Process intensification in lactic acid production

8/12/2019 Process intensification in lactic acid production

1/11

Chemical Engineering and Processing 48 (2009) 15491559

Contents lists available at ScienceDirect

Chemical Engineering and Processing:Process Intensication

j o u rn a l h om epa ge : www.e l sev i e r. com/ loca t e / cep

Process intensication in lactic acid production: A review of membrane basedprocesses

Parimal Pal a,, Jaya Sikder a , Swapan Roy b , Lidietta Giorno ca Environment and Membrane Technology Laboratory, Department of Chemical Engineering, National Institute of Technology,Durgapur 713209, Indiab T.D.B. College, University of Burdwan, Burdwan 713347, Indiac Research Institute on Membranes and Modeling of Chemical Reactors, c/o University of Calabria, Via P. Bucci 17/C, 87030 Rende (CS), Italy

a r t i c l e i n f o

Article history:Received 13 October 2008Received in revised form 7 July 2009Accepted 10 September 2009Available online 16 September 2009

Keywords:Process intensicationLactic acidClean productionMembrane process

a b s t r a c t

Lactic acid the most widely occurring hydroxy-carboxylic acid has traditionally been used as foodpreservative and acidulent. So long, it has been produced through either chemical synthesis route orfermentation route the latter being the dominating one. Despite its tremendous potential for large scaleproduction and use in a wide variety of applications, cost-effective production of high purity lactic acidhas remained a challenge for decades, mainly due to high downstream processing cost. In the recentyears, possibility of integration of highly selective membranes with the conventional fermentors hasopened a golden opportunity for full commercial exploitation of the tremendous application potentialof this wonder chemical. This paper discusses recent developments of such membrane-based processesrepresenting process intensication in production of monomer grade lactic acid while suggesting a verypromising production scheme.

2009 Elsevier B.V. All rights reserved.

1. Introduction

Thechemical and theallied process industries allover theworldare now confronted with the big challenges of developing inno-vative products and processes for survival in an era of emaciatedprot margins amidst highly globalized trade competition and fastgrowing environmental constraints. Thus process intensicationthroughrevolutionarydevelopmentof newproductsand processesthatensure reduced materialand energy consumptionandreducedenvironmental impacts while offering greater exibility in scaleof operation are the need of the hour. Production of monomergrade lactic acid (2-hydroxypropanoic acid) a traditionally usedfood preservative and acidulent has over the last few decades,attracted attention of the world researchers.

By virtue of unique presence of both hydroxyl and carboxylicacid groups, lacticacid canparticipatein a wide variety of chemicalreactions like esterication, condensation, polymerization, reduc-tion and substitution and this has contributed to its tremendouspotential as a platform chemical for a whole range of products thathave very large-volume uses for industrial production and con-sumer products. Biodegradable thermoplastics (polylactic acid),greensolvents (ethyl, propyl, butyl lactates)and oxygenated chem-

Corresponding author.E-mail address: [email protected] (P. Pal).

icals (propylene glycol) are a few examples of lactic acid-derivedproducts, market demands of which are growing exponentiallyover the years [1]. Exploitation of its full potential, however,depends largely on how cost-effectively it can be produced withhigh levels of purity. The major technology barrier in cost-effectiveproduction of high purity lactic acid is its down-stream separationand purication. And this is where; membrane-based processesare stepping in. Being modular in design, membrane-based pro-cesses offer great exibility in scale of production depending onmarket demand. By virtue of high selectivity, membranes canensure high levels of separation and purication. As membranesof chosen selectivity and permeability can easily be integratedwith conventional fermentors, membrane-based processes permitsimultaneous production and purication in the same unit. Thiseliminates the need for separate purication units and results incompactdesignwith reducedcapitalinvestment.Membrane-basedseparation and purication (barring pervaporation) involves nophase change ensuring reduced energy consumption. Thus suchprocesses can meet all the goals of process intensication. In thispaper, we rst briey discuss the traditional processes to high-light the major problems associated with these processes and thenreview the developments in membrane-based processes whichhaveattemptedto overcomethe problems of traditional processes.The objective is to identify an environmentally benign, simple, eco-nomically viable and continuous manufacturing scheme capable of producing monomer grade lactic acid with high productivity.

