Sunteți pe pagina 1din 8

Separation and purification of lactic acid: A study on cross-flow nanofiltration as the final downstream process using different commercial

membranes
Document By: Bharadwaj Visit my website www.Engineeringpapers.blogspot.com More Papers and Presentations available on above site
Abstract This work is focused on the purification of fermentation broths using different commercial nanofiltration membranes in order to optimize the final downstream process for separation of lactic acid produced by fermentation. The processes of lactic acid production include two key stages, fermentation and product recovery. Lactic acid was produced in a 20-liter fermentor and the purification was achieved using a 0.01 m2 filtration unit. Pure sugar cane juice was fermented using Lactobacillus plantarum, NCIM 2912, for production of lactic acid. The cells were then separated from fermentation broth by microfiltration. The produced lactic acid was then purified from fermentation broth by nanofiltration. The commercially available NF2, NF3, and NF20 polyamide composite nanofiltration membranes on polyester backing with polysulfone substrate were used. The studies were performed in a pilot plant equipped with three cross-flow flat-sheet membrane modules arranged in parallel. The effects of transmembrane pressure and cross-flow velocity on the permeate flux, and on the retention of unconverted sugars were investigated. Experiments were performed at five levels of transmembrane pressure (5, 7, 9, 11 and 13 bars) where cross-flow velocity was varied within the range of 1.77 m/s to 2.48 m/s. Higher transmembrane pressure and cross-flow velocity yielded higher permeate flux. The pH value affected both rejection and permeates flux. Rejection increased with pH while flux decreased with this variable. Keywords: Nanofiltation, Lactic acid, Cross-flow module, Transmembrane pressure.

Introduction
In the recent years, there has been a spurt in the demand for lactic acid due to its application potentials in a wide range of fields like foods, pharmaceuticals, cosmetics and in production of biodegradable and biocompatible polylactate [1, 2]. Though it can be manufactured both in chemical synthesis process as well as in fermentation-based process, the latter predominates lactic acid manufacturing industry having the capability of producing desirable L-lactic acid instead of a racemic mixture of L and D lactic acids. But the major cost of lactic acid production is due to its downstream separation and purification steps. For effective exploitation of huge application potential of lactic acid,

efficient separation and purification at relatively low cost remains a challenge. The traditional downstream separation and purification involves quite a series of steps like filtration, acidification, neutralization, crystallization, carbon adsorption, evaporation, ion exchange etc. Such schemes not only involve harsh chemicals they are energy-intensive also [3, 4]. This is where membrane-based nanofiltration steps in as an entirely environmentally benign process [5]. Nanofiltration in downstream purification can replace the multiple purification steps by a single step while yielding a monomer grade lactic aid from a mixture of unconverted sugars and lactic acid. Separation of microbial cells has to be done in a prior microfiltration step. Such a membrane-based process that can be operated at relatively low pressure ranges of 6 to 20 bars vis a vis requirement of high transmembrane pressure (above 15-20 bars always) by reverse osmosis is gaining importance in the recent years. Separation mechanism draws heavily on Donnan-steric effects and understanding these effects is essential to successful modeling and scaling up of the process. To better understand the hydrodynamics of the system and the Donnansteric effects, the present study evaluated the effects of transmembrane pressure and cross-flow velocity on the permeate flux, and on the retention of unconverted sugars while comparing performances of three commercial nanofiltration membranes. The Sepro-made polyamide composite membranes (NF2, NF3, and NF20) on polyester backing with polysulfone substrate were procured in flat sheet forms and tested on crossflow module using fermentation broths directly from a membrane cell recycle bioreactor (MCRB) in the laboratory. The studies were performed in a pilot plant equipped with three cross-flow flat-sheet membrane modules arranged in parallel. The surface area of each flat sheet membrane was 0.01 m2. Experiments were performed at five levels of transmembrane pressure (5, 7, 9, 11 and 13 bar) where cross-flow velocity was varied within the range of 1.77 m/s to 2.48 m/s.

