Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES
Article
Production of L (+) glutamic acid in a fully membrane-integrated hybrid reactor system: direct and continuous production under non-neutralizing conditions Vikramachakravarthi D, Ramesh Kumar, and Parimal Pal Ind. Eng. Chem. Res., Just Accepted Manuscript • Publication Date (Web): 17 Nov 2014 Downloaded from http://pubs.acs.org on November 18, 2014
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Industrial & Engineering Chemistry Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
1 2
Production of L (+) glutamic acid in a fully membrane-integrated hybrid reactor system: direct and continuous production under non-neutralizing conditions
3
Vikramachakravarthi D.1, Ramesh Kumar1, Parimal Pal*
4 5 6
Environmental and Membrane Technology Laboratory, Department of Chemical Engineering, National Institute of Technology Durgapur, India – 713209
___________________________________________________________________________
7
ABSTRACT
8
Experimental investigations were carried out on continuous and direct production of L-
9
glutamic acid in a hybrid reactor system that integrated conventional fermentative production
10
step with downstream membrane-based separation and purification in flat sheet cross flow
11
membrane modules. To ensure constant fermentor capacity, the system was operated at
12
steady state at different set flux values ranging from 76 L m-2 h-1 to 90 L m-2 h-1 at a dilution
13
rate of 0.15 h-1. Overcoming the substrate-product inhibitions of traditional batch production
14
system, this new, compact, flexible and largely fouling-free design ensured steady and
15
continuous production of L-glutamic acid directly from a renewable carbon source at the rate
16
of about 8.4 g L-1 h-1. Provisions of continuous product withdrawal, separation and recycling
17
of unconverted sugar and microbial cells ensured almost inhibition-free production under
18
high cell concentration. Well-screened nanofiltration membranes with high selectivity helped
19
achieve over 97% product purity while ensuring recovery and recycle of more than 95%
20
unconverted carbon source resulting in high yield of 0.95gg-1. The direct production scheme
21
involves no phase change and use of harsh chemicals. The study presents development of a
22
green process of glutamic acid production for sustainable business.
23 24
Keywords
25 26 27 28
Glutamic acid; Continuous process; Direct production; Membrane separation; Cross-flow module; Green design ___________________________________________________________________________ *Corresponding Author (P. Pal) email:
[email protected]/
[email protected]
29 30 31
Tel.: +91 343 2755955 Mobile: +91 943446950 Fax: +91 343 2547375 1
Equal contribution
32 33 34 35 1 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1 2 3
1. INTRODUCTION
4
In the recent years, fermentative production of L-glutamic acid (GA) or 2-aminopentanedioic
5
acid has roused interest among researchers due to its potential application in seasoning
6
throughout the world.1 A wide variety of speciality chemicals are produced from GA as a
7
starting material.2 It has the ability to decrease or prevent nerve damage caused by anticancer
8
drug. GA can also improve the function of nervous centralize and cortical brain for
9
neurasthenia patients. Poly glutamic acid (PGA) is a naturally occurring polymer of GA that
10
is biodegradable, edible and non-toxic towards human and environment.3 GA has the greatest
11
demand of any single amino acid, exceeding 1.5 million tons per year on global basis.4
12
Manufacturing processes of amino acids are categorized as fermentative, enzymatic and
13
extractive (from acid hydrolysates of animal or plant protein). Currently, most of the GA is
14
produced by fermentation. Out of the two optically active forms of GA, only L (+) GA with
15
high optical purity can be used to produce monosodium glutamic acid and PGA.5 Some
16
researchers have tried to immobilize the whole microbial cells in calcium alginate or agar for
17
continuous production of GA6. But product concentration remains low due to inefficient
18
mass transfer, leakage of cells and lack of general matrix for immobilizing different cells. All
19
these make the process unattractive. The carbon source from renewable resources with
20
suitable micro-organisms has always been favoured to produce GA through fermentation
21
route as a myriad of value-added products derived from biological origin are readily accepted
22
by food industries and consumers.7 For fermentative production, the most widely used carbon
23
sources have been carbohydrates ranging from glucose to starch8. Materials like molasses,
24
corn starch and whey are considered as cheap carbon sources and have been frequently
25
examined in several studies. But these substrates demand additional treatments so to avoid
26
membrane fouling9. On the other hand; sugarcane juice has been tried very little though it
27
could be a very promising carbon source being clean, relatively cheap and renewable.