0255-2701/$ see front matter 2009 Elsevier B.V. All rights reserved.

doi: 10.1016/j.cep.2009.09.003


2/11

1550 P. Pal et al. / Chemical Engineering and Processing 48 (2009) 15491559

Fig. 1. Lactic acid production scheme in chemical synthesis route using acetalde-hyde feed.

2. Traditional production process

Traditionally lactic acid has been produced through eitherchemical synthesis route or fermentation route. Chemical synthe-sis route as shown in Fig. 1 uses by-products like lactonitrile fromother industries. Hydrogencyanide (HCN)is added to acetaldehyde(CH 3 CHO) in liquid phase in presence of a base catalyst underhigh pressure when lactonitrile is produced. Crude lactonitrile isthen puried by distillation and subsequently hydrolyzed to lac-tic acid by hydrochloric acid or sulphuric acid [2]. The processis often dependent on other by-product industries and, consid-ered expensive where petroleum based raw material is the majorcost-contributor. Moreover, chemical synthesis route produces amixture of L-lactic acid and D-lactic acid whereas in most of the cases, L-lactic acid is the desirable product. The problems of high cost of raw materials, impurity of the product and depen-dence on other industries for raw materials could be overcomein fermentation based process. Thus majority of the big lacticacid manufacturing industries like Musahino chemical in Japan,

Table 1Widely usedsubstrates in fermentation-basedlactic acid production.

Substrates Reference

Whole-wheat powder [3]Starch [4]Cucumber juice [5]Cheese whey [6]Sugarcane bagasse cellulose [7]Molasses [8]

Sugar cane juice [9]

CCA Biochemical BV of the Netherlands, Natureworks LLC etc.have switched over to fermentation based technology as pre-sented in Fig. 2. The scheme consists of a number of downstreamtreatment schemes like precipitation, conventional ltration, acid-ication, carbon adsorption, evaporation etc. Though possiblecarbon sources may be numerous, mainly the cheap and renew-able carbohydrates as shown in Table 1 are used as the sourceof carbon and produces optically pure form of lactic acid [3] . Thesubstrates are chosen on the basis of their cost, levels of contam-inants, fermentation rate, lactic acid yields, by-product formation,and ability to be fermented with little or no pretreatment andyear round availability. Substrate cost is one of the major cur-rent problems. Downstream purication needs are closely relatedto the nature of these substrates, their degree of conversion andthe reagents or microbes involved in the conversion process alongwith other factors. The most widely used substrates for productionof lactic acid are glucose, sucrose and lactose which are solublecarbohydrates [4] . Cucumber juice (CJ) is also a good substrate forproduction of lactic acid as it contains malic acid with glucose andfructose where malic acid is very rapidly converted to lactic acid.Molar yield of lactic acid from CJ were found to be 11.2 [5] . Cheesewhey, a by-product from cheese manufacturing industry is alsoused as substrate for production of lactic acid as it contains huge

Fig. 2. Typical conventional fermentation-based lactic acid production scheme.


3/11

P. Pal et al. / Chemical Engineering and Processing 48 (2009) 15491559 1551

amount of lactose. Composition of this whey permeate is 9899%total solids, 7981%lactose,810%ash and2.54%protein [6] .Someresearchers used sugarcane bagasse cellulose for batch lactic acidproduction and the yield was 83% but the productivity obtainedwasverylow (0.93g/L/h) and concentration of lacticacidwas 67g/L [7] . Molasses whichis a byproductin production of cane sugar fromsugar cane juice contains 4060% sucrose have also been used asa substrate. It has been reported [8] that lactic acid with yield of 0.880.96 and productivity of 4.3g/L could be produced in a 40 h-batch fermentation of molasses with initial sugar concentration of 190g/L. As the molasses contains enough nitrogen source, need foraddition of yeast extract in such case, is eliminated. Timbuntam etal. [9] used sugar cane juice containing 125g/L sucrose, 8 g/L glu-cose and 6 g/L fructose as substrate in a batch fermentation. Thehighest yield of 96%was obtained and productivity was 2.8g/L/h at3%, w/v sugar cane juice after 10 h of fermentation. Use of naturalsubstrates like starch and cellulose is not economically favourable[10] due to requirement of pretreatment to release fermentablesugars.

