3. Valorization of olive oil waste streams by the development of thin film composite membranes for selective removal of polyphenols BGU, Dr. C.

Transcription

1 3. Valorization of olive oil waste streams by the development of thin film composite membranes for selective removal of polyphenols BGU, Dr. C. Linder

2 Development Objectives Cost effective Extraction of commercially valuable polyphenols form OMWW using a Membrane Aromatic Recovery System (MARS) Develop membrane contactors for MARS with the stability, permeability and selectivity for extracting polyphenols from the other components of the OMWW The first goal is the recovery of tyrosol and hydroxytyrosol

3 Schematic of Membrane based extraction of phenolic compounds from OMWW The ph difference on the feed ( ph ~2) and permeate (ph ~13) side is used to establish a gradient to polyphenol by the formation of polyphenoxide on the permeate side

4 To achieve high flux we investigated stable composite Membranes for MARS prepared by coating of a selective polymer film (1 to 5 microns) on a porous support Example of thin selective top layer of Fluorinated polysiloxane on a UF support 4

5 Supports and Polymers used in membranes preparation Supports; PES UF membranes of 50kDa, 100kDa, 150kDa, and 300kDa Polymers: 1) Poly (dimethylsiloxane), hydroxyl terminated, average Mn ~110,000, (PDMS-1). 2) Poly (dimethylsiloxane), hydroxyl terminated, average Mn ~40,000, (PDMS-2). 3) Polytrifluoropropylmethylsiloxane, silanol terminated. (PTFS). 4) Poly (4-vinylphenol), average Mw ~25,000 5) Tetraethyl orthosilicate PDMS was used as a control for state of art MARS membranes to compare with the more stable and selective membranes from PTFS and P4VP

6 Required Membrane Characteristics Stability: The membrane must be stable to ph extremes of acid (ph 2) on the feed side and base ph on the permeate side. Selectivity: As a 1 st goal: A high permeability to the polyphenols hydroxytyrosol and tyrosol and a low permeability to the other components including acids, base, carboxylate polyphenols and non-aromatic carbon chains (oleuropein degradation products). Low permeability to the acids and bases generating the ph gradient. Permeability rates of polyphenols sufficiently high for cost effectiveness. Fabrication by readily available methods, materials and upscalable

7 Membrane Development Activities - Composite membranes where formed by coating a porous support with a thin dense layer of polymer. This allowed independent tailoring and optimization of the support membrane and the polymer coating film. - The work carried out was on a laboratory scale MARS based on flow through membrane contactors that were developed for characterization of different flat composite membranes with the goal of a) To improve membrane permeability and selectivity by incorporating additives with selective uptake of phenols into the selective films matrix, to enhance their selective permeability for extracting poly phenols from OMWW; b) Membranes stability in strong base and acid conditions used in the feed and permeate compartments. -The affect of different feeds on Membrane performance in both the MARS flow cells and in the dialysis cell. We studied how the feeds from different sources of OMWW affect membrane Mass flux (MF) and Overall Mass Transfer Coefficients (OMTC) of the organic solutes in the feed stream through the membrane. The mass transfer flux of phenol through a membrane J can be described as follows: J= K ov (C f -Cs), (mg/m2/sec K ov --- overall mass transfer coefficient, m/sec C f ---phenol concentration in feed, Cs - phenol concentration in stripping solution.

8 Membrane Parameters For ph stability: Fluorinated polysiloxane was studied as a substitute for polydimethylsiloxane (PDMS) which are not stable in strong acids and bases: Stable Polytrifluoropropylmethylsiloxane (PTFS) was investigated and formed into selective films with, tetraethyl orthosilicate as crosslinker, dibutyltin dilaurate as catalyst and Tetrahydrofuran as solvent. Improving selectivity: For increased phenol permeability membranes were prepared using different combinations of poly (4-vinylphenol) (P4VP) with PTFS. Supports for coating the thin selective films: Porous polyethersulfone ultrafiltration support purchased from Nadir, as well as the more porous pp nonwoven were investigated.

9 Membrane contactors: 1-unmounted; 2-assembled

10 Laboratory MARS

11 Dialysis cell with heat exchanger

12 Supports Permeability was checked without the selective layer Results of supports permeability testing with use of OMWW in dialysis cell Results: The thinner and more open PP nonwoven support had a lower permeability then the thicker and tighter 300K UF membranes. This may be attributed to fouling throughout the entire thickness of the PP and hydrophobic effects.

