the feed vessel is thus kept constant. The second step includes concentrating of the fermentation broth with the factor 2 by stopping the feed of diafiltration water and the level will decrease down to the target value (target value=volume or mass at the beginning of the fermentation broth/concentrating factor). Subsequently, the third step includes a second diafiltration. The permeates collected during these three steps are typically combined to form the permeate. By means of these three steps a lower dilution of the product within the permeate and an increased yield are realized. By increasing the factor of the second diafiltration, the yield may even be increased.

[0128] The following analytical methods are been carried out. [0129] HPLC or GC or GC-MS for the determination of the product, i.e. aroma compounds, and secondary components [0130] Drying balances for measuring the dry content (DC) [0131] APHA for measuring the colour using standard methods, for example DIN EN ISO 6271 [0132] Bradford protein assay for measuring the concentration of protein.

[0133] Hereinafter, the following abbreviations are used: [0134] AC=Active Carbon [0135] UF=Ultrafiltration [0136] DP=Pressure drop along the module (p.sub.feed-p.sub.retentate) [0137] Flux=Permeate flow rate per m.sup.2 and hour (l/m.sup.2 h) [0138] Cross-flow velocity=linear speed of the suspension in membrane channels (m/s) [0139] Membrane load=amount of permeate produced by 1 m.sup.2 of membrane area (m.sup.3/m.sup.2)

[0140] Further, regarding the liquid separation, the following symbols and explanations are used.

TABLE-US-00001 Symbol Meaning Unit Definition C Concentration wt-%, g/L CF Concentration — m.sub.R, t=0/m.sub.R factor DF Diafiltration — m.sub.P/m.sub.R, t=0 factor J Flux LMH = L m.sup.−2 h.sup.−1 m Mass kg P Pressure bar R Retention — 1 − c.sub.permeate/c.sub.retentate TMP Trans-membrane bar (p.sub.feed + p.sub.retentate)/2 − pressure p.sub.permeate

[0141] In the following, examples of the inventive methods with vanillin as the aroma compound are explained in detail to exemplify the invention without limiting it.

[0142] Common elements to the purification experiments for Vanillin described in the following: Separation of biomass from dissolved components was completed in a batch dynamic cross flow lab unit (called MF), running fully automatically, allowing for temperature, TMP, circulation flow control as well as automatic dosing of DF medium to the feed vessel. Permeate flow rate was measured by a scale. As a separation barrier ceramic 50 nm Al.sub.2O.sub.3 membrane from Atech Innovations GmbH (abrasion resistant three layers membrane 50/200/400 nm), mono tube 10/6, with 1.2 m length, having effective membrane area of 0.0223 m2. Only after reaching pre-defined crossflow velocity and predefined filtration temperature, TMP controlled permeate flow was activated.

[0143] Before and after each trial membrane performance (flux) was measured during rinsing with water at 25° C. Rinsing was running for 30 minutes. After each trial the unit was at first rinsed with deionized water, drained, and then cleaned with 0.5% Ultrasil® 125 (membrane cleaner from Ecolab) at 68-70° C. for 30 minutes. After draining the cleaner solution, the unit was rinsed and drained three times to remove completely cleaner solution and finally flux was measured again with deionized water at 25° C.

[0144] Permeates from MF were then treated in a lab UF unit to separate macro molecules: proteins, polysaccharides, RNA, DNA, etc. by means of polymeric Ultrafiltration membrane. In the trials the 4 kD PES membrane UH004 from Microdyn-Nadir was used.

[0145] During UF trials only concentration of MF permeate was applied (no diafiltration). The concentration was running until the minimum retentate amount was left in the unit or the permeate flux was reduced below 5 kg/m2 h.

[0146] UF tests were completed in a flat cell unit having three round membrane cells connected serially, having 0.00793 m2 membrane area. The feed pressure was provided by pressure control of N2 in the feed vessel, the cross-flow velocity was provided by a gear pump. Permeate flow was measured by a scale.

[0147] Similar membrane regeneration procedure as for ceramic membrane was used for polymeric membrane. Rinsing was completed at 30° C. and membrane cleaning at 50° C. for the same time.

[0148] Samples of fermentation broth, retentate and permeate were analyzed for proteins (according to Bradford method), color (APHA on PerkinElmer Lambda 35), dry content by drying samples (Mettler-Toledo HX204, Moisture Analyzer), and vanillin with GC method (see table A below for details). Samples of fermentation broth and MF retentate were filtrated with 0.2-micron filter, to ensure solid free samples.