The most widely used homofermentative bacterial strains usedin such fermentation are Lactobacillus bulgaricus, LactobacillusLeichmanii, Lactobacillus delbrueckii, Lactobacillus amylophilusand Lactobacillus plantarum. Fungal strains like Aspergillus nigerand Rhizopus species have also been used in several studies [11]and have been claimed to be more selective in production of L-lactic acid stereoisomer and less nutrient-demanding than theirbacterial counter parts. In general, fermentative process can selec-tively produce the desirable L-lactic acid stereoisomer instead of a mixture of L and D isomers using specic microbial strains [11]but maintaining high-growth environment for these microbes isdifcult. With production and accumulation of acid, pH falls andaffects productivity of the microbes. In the conventional fermenta-tion process as shownin Fig.2 , pHof thebatch reactor is maintainedat around56 by addition of calcium hydroxide or calcium carbon-ate and the end-fermentation concentration of lactic acid is only10 wt% normally. Sometimes ammonium hydroxide is also used.The problem of addition of lime in controlling pH is that it leads

to production of calcium lactate instead of lactic acid at high pHas pK a value of lactic acid is 3.86 at 30 C. Calcium lactate is thenseparatedfrom the microbial cells by ltrationand further puriedby activated carbon adsorption. The subsequent steps are evap-oration and acidication by sulphuric acid to produce lactic acidand insoluble calcium sulphate (gypsum) as by-product. Gypsumby-product which remains associatedwith organicmassand is pro-duced at the rate of 1 metric tonne per metric tonne of lactic acidis a big environmental problem associated with conventional fer-mentation process. Need for so many steps along with evaporationinvolving phase change to get pure lactic acid naturally involveshigh capital investment as well as high operating cost. The prob-lem of low pH and hence lowproductivity can be largely overcomeif produced lactic acid is continuously removed from the fermen-

torand this is possible if fermentationis carried outin a continuousmode. Continuous removal of lactic acid from fermentation brothcan be done by adsorption [1214] , extraction [15,16] and mem-brane separation. Adsorption and extraction based processes alsoneed quite a few steps as regeneration of adsorbent and recyclingof solvent is necessary. These processes themselves cannot ensureseparation of microbial cells for their recycle without additionalprovision of cell separation and recycle. High cell concentrationin the fermenter is essential for high productivity. Moreover, theextent of process intensication that can be achieved in a mem-brane basedprocessdue to theassociatedadvantagesas mentionedearlier cannot be expected from adsorption or extraction-basedprocesses.Thus therecent years havewitnessed extensiveresearchactivities on solving the problems of traditional manufacturing

processes through membrane based separations. The subsequent

Fig. 3. Schematic ow diagram of continuous membrane based fermentation sys-tem coupled with micro, ultra, nano or reverse osmosis membrane module in asingle stage.

sections review the developments in such membrane basedprocesses.