28
Throughout the year, sugarcane juice is easily available in some major sugarcane growing
29
countries like India and Brazil10. Addition of nutrient supplementation like yeast extracts
30
leads to increase sugar utilization and reduced fermentation time but it also adds to residual
31
impurities in fermentation broth11.To produce pure and monomer grade GA, efficient
32
separation of other impurities from the fermentation broth is essential. Conventional
33
purification scheme involves a number of downstream treatment steps like precipitation,
34
filtration, acidification, extraction, neutralization, carbon adsorption, crystallization and 2 ACS Paragon Plus Environment
Page 2 of 23
Page 3 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
1
evaporation.1 Traditionally, crystallization method has been used for GA separation from
2
fermentation broth at low temperature with or without biomass. But 1-2% GA still remains in
3
isoelectric supernatant. Ion-exchange process has been used for removal of residual GA from
4
isoelectric supernatant but substantial amounts of acid and base are required in regeneration
5
of ion-exchange resins7. In addition to that, conventional batch fermentation also suffers from
6
low volumetric productivity due to both substrate and product inhibitions. It also involves
7
high labour cost due to frequent shutdown and start-up of batch process. These problems may
8
be overcome by using membrane-integrated cell recycle bioreactors with continuous
9
fermentation which promise high cell density, much higher productivity and higher acid
10
concentration in the final product in a continuous process. To separate amino acids from
11
fermentation broths, product properties such as solubility, molecular size and affinity to
12
adsorbent and charge characteristics may be utilized12. The most vital criteria for continuous
13
fermentation are steady state operation with prolonged exponential growth phase which needs
14
to be maintained with proper cell bleeding13. Membrane separation is expected to be one of
15
the most promising candidates in successfully separating target amino acids from large
16
amounts of other impurities such as fermentation medium14. Several attempts have been
17
made over the last two decades in this direction of integration of traditional fermentor with
18
membrane based separation and purification15-17. However, attempt has been made to
19
separate product in a single stage resulting in serious membrane fouling problems after a
20
short period of operation.
21
It has been observed that mass transfer parameters of dissociated and un-dissociated
22
forms of any organic acid through NF membranes are better than those through reverse
23
osmosis membranes18. Ultrafiltration membrane modules have been used in some studies for
24
complete separation of proteins and cells, but not without encountering frequent problems of
25
membrane fouling19. Two-stage membrane separation has been attempted in very few cases.
26
Other methods like electrodialysis that use anion-exchange and cation-exchange membranes
27
have generally not been successful due to significant pH changes in the feed solutions, being
28
undesirable for enzyme reaction when the electrodialysis separation process is incorporated
29
into reactors20, 21. In addition, anion-exchange membranes have been reported to be easily
30
contaminated probably due to an oxidative reaction of a sulfhydryl group present in the
31
amino acids. Studies on multistage membrane cell recycle bioreactor also have been carried
32
out by a few researchers to improve concentration and productivity, but they have used pure
33
glucose only as carbon source22.
3 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1
In many chemical, pharmaceutical, food and biotechnological processes, the purification
2
and recovery of amino acids plays pivotal role.23 A limited studies have been conducted on
3
the efficiency of NF membrane in the context of separation of amino acids. Even these
4
studies considered only model solutions of amino acids24. The major technological barrier in
5
cost-effective production of high purity GA is its downstream separation and purification.
6
And this is where membrane-based processes are stepping in in a big way. With integration
7
of downstream membrane-based purification in production of organic or amino acids,
8
possibility of development of better processes have certainly improved over the years but
9
serious attention still needs to be paid to some areas to evolve a smaller, more compact, more
10
flexible and less energy-intensive plant that could guarantee large scale production of a
11
highly demanding chemical product in an environmentally benign process.25
12
Integration of appropriate membranes as well as modules in fermentor for cell separation
13
and product purification along with provision of logical sequencing of operations are
14
essential in truly achieving such process intensification. This study develops a continuous,
15
GA production process under non-neutralizing conditions in a two-stage membrane
16
integrated system. Judicious combination of microfiltration (MF) membranes in the first stage
17
followed by nanofiltration (NF) membranes in the second stage in flat sheet cross flow
18
membrane module with provision of instant separation of acid from the production media
19
eliminates the need for pH adjustment. To our knowledge, this kind of study related to
20
development of a green and continuous production process for GA in a fully membrane
21
integrated system has not been reported earlier.
22
2. MATERIALS AND METHODS
23
2.1. Microorganism
24
Corynebacterium glutamicum (NCIM 2168) is a Gram positive, facultative anaerobic,
25
heterotrophic GA producing bacterium used throughout this work was brought from National
26
Collection of Industrial Microorganism (NCIM), National Chemical Laboratory (Pune, India)
27
in lyophilized condition. The culture was maintained in nutrient agar slants at 4 oC and
28
subcultured twice a month. Medium for preparing agar slant consisted of 2 g L-1 yeast extract,
29
1 g L-1 beef extract, 5 g L-1 peptone, 5 g L-1 sodium chloride, 15 g L-1 agar at a pH of 7.0. The
30
flask was incubated overnight at 30 oC and used as primary inoculum. Seed culture medium
31
was used with the composition (g L-1): urea, 5; yeast extract, 4; K2HPO4, 1; MgSO4·H2O, 0.5;
32
FeSO4·7H2O, 0.02; MnSO4·H2O, 0.01; biotin (5µg L-1); thiamine HCl (80 µg L-1) and
33
aqueous medium was sugarcane juice containing 11.0% (w/v) fermentable sugar. Biotin, 4 ACS Paragon Plus Environment
Page 4 of 23
Page 5 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
1
thiamine-HCl and urea were cold sterilized by membrane filter (0.2 µm, PALL Life
2
Sciences), whereas rest of the medium was sterilized separately by autoclaving at 15 psi (121
3
o
4
inoculated sugarcane juice with this strain maintained at 30 oC overnight was used as
5
inoculum in membrane integrated reactor system.