Experimental
Materials Membrane The microfiltration PVDF-MFB membrane was supplied by Sepro Membranes (Oceanside, CA 92056) as flat sheets. The important characteristics, as provided by the supplier, are pore size of 0.13 m and normalized water flux 1800 L /m 2 hr bar at pH 7.5 and 250C. The nanofiltration NF2, NF3 and NF20 membranes were also supplied by sepro membranes (Oceanside, CA 92056) as flat sheets. All are made of polyamide, thinfilm composite membranes on polyester backing with a polysulfone substrate. Other important characteristics for different membranes, as provided by the supplier, are 97%, 98%, and 98% rejection of MgSO4 and 40%, 50%, 35% rejection of NaCl respectively (for [MgSO4] and [NaCl] = 2 g/L and P = 10.3 bar), and water flux of 135, 42, 42 L/m 2 hr respectively at specified pressure at 250C and pH 7.5. The same piece of membrane was used throughout the experiments. Fermentation broth preparation Experiment was carried out using microfiltrate fermentation broth (MFB). Pure sugar cane juice (14% w/v), supplied by local farmer, contents 126 sucrose as disaccharides, 8

g/l glucose and 6 g/l fructose as monosaccharide were supplemented with the following medium: peptone 5 g/l , yeast extract 10 g/l, polysorbate80 2 ml/l, Sodium acetate 1.5 g/l ,NH4HCO3 0.75 g/l , MgSO4.7H2O 0.2 g/l , MnSO4.4H2O 0.005 g/l , KH2PO4 1.5 g/l, K2HPO4 1.5 g/l at 370C [2, 6]. All the chemicals except yeast extract (Hi-Media, Mumbai, India) were supplied by Merck Limited (Mumbai, India). A selected strain of Lactobacillus plantarum NCIM 2912 (National Chemical Laboratory, Pune, Maharashtra, India) was used for fermentation. The fermentation broth was then microfiltrate using PVDF-MFB membrane. Methods Experimental method Fig.1. is a schematic diagram of the cross-flow membrane module, fabricated by supercritical solutions, Kolkata, India. Only the relevant module was used for nanofiltration study has been shown here. Effective surface area of the flat sheet membrane was 0.01 m2. The membrane was compacted before each experiment by filtering high-purity water at 15 bars until it reached a constant permeability. The feed was pumped through diaphragm pump (Hydra-cell pump, Minneapolis, MN 55403 USA) from a 20 liters feed vessel, kept at 370C, the temperature of the fermentation broth, into the cell and flowed tangentially to the membrane. A stainless steel control valve is mounted on the retentate outlet to control the transmembrane pressure which was monitored through two digital manometers located on the inlet and outlet of the membrane module. Retentate was recycled back to the feed tank. A volume of 10 ml of permeate was collected for each pressure and timed to estimate the permeation flux. After each run the membranes were cleaned. Fully open the recirculation and permeate valves and flush with tap water for 5 min. Circulate 5 l of 2% potassium metabyphophite for 10 min and again rinse with tap water for15 min. Finally rinse with distilled water for 5 min.

Fig. 1. Experimental set-up for nanofiltration Analytical method The concentrations of lactic acid and residual sucrose, glucose, fructose were determined by High Performance Liquid Chromatography (C18 column, Perkin Elmer, Series 200, USA) equipped with UV- VIS detector and Refractive Index Detector (Perkin Elmer,

Series 200, USA). The column temperature was set to 300C, and the mobile phase was 70% acitonitrile (Sigma Aldrich) and 30% ultrapure water (Milli-Q) at a flow rate of 0.8 ml/min. An injection volume of the sample was 50 l.