3. Membrane based processes

For continuous mode operation of a fermentative process usingrenewable carbohydrate sources for lactic acid production, thecomponents that need to be continuously separated from thefermentation broth are microbial cells, proteins, nutrients (yeastextract, salts of ammonium, potassium, phosphorus, etc.), uncon-verted carbonsources,water andlacticacid as shown in Table2 .Themembranes that can play effective role in separation of these com-ponents are microltration, ultraltration, nanoltration, reverseosmosis and electrodialysis membranes. Microbial cells and pro-teins can quickly foul all membranes though extent of foulingmay be far less in microltration membranes compared to thatin nanoltration or reverse osmosis membranes. However, mem-branes used in some particular modules may be operated forlong without much fouling. Microltration membranes having the

largest pore size (0.10.2 m) among the categories can separatemicrobial cells for their subsequent recycling to the bioreactor toensure high cell concentration and thus high productivity. Ultral-tration membranes with average pore size much less than that of the microltration membranes can retain cells and proteins. Sepa-ration by microltration and ultraltrationmembranes is basedonsizeexclusionand foreffectivecell harvesting, at least100300 kDamolecular weight cut off (MWCO) value should be ensured [17] forthis. Nanoltration membranes being in between reverse osmo-sis and ultraltration membranes with average pore size of 1 nmcan separate cells, proteins, nutrients, salts, and unconverted car-bon sources from lactic acid. Reverse osmosis normally knownas nonporous membrane where separation is based on solutiondiffusion mechanism can separate the same components from fer-

mentation broth as nanoltration membranes do but at muchhigher operating pressure than what is needed in nanoltration.The schematic ow diagram in Fig. 3. shows how such micro,ultra, nano or reverse osmosis membrane modules can be cou-pled with a fermentor permitting continuous removal of acid fromthe broth and separation of cells, nutrients or unconverted car-bon sources for their subsequent recycle. The scheme in the Fig. 3shows a single stage integration of a membrane module as hasbeen investigated in several studies over the last two decades.Separation and recycle of the components depends on the typeof membrane used. For example, if a microltration membranemodule is coupled, only cells are likely to be retained while per-mitting acids, unconverted carbon sources, proteins nutrients andwater to pass to the permeate side. This however, ensures con-

tinuous removal of acid from the fermentor helping to arrest


4/11


Table 2Typical components of lactic acid fermentation broth that can be separated by two-stage membrane lters integrated with fermentor.

Components Effects on membrane ux/fouling Separation mechanism

Microbes seriously affects ux of nal lter through fouling microltration with 15 kgf/cm 2 ) for ltration.An electrodialysis membrane where separation is based on elec-tromigration of ions through a stack of cation and anion exchangemembranes basically involves two stepsconventional electrodial-ysis (CEP) and the bipolar electrodialysis (BED). The rst step (CEP)separates and concentrates lactate salts and the second step (BED)convertslactatesalts intolacticacid.The schematic diagramof Fig.4shows lactic acid production in a fermentative process coupledto electrodialysis membrane modules. The next section reviewsthe developments in the rst stage of separation and purica-tion by microltration and ultraltration membranes and thenmoves over to nanoltration, reverse osmosis and electrodialy-sis.

3.1. Microltration and ultraltration of fermentation broth

Limitations to fermentation based batch production processbasically stems from substrate-product inhibition [18,19] and thetime loss for shut down and start up after every batch of produc-tion. As the concentration of lactic acid in the fermentation cellincreases above 10 g/L, microbial activities start getting reduceddue to increased difculty of survival of the microbes in a low pHmedium. Variation of pH-effects on bacterial growth results fromthe presence of dissociated and undissociated forms of lactic acidin fermentation broth. The undissociated form is more inhibitorythan dissociated form. At low pH ( 6, almostcomplete dissociationof lacticacidtakesplace [20] . Thus the optimum pH value is around 6.

While studying membrane based separation coupled to a fer-mentor, Milcent et al. [21] observed pH as a key parameter. Apartfrom removal of acid from the medium they used NH 4 OH to main-tain pH at 6.2. Lowering of pH resulted in decrease of ux and viceversa. But beyond a pH level of 78, no further improvement ii uxwas observed with increase of pH. They used a ceramic tubularmicroltration membrane (0.1 m pore diameter) module for cellseparation and acid removal from the broth of a fermentor. At anaverage transmembrane pressure of 1 bar and cross ow velocityof 4 m/s, the system permitted a steady-state ux of 85L/(m 2 h)formolasses as carbonsource and Lactobacillus delbreuckii ssp. lacti s asmicrobial strain. It was identied that critical uxes were functionof cross ow velocity. It was observed that increase of cross owvelocity did not lead to an increase of the ux due to irreversiblefouling resulting from adsorption of molasses compounds on to themembrane surface. They didnot reuse the separated cells resultingin low productivity.