6
2.2. Membranes
7
Four different types of NF membranes namely, NF1, NF2, NF3 and NF20 were purchased
8
from Sepro Co. (USA) as flat sheets. These polyamide membranes are of thin-film composite
9
on polyester backing with a polysulfone substrate having different isoelectric points and
10
permeability characteristics. The detailed characteristics and properties of these membranes,
11
as provided by the supplier and collected from literature have been presented in Table 1.
12
2.3. Production medium preparation
13
For the preparation of fermentation media, pure sugarcane juice was purchased from local
14
farmers. Pure sugarcane juice contained 11.0% (w/v) fermentable sugar (97.7 g L-1, 7.2 g L-1,
15
and 5.35 g L-1 sucrose, glucose and fructose respectively) which was pre-filtered to remove
16
suspended particles like fibres and solids. The composition of production medium was same
17
as the seed culture medium but the urea and biotin concentration were 8 g L-1 and 1 µg L-1
18
respectively and plus tween 80 (0.1 mL L-1). Other conditions like temperature, pH and
19
sterilization parameters were also same. Chemical reagents used were purchased from Sigma
20
Aldrich, Merck India Ltd. and Loba.
21
2.4. Experimental equipment
22
Fermentation was carried out in a 30 L pilot plant fermentor which was fabricated with high
23
grade stainless steel (SS316) and was provided with water circulation system (Polyscience,
24
U.S.A.) and thermocouples for measuring and controlling temperature. There was also
25
provision for O2 gas purging system (Fig. 1). Agitation speed was set at 250 rpm whereas
26
temperature was maintained at 30 oC. Measurement and monitoring of pH and dissolved
27
oxygen (DO) concentration were done using pH-Ion meter (ORION, 4star, Thermo, U.S.A.)
28
and DO meter (TOSHBRO, India) attached to the fermentor. A flat-sheet MF membrane
29
fitted in a cross-flow module (having 0.01 m2 surface area) was attached to the fermentor. For
30
monitoring transmembrane pressure, two diaphragm pressure gauges were attached to the
31
inlet and outlet of each module. A peristaltic pump (Entertech, India) was used for circulating
C) for 15 min. To reduce the lag phase of C. glutamicum (NCIM 2168) in fermentation, 2 L
5 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1
the fermentation broth through the membrane module. The desired transmembrane pressure
2
was maintained during MF by controlling the attached diaphragm valve and pressure gauge.
3
For collecting the microfiltrate, a collecting vessel of 15 L was assembled with the system. A
4
series of flat sheet cross flow NF modules was arranged within the system in such a way that
5
the solution both from the fermentor and holding vessel could be directly filtered by the
6
respective module. A high pressure diaphragm pump (Hydra-cell pump, 2.2 kW,
7
Minneapolis, U.S.A.) was used between the holding vessel and NF module for circulation of
8
microfiltered fermentation broth under the desired transmembrane pressure. NF module was
9
fitted with three cross-flow membrane modules (each of 0.01 m2 surface area).
10
Microfiltration of the fermentation broth was carried out with a Nylon 0.45 µm membrane
11
(Membrane Solutions, U.S.A.). Composite polyamide NF1, NF2, NF3 and NF20 membranes
12
(Sepro Membrane, U.S.A.) were used in the NF unit. Before starting the experiment, the
13
fermentor was autoclaved at 121 oC and 15 psi pressure for 15 min and its tubing was
14
sterilized with ultrapure water at 60 oC till neutralization. Membrane cleaning was done after
15
each run with 0.1 N NaOH and 0.01 M HNO3 solution respectively. The membranes were
16
also sterilized with 200 mg L-1 NaOCl solution followed by rinsing with Milli-Q water
17
(MILLIPORE Pvt. Ltd, India).
18
2.5. Fermentation
19
Fermentation was started with 10% (v/v) of inoculum in an 18 L total volume of prepared
20
medium in the membrane-integrated bioreactor of 30 L volume. Fermentation was carried out
21
at a temperature of 30 oC and pH of 6.5 under non-neutralizing conditions. In addition to
22
stirring at 250 min-1, the sterile air flow rate through the spurge was set at 0.1 vvm with the
23
help of a pump attached to 0.22 µm disc filter (Millipore). These conditions were optimum
24
for selective production of GA because very low oxygen transfer rate favours pyruvic acid
25
and lactic acid production whereas higher oxygen transfer rates favour production of α-
26
ketoglutaric acid and succinic acid
27
recycling was initiated via microfiltration membrane module that separated cells and GA
28
from the fermentation broth. By this time period, culture grown inside the reactor was in the
29
last stage of exponential phase and that was the ideal time to make the system continuous
30
while maintaining the same active growth phase of the bacteria by adding new fermentation
31
media. Through continuous removal of GA, product-inhibition could be minimised without
32
the need for adjusting pH with addition of caustic material. The working volume of the
33
reactor was maintained at a constant level by supplying feeding medium in balance with
26
. After an initial start-up period of some 24 h, cell
6 ACS Paragon Plus Environment
Page 6 of 23
Page 7 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
1
permeate flow from the membrane module. Effects of feed dilutions were investigated during
2
continuous fermentation. Dilution rate (h-1) is defined as the ratio of the feed stream inflow
3
rate to the working volume of the reactor. This is the rate at which the fresh media is added to
4
the fermentation broth of the reactor. Productivity (g L-1 h-1) is the final product concentration
5
of GA (g L-1) multiplied by the dilution rate (h-1) in the system. This indicates the plant
6
performance in terms of GA output per unit time. The cross-flow velocity (m s-1) through the
7
membrane may be computed by dividing volumetric flow rate of fluid through membrane
8
module (m3 s-1) by the cross sectional area of the inlet pipe of that module (m2). To minimize
9
membrane fouling problem, cross-flow flat sheet MF modules were operated for continuous
10
10 h in a constant permeate flux mode below the critical flux with partial cell bleeding. The
11
samples were collected and analysed at regular interval of time to check the physiological
12
properties of the broth and product. Retentate from NF was recycled back to holding vessel to
13
make solution properties inside the holding vessel constant. During continuous operation,
14
permeate flow rate from MF unit was tuned in such a way that NF permeate flow rate
15
equalled the fresh feed flow rate to reactor.