TABLE-US-00002 TABLE A GC conditions Column: Fused silica, DB-1 ms UI Length: 30 m Internal diameter: 0.25 mm Film thickness: 0.25 μm Carrier gas: Nitrogen Column Flow/Head pressure 1 mL/min (constant flow) Split ratio: 15:1 Septum purge: 3 mL/min Oven temperature: 50° C., 5 min isothermal 50° C. to 240° C., 5K/min Injector temperature: 250° C. Detector temperature: 300° C. Injection volume: 1 μl

[0149] A fermentation broth of a standard lab strain of E. coli cells grown under typical conditions for a time that corresponds the known times for E. coli for growth through the growth phase and into the bioconversion phase. To simulate a fermentation broth of E. coli that has produced the aroma compound vanillin, vanillin was added to the fermentation broth to a concentration of around 10 g/I and gently mixed.

[0150] This vanillin containing fermentation broth was used for the following experiments.

Example No. 1

[0151] An aliquot of the fermentation broth batch was divided in two portions. One part (Trial 1A) was directly diafiltrated with DF (diafiltration factor) of 4 using deionized water. The second part of fermentation broth (Trial 1B) was acidified with 20% H2SO4 to reach pH 4.5 and then so much of AC active carbon powder Carbopal Gn-P-F was added to obtain 1.0% (g/g) AC concentration. After that this suspension was identically diafiltrated with DF of 4 using deionized water. MF conditions with fermentation broth without pretreatment (Trial 1A):

[0152] T=35° C.

[0153] TMP=1 bar

[0154] Cross-flow velocity 4 m/s (4001/h)

[0155] MF conditions with fermentation broth with pretreatment (Trial 1B):

[0156] T=35° C.

[0157] TMP=1 bar

[0158] Cross-flow velocity 3.5 m/s (3501/h)

[0159] UF conditions with both permeates

[0160] T=10° C.

[0161] TMP=10 bar

[0162] Cross-flow velocity 1.5 m/s

[0163] Results from Separation of Biomass from Dissolved Components by a First Membrane Filtration by Microfiltration (MF)

[0164] Table 1 presents the performance data of biomass separation with ceramic membrane for fermentation broth without pretreatment (Trial 1A) and with pretreatment (Trial 1B). Membrane performance is 2.5 times higher, when fermentation broth was pretreated with pH adjustment and with addition of 1% of AC powder even though the cross-flow velocity is reduced. Due to this reduction in the crossflow velocity, the pressure difference (DP) along 1.2 m long membrane is reduced by 25%. This significantly reduces electrical energy demand by an industrial unit.

TABLE-US-00003 TABLE 1 Performance data Flux DP cross-flow Trial [kg/m.sup.2h] [bar] [m/s] A 124 1.01 4.0 B 305 0.85 3.5

[0165] Table 2 presents membrane performance, measured with deionized water, before each trial, during rinsing after the trial, and after membrane regeneration. The results indicate that the membrane performance was easy re-established, slightly better after the trial with 1% AC.

TABLE-US-00004 TABLE 2 Membrane regeneration data Flux- Flux [kg/m.sup.2h] Regeneration Improvement Trial before trial after trial after CIP [%] [%] 1A 509 150 477 94 1B 477 251 532 112 67

[0166] Table 3 (see following page) presents mass balance for both trials for APHA, protein, dry content of dissolved components, and vanillin. The data show clearly that pretreatment of fermentation broth significantly reduces color components not only in permeate but also in feed and retentate. Concentration of proteins is reduced by a factor of >5 and their retention is significantly higher in comparison to the trial without pretreatment.

[0167] Whereas protein amount in permeate in trial 1A is similar to the amount in feed, the amount of proteins is reduced by factor 3 in permeate from trial 1B.

[0168] The analyzed vanillin concentration in feed is significantly lower in feed in pretreated fermentation broth, indicating that vanillin partly adsorbs on AC. However, during diafiltration step it was possible to wash it out, so that similar amount of vanillin could be found in permeate.

[0169] FIG. 3 presents flux trends as a function of diafiltration factor for both trials. The trends indicate that even with the higher solids load (1% AC) the membrane performance in trial 1B is significantly higher. A stable filtration performance was reached with or without active carbon.