Taniguchi et al. [22] observed a rapid decrease in the ratio of specic growth rate to maximum specic growth ( / m ) at a lac-tate concentrationof 22g/L and no microbial growthwas observedat 35 g/L of lactate concentration. At the outset, two most impor-tant tasks to ensure high productivity in a fermentor are removalof acid from the fermentor medium to maintain optimum pH(around 6) and separation and recycle of the microbial cells at thelate-logarithmic growth phase of the microbial cells. Coupling of

membrane separation with a fermentor in an external unit permitscontinuous separation and removal of lactic acid from the fermen-tation broth preventing lowering of medium pH to an inhibitionlevel and this simultaneously permits cell separation and recyclealso. Taniguchi et al. [22] achieved 29-fold increase (compared tothe system without ltration) by removing lactic acid from thebroth by a cross ow ltration and by recycling the cells retainedon the microlter to the fermentor. This ensured high cell concen-tration (81.5g dry cell/L) in the fermentor. They used a ceramicmicrolter in tubularmodule as shown in Fig.5 . pH was maintainedat 6.5 by addition of NH 4 OH. Use of ceramic membranes permittedeasy disinfection also. But ceramic membranes suffer from quickfouling and can retain only cells where unconverted substrate getslost.

Fig. 5. A tubular module with two concentric hollow membrane tubes cast on the

inner and outer walls of two concentric porous support tubes.


5/11


In a design aimed at total substrate conversion, Moueddeb etal. [23] studied conversion of lactose into lactic acid using Lacto-bacillus rhamnosus in a microltration membrane bioreactor thatconsistedof twocoaxial alumina tubes having alpha alumina mem-brane (pore size: 2.0 10 7 m) on the inner wall of the inner tubeand on the outer wall of the outer tube. The micro-organisms werexed in the macroporous support andconnedin the annular spaceof two coaxial porous tubes of a tubular membrane. The substratesolution was fed into the reactor inner compartment whereas theliquid percolatedin the radial directionacross the two membranes.Lactic acid was produced in the porous space between the twomicroltration layers. Though the new design focused on elimi-nation of bacterial inhibition giving total substrate conversion, uxdecreased rapidly due to membrane plugging by microbes whichcould, however, be reduced to some extent by sterilization of theceramic membranes.

To eliminate the inhibition problem through direct removal of acid from the fermentation broth, Giorno et al. [24] integrated onecrossow membranemodule tted withmicroltration or ultral-tration capillary membranes with a 2.5L stirred cell. They studiedconversion of glucose to lactic acid using Lactobacillus bulgaricus .They used capillary polysulfone ultraltration (UF) membranesof molecular weight cut off diameter (MWCO) of 100 kDa andpolyamideUF membrane of MWCO value 50kDa. In thesame mod-ule, they tested 0.1 m pore size polyamide microltration (MF)membrane also for separation and recycle of cells from the broth.With an initial glucose concentration of 17 g/L, the achieved lacticacid concentration was 10.8 g/L the average yield being 62% andproductivity, 0.43 g/L. Polyamide membranes showed lower uxreduction than polysulfone types but direct ultraltration of brothwithout prior microltration resulted in quick reversible foulingas the system was operated at cross ow velocity to protect themicrobes from shear.

Compared to polymeric membranes, the ceramic membraneshave the advantage of easy disinfection. Ultraltration membranescanretain both cells andproteins. Xavier et al. [25] coupled ceramicultraltration membrane to a fermentor to separate both cells and

proteins for their recycling and simultaneous removal of acid fromthe medium. They achieved 36g/L h productivity and a lactic acidconcentration of 90g/L in a tubular module. In continuous mode of operation, theux was 15L/m 2 h.However,operationof thesystemfor 90 h at a dilution rate of 0.40/h ux decreased rapidly due toclogging of the ultraltration membrane.