16
2.6. Assay
17
Samples were collected at regular intervals then ultra-centrifuged (Sigma Instruments) at
18
12,000 rpm for 15 min and supernatants were collected for analysis of GA, sucrose, glucose
19
and fructose. Pellet was re-suspended with Milli-Q water with same volume and measured for
20
cell growth by taking optical density (OD) through UV spectrophotometer (Agilent, Cary 60,
21
U.S.A.) at 660 nm. Concentration of GA with all carbohydrates was measured using High
22
Performance Liquid Chromatography (Agilent, 1200 series). Analysis of GA was done by
23
MCI GEL CRS 10W column (Mitsubishi Chemical Corporation, Japan) with mobile phase
24
was 2 mM CuSO4 at flow rate 1 mL min-1 with Diode Array Detector (254 nm) at retention
25
time (RT) 16 min. The injected sample volume was 5 µL after filtration with 0.2 µm Nylon
26
syringe filter membrane (13 mm diameter). The measurement of all carbohydrates (sucrose,
27
glucose and fructose) was done by Refractive Index Detector with Agilent Zorbax
28
Carbohydrate Analysis Column at column temperature 30 oC with mobile phase acetonitrile:
29
water (Milli-Q) (75:25 ratio) at a flow rate 1.4 mL min-1 with sample injection volume 10 µL.
30
The RT of fructose, glucose and sucrose were 4.08 min, 5.32 min and 8.12 min respectively.
31
Minerals (Na+, K+ and Ca+) were quantified with Flame Photometer (Labard, India). Purity of
32
the nanofiltrated sample was determined by peak purity software tool of HPLC (Agilent,
33
series 1200). 7 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1
3. RESULTS
2
3.1. Constant permeate flux filtration runs during fermentation
3
3.1.1. Effects of transmembrane pressure on flux during microfiltration
4
The cross-flow module integrated with flat-sheet MF membrane successfully separated cells
5
(as evident in cell analysis of the permeate) for recycling while generating clear permeate
6
from the fermentation broth without significant flux decline over 24 h. Filtration flux during
7
MF study varied with both transmembrane pressure as well as cross-flow velocities as shown
8
in the Fig. 2a. During MF, operating transmembrane pressure was varied from 0.5 bar to 3.5
9
bar and experiment was carried out at three cross-flow velocities of 0.55, 0.90 and 1.25 m s-1
10
with corresponding volumetric flow rates of 150, 250 and 300 L h-1 respectively. The cross-
11
flow velocities (m s-1) were calculated as volumetric flow rate (m3 s-1) of the retentate per
12
cross sectional area. Maximum fermentation broth flux during microfiltration as obtained at
13
2.5 bar in the case of 1.25 m s-1 cross-flow velocity was 135 L m-2 h-1. The permeate fluxes of
14
fermentation broth with increasing pressure up to 2.5 bar at 0.55 m s-1 cross-flow velocity
15
increased almost linearly but at higher cross-flow velocities like 0.9 and 1.25 m s-1 the
16
relationship curves turned quite sigmoid as shown in the Fig. 2a. The effect of time (up to 24
17
h) of operation had been investigated of fermentation broth flux at fixed transmembrane
18
pressure 2.5 bar at three cross-flow velocities of 0.55, 0.9 and 1.25 m s-1 as shown in the Fig.
19
2b. For over 24 h, the fermentation broth flux declined with the increase of cross-flow
20
velocities and it was minimum at 0.55 m s-1 lowest cross flow velocity. Moreover, the steady
21
state flux could be maintained for longer time at lower cross-flow rate than higher one at
22
fixed transmembrane pressure. Thus regulation was more prominent in case of higher cross-
23
flow rate. At higher operating cross-flow velocity, greater sweeping action resulted in higher
24
flux than that attained during the operation at lower cross-flow velocity under identical
25
transmembrane pressure. Higher cross-flow velocity reduces membrane fouling by higher
26
sweeping action up to certain level but beyond certain level, further increase in cross-flow
27
rate induces higher membrane fouling by virtue of higher hydrostatic pressure, adsorption on
28
membrane surface and pore blocking.