Results and discussion


Permeate collection and measurement were done after running the cross flow membrane module (Fig.1.) with ultra pure water (Milli-Q) for first 30 minutes under a constant pressure of 13 bars to allow for complete wetting and normal compaction of the membrane under normal nanofiltration pressure regime. The pure water flux and flux of fermentation broth were then measured in our experimental cross flow membrane module over a trans-membrane pressure range (TMP) of 5-13 bars and cross-flow velocities varying from 1.77 to 2.48 m/s. The filtration area of the membrane was 0.01 m2. Pure water flux was calculated as PWF (litre/m2hr) = Q (amount of water permeated through membrane, liter) / [A (effective membrane cross sectional area, m2) T (sampling time, hr)]. Permeate flux (PMF) of the lactic acid fermentation broth was calculated using the same relation.

400 Water flux/ Broth Flux (L/m2hr)

300 NF2/PWF NF3/PWF NF20/PWF NF2/FBF NF3/FBF NF20/FBF

200

100

0 0 5 10 15 Transm em brane Pressure (bar)

Fig. 2. Effect of trans-membrane pressure on pure water flux, PWF (___) and fermentation broth flux, FBF (----) for different membranes at 370C and 2.48 m/s. Flux study results as presented in Fig. 2 show that all the three membranes (NF2, NF3 and NF20) show increase of pure water flux (PWF) as TMP increases from 5 bars to 13 bars. The NF2 exhibits the highest flux of 360 L/ m 2 hr at 13 bar pressure and NF20 exhibits the lowest flux and the highest attained value being 135 L/ m2 hr at 13 bars. For all selected membranes fermentation broth flux were lower than the pure water flux due to osmotic pressure difference induced by the separation, higher viscosities of the fermentation broth than water and reversible fouling of the membranes [7]. After certain

pressure (9 bar), no significant effect on permeate flux was observed due to cake formation over the membrane. Compared to all the other membranes, although NF3 started from low microfiltrate broth permeate flux exhibits the highest microfiltrate broth permeate flux, 111 L/m2 hr at 13 bar, in contrast, NF20 showed the lowest broth permeate flux (78 L/ m2 hr) at the same pressure. Broth fluxes of NF2 and NF3 membrane are nearly same. These results help in setting the best membrane operating cross flow velocity at around 2.48 m/s and 13 bar pressure for NF3 membrane. The flux obtained by NF3, 111 L/m2 hr at 13 bar, is suitable for industrial application as this flux are higher than those obtained in literature [7, 8,9].

120 100 Broth Flux (L/m2 hr) 80 60 40 20 0 NF2 NF3 NF20 1.77 2.48

Fig. 3. Effect of cross-flow velocity on fermentation broth permeates flux at constant transmembrane pressure of 13 bars for different membranes. As shear rate or cross-flow rate plays an important role in reducing membrane fouling, optimization of cross flow is essential in minimizing membrane area requirement [7, 8]. Fig. 3 shows that flux increases with increase of cross flow velocities for all selected membranes. This is because of greater convective force that minimizes cake formation and concentration polarization effects. Though compared to all the other membranes, NF3 exhibits the highest broth permeate flux (111 L/ m2 hr) at 13 bars and 2.48 m/s crossflow velocity. Lactic acid and sugar rejection was calculated as rejection (%) = [1- (concentration of component in permeate / concentration of component in the feed stream)] 100. Fig. 4 shows lactic acid and sugar rejection as a function of permeate flux of microfiltrate fermentation broth. For both component fermentation broth rejection increases with increasing transmembrane pressure. The observed phenomena can be explained as the pressure increases flux also increases resulting a dilution of the components in permeate, though retentate concentration did not change significantly [8]. Lactic acid rejection for microfiltrate fermentation broth increased in NF3 membrane from 53% to 67% when flux increased from 69 to 111 L/ m2 hr and transmembrane pressure increased from 5 bars to

13 bars. The increased rejection of selected membranes were observed in the order of NF3>NF2>NF20.