As productivity and acid concentration are two mutually exclu-sivegoals in fermentativeprocess,achieving bothis a bigchallenge.To take up this challenge Kwon et al. [26] connected twomembranecell recycle bioreactors (MCRB) in plate and frame membranemod-ule in series with cell recycle in each stage. Fig. 6 shows such amodule. The bioreactor was a water-jacketed stirred glass reactor.Using Lactobacillus rhamnosus strain on glucose feed stock, theyachieved a high productivity of 57 g/L h and lactic acid concentra-

tion of 92 g/L. Flat sheet plate and frame module ( Fig. 9) is like theconventional plate and frame lter press where a series of annularmembrane discs can be placed one on either side of polysulfonesupport plates that provide channels for withdrawal of permeate.Spacer plates separate the sandwiches of membranes and sup-port plated. The spacer plates having central and peripheral holesdirect feed liquor overthe membranesurface.Permeate is collectedfrom each membrane pair. Thoughplate andframe module has theadvantage of easy membrane cleaning option it suffers from theproblem of low membrane surface area per unit volume.

As a hollow ber module shown in Fig. 7 offers high surfacearea per unit volume and avoids bypassing of feed unlike in spiralwound module Levente et al. [27] integrated anamiconhollow bermicroltration membrane (0.1 m pore size) module with a biore-

actor. This was a cell recycle system with cell bleeding introduced

Fig. 6. Plate and frame membrane module.

Fig. 7. Hollow ber membrane module.

at the47th hour when biomass concentrationapproached30 g/Ltoensuresteady state in terms of biomass concentration so that fall in

ux could be arrested. The researchers claimed that for over 100h,steady state operation could be maintained when ux stabilized at5L/m 2 h.Theyachieved 99%substrate utilization using Lactobacillusrhamnosus strain and better purity (economic advantage) by con-tinuous cell bleeding. However, the study used glucose as feed andin this case also, pH hadto be maintained by continuous addition of NaOH with the help of an automatic pH-stat instrument (Metrohmsystem,BrinkmannInstruments, Canada).But their system sufferedfrom high membrane fouling problem also.

To overcome the problem of membrane fouling in microltra-tion stage,Toranget al. [28] suggesteda shear-enhanced cross-owultraltration module for separation of cells and proteins fromfermentationbroth. 100% protein retentionwas doneby hydropho-bic (polyethersulfone) membrane (MWCO 25kDa) but due to the

blocking of the pores by protein adsorbed on to the hydrophobicmembrane surface the ux was higher for hydrophilic (regener-ated cellulose acetate) membrane with MWCO of 20 kDa and uxof 1285L/m 2 h. Hydrophilic membranes used had low proteinbinding tendency. The chemical stability and as well as proteinseparation was better for hydrophobic membrane so microltra-tion or centrifugation were suggested before the ultraltration tomake the ultraltration step more efcient by avoiding the foulingof the membrane by high molecular weight protein. Cell harvest-ing by microltration or ultraltration for its subsequent recyclingleads to high cell concentration in the fermentor but often exces-sive build up increases viscosity and causes lowering of ux. Toovercome this problem Crespo et al. [29] suggested cell bleeding.