[0170] Results from Concentration of the Permeates of the First Membrane Filtration with Ultrafiltration (UF) as Second Membrane Filtration

[0171] Table 4 presents reached concentration factors and average flux during the trial. The results clearly indicate the advantages of pretreatment of the fermentation broth. Concentration factor is by factor 7 higher as well as the average membrane performance is almost three times higher when fermentation broth was pretreated.

TABLE-US-00005 TABLE 4 Performance data Flux Trial CF [kg/m.sup.2h] 1A 4.6 10 1B 34.8 28

[0172] Table 5 provides membrane regeneration data, indicating that the chosen membrane can be easily regenerated even it was strongly fouled as in trial 1A.

TABLE-US-00006 TABLE 5 Membrane regeneration data Flux [kg/m.sup.2h] Regeneration UF trial before trial after trial [%] 1A 72 66 92 1B 62 59 95

TABLE-US-00007 TABLE 3 First membrane filtration: Mass balance of both trials Mass Protein R.sub.protein DC (dissolved) Vanillin R.sub.Vanillin Trial Stream [g] APHA [mg/l] [g] [%] [%] [g] [g/l] [g] Yield [%] 1A Feed 1611.0 1988 198 0.32 6.48 104.4 9.0 14.5 Permeate 6456.5 418 63 0.41 68 1.49 96.2 2.0 12.9 Retentate 1606.9 78 23 0.04 0.13 2.1 0.6 1.0 0.07 32 1B Feed 1210.7 735 114 0.14 6.31 76.4 5.0 6.1 Permeate 4859.4 160 11 0.05 90 1.39 67.5 2.2 10.7 Retentate 1207.7 12 17 0.02 0.14 1.7 0.2 0.2 0.04 20 Yield is given as % (w/w); R.sub.Protein is the retention of proteins, R.sub.Vanillin is the retention of Vanillin; same applies to table 6.

TABLE-US-00008 TABLE 6 Second membrane filtration: Mass balance of both trials Mass Protein R.sub.protein DC (dissolved) Vanillin Yield R.sub.vanillin Trial Stream [g] APHA [mg/l] [g] [%] [%] [g] [g/l] [g] [%] [%] 1A Feed 2712.2 418 63 0.17 1.59 4.3 2.0 5.4 Permeate 2116.6 245 60 0.13  5 1.20 2.5 2.0 4.2 78 Retentate  595.6 1187 181 0.11 3.09 1.8 3.0 1.8 33 27 1B Feed 4454.0 160 11 0.05 1.53 6.8 2.0 8.9 Permeate 4326.1 115 4 0.02 65 1.24 5.4 2.0 8.7 97 Retentate  127.9 1903 152 0.02 7.81 1.0 2.0 0.3  3  0

[0173] Table 6 (see previous page) presents the mass balance for both trials for APHA, dry content, proteins and vanillin. In permeate after pretreatment (trial 1B) APHA is reduced by factor 2, concentration of proteins by factor 25 in comparison to trial 1A (without pretreatment). Dry content in trial 1B is slightly higher, indicating that dissolved components pass the membrane with lower retention, because of significantly lower protein concentration and its fouling tendency. During that investigation period, polysaccharides were not analyzed but they surely play important role by membrane fouling. Vanillin recovery in trial 1B was with 97% significantly higher than in trial 1A with 78%. Its retention in trial 1B was 0%, vs. 27 in trial 1A.

[0174] FIG. 4 provides trends of flux for both trials as a function of concentration factor. Membrane performance in trial 1B was still twice as high after CF of 34.8 as in trial 1A after CF of 4.6. Because of online data collection problem in trial 1B a part of flux values is missing. Because the operation volume of the UF unit was limited to 1.5 liters, the unit had to be stopped and refilled with feed solution. This was done three times during the 1B trial. After each restart the flux increased but relatively quick reached the previous values.

[0175] Subsequent Reverse Osmosis

[0176] Permeates from the UF trials 1A and 1B were concentrated by reverse osmosis in stirred test cells using different reverse osmosis membranes. Two types of Membranes (AlfaLaval R099 and Microdyn-Nadir ACM2) were used for both samples

[0177] Both experiments started with feeds with approximately same solid contents (1.24% for the experiments with active carbon, 1.20% for the experiments without active carbon). The results indicate that the reverse osmosis shows a better performance for broths that were treated with pH adjustment and active carbon. If the samples were pre-treated with pH adjustment and active carbon before the first membrane filtration, the solid retention and the Vanillin retention were higher for both membranes, and fluxes were continuously improved.