100
Lactic acid/ Sugar rejection (%)

90 80 70 60 50 40 0 5
TMP (bar) NF2/LACR NF3/LACR NF20/LACR NF2/SUGR NF3/SUGR NF20/SUGR

10

15

Fig. 4. Rejection of lactic acid (____) and sugar (-----) of selected membranes of microfiltrate fermentation broth as a function of transmembrane.

100 90 80 Rejection (%) 70 60 50 40 30 20 10 0 2.7 (LA) 2.7 (SUG) pH 5.5 (LA) 5.5 (SUG) NF2 NF3 NF20

Fig. 5. Comparative lactic acid rejection of the selected membranes as a function of pH at 13 bars pressure. In, Fig 5, the effect of pH on lactic acid rejection is compared for the three membranes considered. The fermentation broth was acidified to pH 2.7, and the microfiltrate fermentation broth pH was 5.5. The results shown were obtained at a constant transmembrane pressure of 13 bar. It can be observed that rejection increases with pH for

three membranes. The membrane charge becomes more negative with increase in pH resulting increase in electrostatic repulsion and higher rejection of lactic acid (negatively charged).

Conclusions
When the microfiltrate fermentation broth was treated by NF the electrostatic effect is a limiting factor as at the acidic pH lactic acid rejection was low. Lactic acid transport through three membranes was affected by pH. Permeate flux increased with pressure. NF3 membrane showed 96% sugar removal and 67 % lactic acid rejection at 13 bar pressure attaining flux 111 L/ m2 hr. The successful integration of the membrane with a cross-flow module in a continuous lactic acid production and purification system paves the way for scale up of the hybrid system not only for production of lactic acid but for other similar organic acids as well.

Acknowledgements
The authors are thankful to the Department of Science and Technology, Government of India (DST) for the grants under DST-FIST Program (SR/FST/ET1-204/2007) and Green Chemistry/Technology Program (SR/S5/GC-05/2008) with which the infrastructure for the present research was developed and necessary research materials were procured. References: 1. Datta R., Henry M., Lactic acid: recent advances in products, processes and technologies- a review, J. Chem. Tech. and Biotech., 81 p.1119-1129 (2006). 2. Timbuntam W. , Sriroth K., Tokiwa Y., Lactic acid production from sugarcane juice by a newly isolated Lactobacillus Sp., Biotechnol. Lett. 28 p. 811-814 (2006). 3. Joglekar H.G., Rahman I., Babu S., Kulkarni B.D., Joshi A., Comparative assessment of downstream processing options for lactic acid, Separation. and Purification Technol.., 52 p.1-17 (2006). 4. Gonzalez M. I., Alvarez S., Riera F., Alvatez R., Economic evaluation of an integrated process for lactic acid production from ultrafiltered whey, J. of Food Engg, 80 p.553-561 (2007). 5. Pal P., Sikder J., Roy S., Giorno L., Process Intensification in lactic acid production: a review of membrane-based processes Chemical Engineering and Processing: Process Intensification, DOI: 10.1016/j.cep.2009.09.003, (in press) 6. Fu W., Mathews A.P., Lactic acid production from lactose by Lactobacillus plantarum: kinetic model and effects of pH, substrate and oxygen, Biochem. Engg. J., 3 p.163-170 (1999). 7. Bouchoux A., Balmann H. R., Lutin F., Investigation of nanofiltration as a purification step for lactic acid production processes based on conventional and bipolar electrodialysis operations, Sepa. and Purif. Techl., 52 p. 266-273 (2006). 8. Gonzalez M. I., Alvarez S., Riera F. A., Alvarez R., Lactic acid recovery from whey ultrafiltrate fermentation broths and artificial solutions by nanofiltration, Desalination 228 p. 84-96 (2008).

9. S.H. Kang, Y. K. Chang, H. N. Chang, Recovery of ammonium lactate and removal of hardness from fermentation broth by nanofiltration, Biotech. Progress 20 p. 764-770 (2004).

Document By: Bharadwaj Visit my website www.Engineeringpapers.blogspot.com More Papers and Presentations available on above site

S-ar putea să vă placă și