29
During MF run, it was essential to establish a continuous operational pattern by
30
maintaining the same reactor volume as well as same holding vessel volume under constant
31
permeates flux. To achieve a steady state condition of constant permeate flux below critical
32
flux was very convenient in clarification process involving two stage membrane filtration.27 8 ACS Paragon Plus Environment
Page 8 of 23
Page 9 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
1
The flux behaviour of the membrane modules at constant transmembrane pressure and at
2
different cross-flow velocities was investigated and it was observed that the steady state flux
3
could be attained after 3 h, 4 h and 6 h of operation for 0.55 m s-1 (dilution rate = 0.15 h -1),
4
0.9 m s-1 (0.17 h -1) and 1.25 m s-1 (0.18 h-1) cross-flow velocities respectively as shown in
5
Fig. 2b. The steady state fluxes were 76 L m-2 h-1, 84 L m-2 h-1 and 90 L m-2 h-1 for the above
6
mentioned three cross-flow velocities.
7
To operate the MF module at set flux rate, transmembrane pressure had to be adjusted at
8
regular interval of time for each cross flow velocities. In case of higher cross-flow velocity
9
like 1.25 m s-1 to maintain higher flux over a period of time, increase of transmembrane
10
pressure with time was quite sharper than the others as under higher operating pressure with
11
higher set flux, possibility of biomass adsorption on the membrane surface increases. Thus
12
high flux due to higher cross-flow velocity was translated to higher rate of membrane fouling
13
beyond a certain level. Cell viability and other physiological condition of C. glutamicum
14
(NCIM 2168) were not affected by variation of the operating conditions of three cross flow
15
velocities.
16
3.1.2. Effects of transmembrane pressure on flux during nanofiltration
17
Permeate collection and measurements were done after running the cross-flow NF membrane
18
module with four different types of NF membranes. The pure water flux and flux of micro-
19
filtered fermentation broth were measured in cross-flow membrane module over a
20
transmembrane pressure range of 2.5-15 bar at volumetric cross-flow velocity 4.44 m s-1. A
21
positive correlation between flux and transmembrane pressure has been observed in the Fig.
22
3. Transmembrane pressure (TMP) up to 15 bars, flux behaviour for fermentation broth was
23
in line with that of pure water flux. At lower levels of TMP, osmotic pressure difference
24
plays significant role in flux behaviour13. Pure water flux was found to vary linearly with
25
transmembrane pressure indicating absence of significant fouling. The NF2 exhibited highest
26
flux 350 L m-2 h-1 at 15 bar pressure whereas NF1 exhibited the lowest flux with the highest
27
attained value being 116 L m-2 h-1 at 15 bars transmembrane pressure indicating that NF2
28
membrane was the loosest membrane and NF1 was the tightest membrane (Table 1). Similar
29
trend was achieved during the NF of microfiltered fermentation broth with four different NF
30
membranes. At 15 bar transmembrane pressure, maximum and minimum fluxes obtained
31
were 220 and 70 L m-2 h-1 by NF2 and NF1membranes respectively. In this investigation,
32
operating pressure of 15 bars at cross-flow velocity of 4.44 m s-1 was optimum to maintain
33
system stability. Modules should be operated at their critical levels of cross-flow and this is 9 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1
essential to minimizing membrane surface area as well as shear rate or cross-flow rate that
2
plays an important role in reducing fouling28.
3
1.3. Effect of transmembrane pressure on glutamic acid and sugar rejection
4
GA and sucrose rejection was calculated as R (%) = [1 – (Cp/Cf)] × 100, where R = rejection;
5
Cp = concentration of GA or sucrose in permeate; Cf = concentration of GA or sucrose in
6
feed. Fig. 4 (a) and (b) show GA and sucrose rejection as a function of transmembrane
7
pressure of microfiltered fermentation broth. It was observed that the rejection increased with
8
increasing transmembrane pressure for both components. GA rejection by NF membrane
9
increases linearly with increasing transmembrane pressure and the maximum rejection of
10
97% by NF1 is observed at 15 bar transmembrane pressure (Fig. 4a). During nanofiltration
11
with respect to sucrose, NF1 and NF20 membranes show maximum rejection whereas the
12
NF2 membrane retains the minimum (Fig. 4b). Significant differences are observed in
13
rejection behaviours of the used nanofiltration membranes for sucrose and glutamic acid
14
because of the neutral and ionic characteristics of the two major target solutes here as evident
15
in Fig.4a and 4b. NF20 membrane rejects very little GA (only 15%) allowing more than 85%
16
to the permeation side facilitating its recovery with the permeate stream. The same NF20
17
membrane at a transmembrane pressure of 15 bars, rejects sucrose by 95% for its continuous
18
recycle to the fermenter. As sugar is neutral solute, its retention is only due to steric effects.
19
Sugar retention depends on membrane pore radius and hydrodynamic radius or a combination
20
of the two. The variation of retention by the selected membranes was due to variation of the
21
mean pore sizes of the membranes. Sugar rejection was largely found to remain independent
22
of pH of the medium. However, rejection increased with the increase of transmembrane
23
pressure in all cases. This may be attributed to the solution diffusion mechanism that applies
24
to NF membrane. During the diffusion of solution through a NF membrane, solute flux and
25
solvent flux remain uncoupled and hence increased solvent flux at a higher transmembrane
26
pressure does not result in increase of solute flux. Rather, increased solvent flux results in
27
increased obstruction to the transport of solute through the membrane. Thus NF membrane
28
should be selected for downstream processing of fermentation broth in such a way that they
29
should offer low rejection of GA and a high rejection of sugars which makes NF20
30
membrane as the best choice here.