Summary: Advantages of Pre-Treatment

[0178] Biomass separation by pre-treatment and first membrane filtration: [0179] 2.5-time higher flux [0180] Lower pressure difference through membrane length [0181] 5.5-time less proteins in permeate [0182] 2.5-time lower APHA [0183] Higher vanillin yield in permeate

[0184] UF separation of macromolecules: [0185] 7-time higher concentration factor of the feed [0186] 2.8-time higher average flux [0187] 2-times lower APHA [0188] 25-time less proteins in permeate [0189] Significantly higher vanillin yield (97% vs 78%)

[0190] Reverse Osmosis [0191] Better solid retention [0192] Better Vanillin retention

Example No. 2

[0193] A new fermentation broth batch was again divided in two portions after the addition of vanillin to around 10 g/l. The trials were conducted identical as in example 1 except a different amount of activated charcoal was used compared to example 1. One part (Trial 2A) was directly diafiltrated, with DF of 4 using deionized water. The second part of fermentation broth (Trial 2B) was acidified with 20% H2SO4 to reach pH 4.5 and then so much of AC active carbon powder Carbopal Gn-P-F was added to obtain 1.5% of its concentration. After that the pretreated fermentation broth was diafiltrated with DF of 4 using deionized water.

[0194] MF conditions with fermentation broth without pretreatment (Trial 2A):

[0195] T=35° C.

[0196] TMP=1 bar

[0197] Cross-flow velocity 4 m/s (4001/h)

[0198] MF conditions with fermentation broth with pretreatment (Trial 2B):

[0199] T=35° C.

[0200] TMP=1 bar

[0201] Cross-flow velocity 3.5 m/s (3501/h)

[0202] UF conditions with both MF permeates

[0203] T=10° C.

[0204] TMP=10 bar

[0205] Cross-flow velocity 1.5 m/s

[0206] Results from Separation of Biomass from Dissolved Components (MF)

[0207] Table 7 presents the performance data of biomass separation with ceramic membrane for fermentation broth without pretreatment (Trial 2A) and with pretreatment (Trial 2B). Membrane performance is 2.5 times higher, when fermentation broth was pretreated with pH adjustment and with addition of 1% of AC powder even though the cross-flow velocity is reduced, identical to the trail 1B in the first example.

TABLE-US-00009 TABLE 7 Performance data Flux DP cross-flow Trial [kg/m.sup.2h] [bar] [m/s] 2A 171 1.00 4.0 2B 286 0.85 3.5

[0208] Table 8 presents membrane performance, measured with deionized water, before each trial, during rinsing after the trial, and after membrane regeneration. The results indicate that the membrane performance was easy re-established.

TABLE-US-00010 TABLE 8 Membrane regeneration data Flux- Flux [kg/m.sup.2h] Regeneration Improvement Trial before trial after trial after CIP [%] [%] 2A 532 182 501 94 2B 501 228 473 94 25

[0209] Table 9 (see following page) presents mass balance for both trials for APHA, protein, dry content of dissolved components, and vanillin. The data show clearly that pretreatment of fermentation broth with pH adjustment to 4.5 and the addition of 1.5% AC significantly reduces color components not only in permeate but also in feed and retentate. In permeate from pretreated fermentation broth: APHA is reduced by factor of 3, concentration of proteins is reduced by a factor of 35 and their retention is significantly higher. However, Vanillin concentration in feed is much lower in feed in pretreated fermentation broth, indicating that vanillin partly adsorbs on AC. However, during diafiltration step it was possible to wash it so that 15% more vanillin could be recovered, and its concentration was 2.5 times lower in retentate. Vanillin yield and retention for trial 2B were calculated with the dissolved vanillin concentration in the feed, and that is why they do not depicture the reality.

[0210] FIG. 5 presents flux trends as a function of diafiltration factor for both trials. The trends indicate that even by higher solids load (1.5% AC) the membrane performance in trial 2B is significantly higher. In both cases a stable filtration performance was reached. Similar trends as in the first example were measured.

[0211] Results from Concentration of MF Permeates with UF

[0212] Table 10 presents reached concentration factors and average flux during the trial. The results clearly indicate the advantages of pretreatment of the fermentation broth. Concentration factor is by factor 13 higher as well as the membrane performance is five times higher when fermentation broth was pretreated.