13 Effect of UF support on PTFS composite membrane permeability of phenol (Selective PTFS layer thickness was estimated as 1-3µm) Mass flux,mg/m2*sec,10-3 OMTC,m/sec, Cell103(300kDa) Cell109(150kDa) Cell110(100kDa) Cell111(50kDa) Membrane support MWCO, kda Results: The composite membranes on the more open UF supports had higher mass flux and OMTC than the tighter UF membranes.

14 Effect of UF support on composite membrane (PTFS:P4VP=8:2) phenol permeability (Selective PTFS:P4VP layer thickness was estimated as 1-3µm) Mass flux,mg/m2*sec,10-1 OMTC,m/sec, Cell104(300kDa) Cell107(150kDa) Cell113(100kDa) Cell106(50kDa) Membrane support MWCO, kda The composite membranes on the more open UF supports had higher mass flux and OMTC than the tighter UF membranes.

15 The different membranes checked for stability at ph 13 #71-- PDMS as 5µm thick coating on a PP support. #72 PDMS >5µm thick coating on a PP support. #73-- PDMS+PTFS 5µm thick coating on a PP support. #74-- PDMS+PTFS >5µm thick coating on a PP support. PDMS- polydimethylsiloxane PTFS- polytetrafluorosiloxane

16 Mass flux,mg/m2*sec,x10-3 PDMS and PDMS/PTFS membrane stability characterized by changes in Phenol mass flux in an aqueous solution of ph= # 72# 73# 74# Immersion time, weeks Results: membranes made from only PDMS (71 and 72) have a much larger initial drop in mass flux than membranes with PTFS ( 73 and 74). The difference of coating thickness of 5 vs. >5 microns does not seem to have a large affect possibly because of surface polymers entering the PP non woven support.

17 Mass flux,mg/m2*sec,x10-3 Membrane Stability improvements under basic ph by the use of polytrifluoropropylmethylsiloxane (PTFS) and poly4-vinylphenol on non-woven PP support. Phenol mass flux vs. time #75 is PTFS #76 (8/2 PDMS/P4VP) #77 (6/4 PDMS/P4VP) 75# 76# 77# Dana's result =2.38 for silicone tubing with Exposition time in ph 13, weeks The results show for membranes 75,76 and 77 better long term stability and 75 and 76 higher phenol mass flux than membranes (previous slide). In all cases they have significantly higher mass flux then silicone tubing which is not expect to be stable under these basic conditions

18 The affect of selective membrane crosslinker concentration on composite Membranes ( 50K UF support): Phenol flux and Salt rejection (1000 ppm NaCl) (dialysis cell experiment) The results show that increasing the concentration of crosslinker TEOS, increases NaCl rejection without significantly hurting phenol permeability. The high salt rejections are needed for keeping the ph gradient. The results from Dana and Livingston were carried out with silicone tubing and have significantly lower ( 3 to 4 times) OMTC.

19 The affect of polyvinyl phenol (P4VP) concentration in the selective layer on phenol (9gr/L) flux (Support 50kDa UF) Under the concentration ratios studied there appears to be a small increase in mass flux and OMTC at an optimum ratio of PTFS:PVP of 8:2

20 Effect of adding P4VP on membrane permeability of phenol Under the concentration ratios studied there appears to be a small increase at an optimum ratio of PTFS:P4VP of 8:2 (50K UF support)