31
3.2. Continuous fermentation with cell recycles by microfiltration and nanofiltration
32
Fermentation was done at 30 oC in a 30 L pilot plant fermentor using 20 L of its working
33
volume (18 L fermentation media and 2 L inoculum, C. glutamicum NCIM 2168). 10 ACS Paragon Plus Environment
Page 10 of 23
Page 11 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
1
Microorganisms took some 24 h to reach an active stage of exponential growth phase. Lag
2
phase of microorganisms was not prominently observed during the considered period of
3
continuous fermentation (excluding the very initial 24 hours) because of direct use of pre-
4
inoculated (10% inoculums v/v) and nutrient-supplemented sugarcane juice as inoculums.
5
The changes in the bacterial growth (OD660), pH, GA concentration and total sugar
6
concentration with time have been presented in the Fig. 5. pH of the fermentation broth
7
dropped from 6.5 to 3.7 as the GA concentration reached 45 g L-1 during the initial stage.
8
Three continuous fermentation experiments were carried out at steady state fluxes of 76 L m-2
9
h-1, 84 L m-2 h-1 and 90 L m-2 h-1 respectively. To maintain the fermentor volume constant,
10
fresh feed was introduced at the same rates into the fermentor with dilution rate of 0.15 h-1,
11
0.17 h-1 and 0.18 h-1. After few hours of cell recycle with introduction of fresh medium with
12
different dilution rates to keep fermentor volume constant, the overall mass concentration
13
started to decrease, but this condition had again started to increase as microbes got adapted
14
themselves to the prevalent conditions as shown in the Fig. 5 a, b and c. The dry cell mass
15
measured during steady state at three different conditions were 2.95±0.5 g L-1, 3.55±0.6 g L-1
16
and 4.1±0.8 g L-1 at cross flow velocities of 0.55 m s-1, 0.9 m s-1 and 1.25 m s-1 respectively.
17
In attaining steady state conditions cell bleeding played significant role. Variation in pH was
18
insignificant during these three independent runs and only small variation 3.6 to 3.8 was
19
observed. With the onset of MF cell recycle at a relatively high dilution rate of D = 0.18 h-1,
20
total sugar concentration initially increased, but again decreased to finally settle at 50.4 g L-1
21
under steady state conditions. However, different sugar concentration profiles are observed in
22
Fig. 5 a and b for dilution rates of D = 0.15 h-1 and D = 0.16 h-1. Ultimately average total
23
sugar concentrations reached 52.3 g L-1 and 54.1 g L-1 respectively at steady state conditions.
24
Fig. 5 c shows a similar trend. Under steady state conditions, a higher GA concentration (55.6
25
g L-1) with maximum product yield of 95% was achieved at a cross-flow rate 0.55 m s-1. At
26
this cross-flow velocity, continuous operation could be easily sustained in the face of
27
changing operating conditions introduced with membrane cell recycle than higher cross-flow
28
velocity regime. In other words, to achieve steady state conditions quickly higher cross-flow
29
velocity should be avoided. Table 2 shows the product concentration, yield and productivity
30
achieved during three different individual runs. During continuous operation, the highest
31
productivity achieved was 8.39 g L-1 h-1 at u = 1.25 m s-1 and dilution rate D = 0.18 h-1. The
32
final concentration of GA in those three runs after nanofiltration were 55.6 g L-1, 50.4 g L-1
33
and 47.8 g L-1 at cross-flow velocities 0.55 m s-1, 0.9 m s-1 and 1.25 m s-1 respectively in
34
fully membrane integrated system. The overall productivities after nanofiltration were 11 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1
computed as 8.3 g L-1 h-1, 8.1 g L-1 h-1and 8.6 g L-1 h-1 at the dilution rate 0.15 h-1, 0.16 h-1
2
and 0.18 h-1 respectively with >97% product purity. This is slightly less than the fermentative
3
productivity.
4
4. DISCUSSION
5
Glutamic acid with high purity could be produced continuously without any need for pH
6
adjustment in a fully membrane-integrated fermentation process using a cheap and renewable
7
carbon source. As membranes can be tailor-made, the new process will always have high
8
potential of ensuring high degree of purity of the acid product. Steady operation could be
9
attained for a constant fermentor capacity as well as constant transmembrane pressure. The
10
highest GA concentration achieved was 55.6 g L-1 with maximum product yield of 95% and
11
product purity 97% at a dilution rate of 0.15 h-1 although maximum productivity (8.6 g L-1 h-
12
1
13
for definite improvements over the existing ones in terms of carbon source, product purity,
14
yield, concentration and process intensification. In the literature, yield of GA like 66.3 g g-1
15
(GA concentration 25 g L-1)20, and 48 g g-1 (GA concentration 41.4 g L-1)29 have been
16
reported in batch mode fermentation, which are comparatively low.