TABLE-US-00011 TABLE 10 Performance data Flux Trial CF [kg/m.sup.2h] 2A 2.9 6 2B 39.0 31

TABLE-US-00012 TABLE 9 First membrane filtration: Mass balance of both trials Mass Protein R.sub.protein DC (dissolved) Vanillin R.sub.vanillin Trial Stream [g] APHA [mg/l] [g] [%] [%] [g] [g/l] [g] Yield [%] 2A Feed 1180.4 1259 226 0.27 6.90 81.4 9.0 10.6 Permeate 4746.2 375 46 0.22 80 1.48 70.2 2.0 9.5 Retentate 1183.1 73 3 0.01 0.09 4.3 0.5 2.4 0.22 62 2B Feed 1292.9 415 33 0.04 6.26 80.9 3.0 3.9 Permeate 5263.8 126 3 0.02 91 1.37 72.1 2.1 11.1 Retentate 1266.3 14 11 0.06 0.11 5.7 0.2 1.0 0.27 67 Yield is given as % (w/w); R.sub.Protein is the retention of proteins, R.sub.Vanillin is the retention of Vanillin; same applies to table 12

TABLE-US-00013 TABLE 12 Second membrane filtration: Mass balance of both trials Mass Protein R.sub.protein DC (dissolved) Vanillin Yield R.sub.Vanillin Trial Stream [g] APHA [mg/l] [g] [%] [%] [g] [g/l] [g] [%] [%] 2A Feed 2109.0 375 46 0.10 1.63 3.4 2.0 4.2 Permeate 1384.6 195 38 0.05 17 1.16 1.6 3.0 2.8 66 Retentate  724.4 724 122 0.09 2.48 1.8 3.0 2.2 52 38.0 2B Feed 4921.8 126 3 0.01 1.48 7.3 2.0 9.8 Permeate 4768.7 98 2 0.01 33 1.29 6.2 2.0 9.5 97 Retentate  153.1 1662 140 0.02 7.87 1.2 2   0.3  3  5.0

[0213] Table 11 provides membrane regeneration data, indicating that the chosen membrane can be easily regenerated even it was strongly fouled as in trial 2A.

TABLE-US-00014 TABLE 11 Membrane regeneration data Flux [kg/m.sup.2h] Regeneration UF trial before trial after trial [%] 2A 73 71 97 2B 67 65 97

[0214] Table 12 (see previous page) presents the mass balance for both trials for APHA, dry content, proteins and vanillin. In permeate after pretreatment (trial 2B) APHA is reduced by factor 2, concentration of proteins by factor 19 in comparison to trial 2A (without pretreatment). Dry content in trial 2B is slightly higher, indicating that dissolved components pass the membrane with lower retention. Vanillin recovery in trial 2B was with 97% significantly higher than in trial 2A with 66%. Its retention in trial 2B was 5%, vs. 33 in trial 2A.

[0215] FIG. 6 provides trends of flux for both trials as a function of concentration factor. Membrane performance in trial 2B was still 2.5 times higher after CF of 39 as in trial 2A after CF of 2.9. After each restart the flux increased but relatively quick reached the previous values.

Summary: Advantages of Pre-Treatment

[0216] Biomass Separation: [0217] 70% higher flux [0218] 15% lower pressure difference through membrane length [0219] 15-time less proteins in permeate [0220] 3-time lower APHA [0221] At least 10% higher vanillin yield in permeate

[0222] UF Separation of Macromolecules: [0223] 13-time higher concentration factor of the feed [0224] 5-time higher average flux [0225] By factor 2 lower APHA [0226] 19-time less proteins in permeate [0227] Significantly higher vanillin yield (97% vs 67%)

Example 3

[0228] In this example the dependence of active carbon (AC) concentration on vanillin separation was investigated. AC concentration was adjusted to 0.7% in trial 3A and 1.5% in trial 3B. Additionally the procedure for biomass separation was modified: In the first membrane filtration, after DF=5 an additional CF step was conducted until reaching dead volume of the unit (CF in trail 3A=2.65, in trial 3B=2.71).

[0229] The procedure for the UF step did not changed.

[0230] Results from Separation of Biomass from Dissolved Components (MF)

[0231] Table 13 presents the performance data of biomass separation with ceramic membrane for fermentation broth with two AC concentrations: Trial 3A: AC=0.7% and Trial 3B: AC=1.5%. Membrane performance is similar in both trials indicating that already 0.7% AC significantly improves the membrane performance. (Compare the results from example 1 and 2).