21 Selectivity membrane (PTFS on 50K UF) permeability used in MARS on OMWW feed from NF vs. RO concentrate Feed source Concentrate after NF with use NF-270 at 20 bar Concentrate after NF with use NF-270 at 20 bar Concentrate after NF with use NF-270 at 20 bar Concentrate after reverse osmosis RO-1 Concentrate after reverse osmosis RO-1 Concentrate after reverse osmosis RO-1 Concentrate after reverse osmosis RO-2 Concentrate after reverse osmosis RO-2 Concentrate after reverse osmosis RO-2 Feed components, ppm Permeate components, ppm Mass transfer flux, mg/m 2 *sec Tyrosol 14.7 Tyrosol * *10-6 Hydroxytyrosol 3.2 Hydroxytyrosol * *10-6 Gallic Acid 20.8 Gallic Acid * *10-6 Tyrosol 565 Tyrosol * *10-7 Hydroxytyrosol 853 Hydroxytyrosol * *10-7 Gallic Acid 82 Gallic Acid * *10-7 Tyrosol 840 Tyrosol * *10-8 Hydroxytyrosol 1385 Hydroxytyrosol * *10-8 Gallic Acid 69 Gallic Acid * *10-7 Overall mass transfer coefficient, m/sec Results: The membrane OMTC is significantly higher on NF feed streams than RO feed streams and there is selective passage of tyrosol and hydroxytyrosol compared to Gallic acid.

22 MARS Membranes Selectivity for different membranes with OMWW Membranes selectivity by biophenols recovery from OMWW in MARS GC-MS results (Membranes support is UF membrane 50kDa))

23 Selectivity membrane (PDMS: P4VP=8:2 on UF50kDA ) permeability used in MARS on concentrated OMWW feed from NF Feed source Feed components, ppm Permeate components, ppm Mass transfer flux, mg/m 2 *sec Overall mass transfer coefficient, m/sec Concentrate after NF with use NF-270 at 20 bar Tyrosol 14.3 Tyrosol * *10-6 Concentrate after NF with use NF-270 at 20 bar Hydroxytyrosol 5.2 Hydroxytyrosol * *10-6 Concentrate after NF with use NF-270 at 20 bar Gallic Acid 18.4 Gallic Acid * *10-6 Results: The membrane of PDMS/P4VP (8:2) did not change the significantly the OMTC as compared to the membrane of PTFS and did not have the same selective passage of tyrosol and hydroxytyrosol compared to Gallic acid.

24 Increasing the P4VP-Selectivity membrane #77 (PDMS: PV4P=6:4) on UF50kDa support permeability in MARS on concentrated OMWW feed from NF membrane Feed source Feed components, ppm Permeate components, ppm Mass transfer flux, mg/m 2 *sec Overall mass transfer coefficient, m/sec Concentrate after NF with use NF-270 at 20 bar Concentrate after NF with use NF-270 at 20 bar Concentrate after NF with use NF-270 at 20 bar Tyrosol 13.6 Tyrosol * *10-6 Hydroxytyrosol 3.8 Hydroxytyrosol * *10-6 Gallic Acid 22.4 Gallic Acid * *10-6 In comparing membranes 76 (previous slide) to 77 the further addition of PV4P increases the selectivity of the membrane for tyrosol and hydroxytyrosol compared to Gallic acid. Membrane 77 also has improved selectivity compared to membrane 75 (PTFS)

25 To improve further on permeability and selectivity: Effect of Tyrosol addition to selective layer of P4VP ( support 50kDa) on total phenol permeability (Feed is OMWW concentrate) PTFS PTFS/P4VP=8/2 P4VP+TYROSOL Total Phenol in MARS permeate, mg/l h 2 h 24 h MARS work time, hours The results show that total phenol permeability can be significantly enhanced by the addition of rational additives that form channels for phenol passage. The permeability of a selective layer of only P4VP is very low with little total phenol passage under the above conditions. The addition of tyrosol to the selective layer significantly enhances phenol permeability. This approach will be tried for PTFS and PTFS/P4VP selective layers. The affect on selectivity will also be investigated.

26 Conclusion and Future work The composite membranes with thin chemically (ph) crosslinked fluorinated silicone selective layers on a UF support were shown to have good chemical stability and permeability for polyphenols as compared to silicone tubing. Selectivity passage of tyrosol and hydroxytyrosol vs. Gallic acid was demonstrated. Methods for increasing permeability to phenols and high salt rejections which allows for the maintenance of a ph gradient across the membrane where demonstrated by choice of membrane composition. The results indicate that selective permeability of given phenols can be achieved by the use of specific modification of the selective barriers by formation of specific channels. This work is now under development for improving selectivity. As a 1 st goal OMWW valorization: We will develop a high permeability to the polyphenols hydroxytyrosol and tyrosol and a low permeability to the other components including acids, base, carboxylate polyphenols and non-aromatic carbon chains (oleuropein degradation products).