17
In place of a traditional multi-step process, this new eco-friendly scheme is quite simple as
18
well as flexible. Literature shows that membrane integrated processes are rarely used for the
19
continuous production of GA. Benefits of such a membrane-based process are many and
20
likely to percolate down to the cane-growing farmers also. Single stage membrane separation
21
integrated with fermentor has always suffered from severe membrane fouling. Thus two stage
22
membrane separation like MF and selective NF (NF20) with appropriate mode (flat-sheet
23
cross-flow) not only produces GA with reasonably high productivity, yield, concentration and
24
purity but also offers an industrially acceptable flux of 65-78 L m-2 h-1. For running the
25
system in continuous mode over a long period of time, set flux method adopted for both MF
26
and NF modules is quite effective.
27
5. CONCLUSIONS
28
The study culminates in the development of a novel, fully membrane-integrated hybrid
29
process for continuous and fast production of glutamic acid overcoming the difficulties of
30
substrate-product inhibitions from a renewable carbon source. Selection of appropriate
31
membranes as well as modules in cell separation and product purification along with
32
provision of logical sequencing of operations ensures fermentation under high cell density
) was associated with a highest dilution rate of 0.18 h-1. Encouraging results were obtained
12 ACS Paragon Plus Environment
Page 12 of 23
Page 13 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
1
and purification to a high degree by highly selective NF membrane. Such a plant represents
2
high degree of process intensification which modern chemical process industries are
3
desperately seeking for their survival in highly competitive and environmentally conscious
4
world market.
5
Acknowledgment
6
Authors are thankful to the Department of Science and Technology, Government of India for
7
financial support under Start-Up Research Grant for Young Scientist (Science and
8
Engineering Research Board) (SB/FTB/ETA-59/2013).
9
REFERENCES
10
1.
Kumar, R.; Vikramachakravarthi, D.; Pal, P. Production and purification of glutamic
11
acid: A critical review towards process intensification. Chem. Eng. Process. Process
12
Intensif. 2014, 81, 59-71.
13
2.
Çalık, G.; Ünlütabak, F.; Özdamar, T. H. Product and by-product distributions in
14
glutamic acid fermentation by Brevibacterium flavum : effects of the oxygen transfer.
15
Bioche. Eng. J. 2001, 9, 91–101.
16
3.
Saudi Pharm. J. 2013, 21, 337–343.
17 18
4.
5.
6.
Nampoothiri, K. M.; Pandey, A. Immobilization of Brevibacterium cells for the production of l-glutamic acid. Bioresour. Technol. 1998, 63, 101–106.
23 24
Shih, I.; Van, Y. The production of poly- (γ-glutamic acid ) from microorganisms and its various applications. Bioresour. Technolo. 2001, 79, 207-225.
21 22
Hermann, T. Industrial production of amino acids by coryneform bacteria. J. Biotechnol. 2003, 104, 155-172.
19 20
Dutta, S.; Ray, S.; Nagarajan, K. Glutamic acid as anticancer agent: An overview.
7.
Zhang, J.; Tang, L.; Zhang, H.; Yang, Y.; Mao, Z. A novel and cleaner technological
25
process of extracting L-glutamic acid from fermentation broth by two-stage
26
crystallization. J. Clean. Prod. 2012, 20, 137–144.
27
8.
Pal, P.; Sikder, J.; Roy, S.; Giorno, L. Process intensification in lactic acid production:
28
A review of membrane based processes. Chem. Eng. Process. Process Intensif. 2009,
29
48, 1549–1559.
30
9.
Nayak, J.; Parimal Pal, P. Transforming Waste Cheese-Whey into Acetic Acid through
31
a Continuous Membrane-Integrated Hybrid Process. Ind. Eng. Chem. Res. 2013, 52,
32
2977−2984. 13 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1
10.
Sikder, J.; Roy, M.; Dey, P.; Pal, P. Techno-economic analysis of a membrane-
2
integrated bioreactor system for production of lactic acid from sugarcane juice.
3
Biochem. Eng. J. 2012, 63, 81–87.
4
11.
Dey, P.; Linnanen, L.; Pal, P. Separation of lactic acid from fermentation broth by
5
cross flow nanofiltration : Membrane characterization and transport modelling.
6
Desalination, 2012, 288, 47–57.
7
12.
Timmer, J. M. K.; Speelmans, M. P. J.; van der Horst, H. C. Separation of amino acids
8
by nanofiltration and ultrafiltration membranes. Sep. Purif. Technol. 1998, 14, 133–
9
144.
10
13.
Bouchoux, A.; Roux-de Balmann, H.; Lutin, F. Investigation of nanofiltration as a
11
purification step for lactic acid production processes based on conventional and bipolar
12
electrodialysis operations. Sep. Purif. Technol. 2006, 52, 266–273.
13
14.
broth by nanofiltration. J. Membr. Sci. 2003, 222, 191–201.
14 15
Li, S.; Li, C.; Liu, Y.; Wang, X.; Cao, Z. Separation of l-glutamine from fermentation
15.
Xu, G. et al. Development of a continuous cell-recycle fermentation system for
16
production of lactic acid by Lactobacillus paracasei. Process Biochem. 2006, 41,
17
2458–2463.
18
16.
Giorno, L.; Chojnacka, K.; Donato, L.; Drioli, E. Study of a Cell-Recycle Membrane
19
Fermentor for the Production of Lactic Acid by Lactobacillus bulgaricus. Ind. Eng.
20
Chem. Res. 2002, 41, 433–440.
21
17.