TABLE-US-00015 TABLE 13 Performance data AC Flux DP cross-flow Trial [%] [kg/m.sup.2h] [bar] [m/s] 3A 0.7 232 0.86 3.5 3B 1.5 247 0.85 3.5

[0232] Table 14 presents the membrane regeneration data for both trials. Slightly better performance was obtained in trial 3B.

TABLE-US-00016 TABLE 14 Membrane regeneration data Flux- Flux [kg/m.sup.2h] Regeneration Improvement Trial before trial after trial after CIP [%] [%] 3A 412 190 395 96 3B 421 205 459 109 8

[0233] Table 15 (see following page) presents mass balance for both trials for APHA, protein, dry content of dissolved components, and vanillin. The data show clearly that pretreatment with higher AC concentration (trial 3B with 1.5% AC) much better reduces color components. The concentration of color components (as APHA) drops in the pretreated fermentation broth already by factor 4 when using 0.7% AC and by factor 12 when adding 1.5% AC. Further color reduction is reached during MF process, when using 1.5% AC an APHA value in permeate is two times lower than when using 0.7% AC.

[0234] Due to the pre-treatment (acidic conditions and AC) the concentration of dissolved proteins drops in the pretreated fermentation broth already by a factor 8 independently of the AC concentration. Further protein reduction takes place during MF process resulting in ca. 50% protein reduction in permeate when using 0.7% AC and in ca. 70% protein reduction when using 1.5% AC.

TABLE-US-00017 TABLE 15 Mass balance of both trials Mass Protein R.sub.protein DC (dissolved) Vanillin R.sub.Vanillin Trial Stream [g] APHA [mg/l] [g] [%] [%] [g] [g/l] [g] Yield [%] 3A Ferm. broth 1132.0 5837 648 0.73 9.1 10.3 Feed pretr. 1148.3 1416 81 0.09 7.83 89.9 7.4 8.5 Permeate 6436.4 214 5.6 0.04 82 1.34 86.2 1.5 9.7 >98 Retentate 452.3 24 95 0.04 0.09 0.4 <0.5 <0.2 <0.2 <26 3B Ferm. broth 1187.0 5837 648 0.77 9.1 10.8 Feed pretr. 1200.7 487 80 0.10 7.54 90.5 4.4 5.3 Permeate 6751.8 107 3.9 0.03 95 1.28 86.4 1.6 10.8 >98 Retentate 457.5 4 65 0.03 0.08 0.4 <0.5 <0.2 <0.2 <39 Yield is given as % (w/w); R.sub.Protein is the retention of proteins, R.sub.Vanillin is the retention of Vanillin

[0235] During the DF step proteins and color components have not been redissolved and even their concentration was further reduced, resulting in much cleaner permeate.

[0236] Vanillin in both retentates was below detection limit, indicating that the membrane procedure consisting of DF=5 followed by CF=2.5 reaches vanillin yield in permeate of >98%. Such high yield was already observed in previous examples. Nevertheless, AC adsorbs vanillin, vanillin is washed out during DF step with demineralized water as diafiltration medium.

[0237] FIG. 7 presents flux trends as a function of relative permeate amount for both trials. The trends indicate that both fluxes are practically equal, and that higher AC concentration did not result in higher flux. During the final concentration step, which begins when reaching relative permeate amount of 4 the flux remains constant.

SUMMARY

[0238] Fermentation broth pretreatment (acidification and AC addition) results in higher flux. However, no significant difference in membrane performance was observed in both trials with 0.7 and 1.5% AC [0239] Pretreatment is very effective for reduction of color components and proteins present in the fermentation broth [0240] Reduction of color (APHA) increases with increasing AC concentration [0241] Reduction of proteins is similar for both AC concentrations [0242] A cross flow filtration (MF) additionally allows for reduction of color and proteins membrane [0243] Even if vanillin adsorbs on AC, it is possible to reach almost 100% vanillin yield in permeate. By proper adjusting of diafiltration and concentration steps the amount of permeate can be reduced.

CITED PUBLICATIONS

[0244] CN105132472 [0245] CN105219806 [0246] EP 2 379 708 [0247] EP1081212 [0248] EP2583744 [0249] KR10-1163542 [0250] US20110028759 [0251] U.S. Pat. No. 6,133,033—CN105132472, [0252] U.S. Pat. No. 9,115,377 [0253] Vandamme & Soetart, Journal of Chemical Technology and Biotechnology 77:1323-1332 2002 [0254] WO2007099230 [0255] WO2015002528 [0256] WO2020/223417 [0257] WO2020/223418