Torang, A.; Jonsson, A.; Zacchi, G. Separation of cells and proteins from fermentation
22
broth in a shear-enhanced cross-flow ultrafiltration module as first step in refinement of
23
lactic acid. Appl. Biochem. Biotechnol. 1999, 76, 144–157.
24
18. Wang, X.-L.; Shang, W.-J.; Wang, D.-X.; Wu, L.; Tu, C.-H. Characterization and
25
applications of nanofiltration membranes: State of the art. Desalination 2009, 236,
26
316–326.
27
19. Ward, R. A.; Klein, E.; Feldhoff, P. W.; Turnham, T. J. Ultrafiltration: separation enhancement by counter-current cascading. J. Membr. Sci. 1987, 33, 97–111.
28 29
20.
Nampoothiri, K. M.; Ashok Pandey, A. Fermentation and recovery of L-glutamic acid
30
from cassava starch hydrolysate by ion-exchange resin column. Revista de
31
Microbiologia 1999, 30, 258-264.
32 33
21.
Shen, J.; Duan, J.; Yu, L.; Xing, X.; Xu, P. Desalination of glutamine fermentation broth by electrodialysis. Process Biochem. 2006, 41, 716–720.
14 ACS Paragon Plus Environment
Page 14 of 23
Page 15 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1
Industrial & Engineering Chemistry Research
22. Russo, L. J.; Kim, H. Membrane-based process for the recovery of lactic acid by fermentation of carbohydrate substrates containing sugars, US Patent no. 5,503,750.
2 3
23.
Leuchtenberger, W.; Huthmacher, K.; Drauz, K. Biotechnological production of amino
4
acids and derivatives : current status and prospects. Appl. Microbiol. Biotechnol. 2005,
5
69, 1-8.
6
24. Kov´acs, Z.; Samhaber, W. Contribution of pH dependent osmotic pressure to amino
7
acid transport through nanofiltration membranes. Sep. Purif. Technol. 2008, 61, 243–
8
248.
9
25.
Wang, K.; Li, W.; Fan, Y.; Xing, W. Integrated Membrane Process for the Purifi
10
cation of Lactic Acid from a Fermentation Broth Neutralized with Sodium Hydroxide.
11
Ind. Eng. Chem. Res. 2013, 52, 2412-2417.
12
26.
Çalık, G.; Ünlütabak, F.; Özdamar, T. H. (2001). Product and by product distributions
13
in glutamic acid fermentation by Brevibacterium flavum: effects of the oxygen transfer.
14
Biochem. Eng. J. 2001, 9, 91–101.
15
27.
Carrère, H.; Blaszkow, F. Comparison of operating modes for clarifying lactic acid
16
fermentation broths by batch cross-flow microfiltration. Process Biochem. 2001, 36,
17
751–756.
18
28.
Sikder, J. et al. Synthesis and characterization of cellulose acetate-polysulfone blend
19
microfiltration membrane for separation of microbial cells from lactic acid
20
fermentation broth. Desalination 2009, 249, 802–808.
21
29.
Choi, S-U.; Nihira, T.; Yoshida, T. Enhanced Glutamic Acid Production of
22
Brevibacterium sp. with Temperature Shift-Up Cultivation. J. Bioscie. Bioeng. 2004,
23
98, 211-213.
24
30.
Artug, G.; Hapke, J. Characterization of nanofiltration membranes by their
25
morphology, charge and filtration performance parameters. Desalination 2006, 200,
26
178-180.
27
31.
fractionation with NF membranes. J. Membr. Sci. 2006, 283, 57-64.
28 29
Tanninen, J.; Manttari, M.; Nystrom, M. Effect of salt mixture concentration on
32.
Sikder, J.; Chakraborty, S.; Pal, P.; Drioli, E.; Bhattacharjee, C. Purification of lactic
30
acid from microfiltrate fermentation broth by cross-flow nanofiltration. Biochem. Eng.
31
J. 2012, 69, 130–137.
32
15 ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Page 16 of 23
Table 1 Characteristics of membranes used in this work according to membrane manufacturer’s data, literature data and obtained in the present work. ____________________________________________________________________________________________________________________ Characteristics Membranes _________________________________________________________________________ NF1 NF2 NF3 NF20 ____________________________________________________________________________________________________________________ Manufacturer Sepro Co. (USA) Sepro Co. (USA) Sepro Co. (USA) Sepro Co. (USA) Geometry Flat-sheet Flat-sheet Flat-sheet Flat-sheet Physical properties: Membrane width (m) 1.02 1.02 1.02 1.02 Thickness including fabric backing (µm) 165 165 165 165 Roll length (m) 250 250 250 250 Material polyamide polyamide polyamide polyamide Test pressure (bar) 10.3 10.3 10.3 10.3 2 2 2 2 Solute concentration (g L-1) Solute rejection (%) MgSO4 99 97 98 98 NaCl 90 50 60 35 Average molecular weight cut-off (Da) 150-300 150-300 150-300 150-300 pH range 3-10 3-10 3-10 3-10 o Cleaning pH range (at 50 C) 2-11 2-11 2-11 2-11 Maximum operating temperature (oC) 50 50 50 50 Maximum operating pressure (bar) 83 83 83 83 2.7b 6.6c Isoelectric point (pH)