Nishiwaki and Dunn [30] developed a continuous process for

production of lactic acid by two stage fermentor combined in each


6/11


Fig. 8. Cross ow at sheet membrane module.

recycle loop with cell separators and analyzed numerically. Thebleed culture from the rst stage was introduced to the secondfermentor. By relatingthe maximumoverallproductivitywith dilu-tion rate it was observed that efciency was greater for two-stagesystem. A hybrid cell-recycle ultraltration membrane fermen-tor system was developed for production of lactic acid. Hollowber ultraltration module ( Fig. 7) after 61hr operation from thebeginning a second membrane module was introduced and it wasobserved that the permeate ux of the second module was higherthan the rst module. These results were explained by the factthat fouling properties of the fermentation broth during growthphase of bacteria Streptococcus faecalis were different from theproperties during the non-growing phase. The difference could bein the protein concentration which was high during the lag phase,decreasesduringthe exponential growthandthe stationary phases,andremainalmostconstant during thedeathphase. Thecontinuousfermentation was nished after 83.5 h of operation and concentra-tion, productivity and yield of lactic acid was 16 g/L, 0.19g/L h and0.94g/g respectively using glucose as feed stock. The reversiblefouling was observed due to the need to work at axial velocitieslower than 1 m/s.

Xuet al. [31] claimed to have controlled the problem of concen-trationpolarizationand foulingeffect by controllingtangentialowin the PVDF microltration membrane (Millipore) attached with afermentor for downstream separation of cells Lactobacillus paraca-sei . They claimed to have achieved a productivity of 31.5g/L h andyield of 0.85 g lactic acid/g glucose. The best results were obtainedthat the maximum productivity was 31.5 g/L h; yield was 0.85 forprolonged operation time of over 155h. Operation over such longperiod wasdone withon-line sterilizationand cleaningof themem-brane using NaOCl and distilled water.

A tubular module ( Fig. 5) where the membrane is cast on theinside of a porous support tube housed in a perforated stainlesssteel shroud is considered effective in purication of a solutionhaving high concentration of solids. A tubular module can also be

operated at high pressure. But normally microltration does notrequire highpressure. Another major problem withtubularmoduleis quick fouling. Cleaning is often difcult though easy disinfec-tion of ceramic membrane solves the plugging problems to someextent. Overall mass transfer area in a tubular module is low andit involves high volumetric hold up. To overcome the difculty of fouling and high volumetric hold up, Sikder et al. [32] integrateda crossow, at sheet membrane module as shown in Fig. 8, witha fermentor where a laboratory-synthesized polysulfone-celluloseacetate blend microltration membrane very effectively separatedcells from fermentation broth without the problem of fouling evenafter long hours of operation. The tangential ow pattern in suchat sheet membrane module largely removes concentration polar-ization problems of dead end lters as shown in Fig. 9. Integration

of ultraltration membrane particularly in tubular module with

Fig. 9. Flat sheet stirred cell operating in dead-end mode.

a fermentor directly in a single step results in rapid fouling of the membranedue to concentrationpolarization. In microltrationmembranes, this fouling problem is not so acute but in both cases,integration of either microltration or ultraltration alone resultsin huge loss of valuable carbon sources andnutrients though main-tenance of desired pH level and high cell concentration throughrecycling greatly improvesproductivityovercomingtheproblemof pH-inhibition. However, nanoltration and reverse osmosis mem-branes can retain these unconverted carbon sources and nutrientsfor their subsequent recycling along with microbial cells whilesimultaneously permitting separation and purication of lacticacid. The next section reviews the developments in such nanol-tration (NF) and reverse osmosis (RO) based separations coupled toa fermentative process.

3.2. Nanoltration and reverse osmosis

Timmer et al. [33,34] integrated NF and RO membranes with a

fermentor. They used a spiral wound module as shown in Fig. 10.RO or NF membranes were tted with this module for separationof lactic acid from fermentation broth. The rejection parametersof NF membranes CA960 and HC95 at a pH of 4.93 for undissoci-ated and dissociated lactic acid were 0.675, 0.898 and 0.497, 0.953respectively compared to 0.73, 0.989 and 0.974, 0.998 respectivelyin RO membrane varieties CA995 and HR95 at the same pH. Butthe mass transfer parameter of NF membranes (CA960 and HC95)at 4.93 pH for undissociated and dissociated lactic acid was higherthan those of the RO membranes. It was observed that lactic acidseparation characteristicsof NF were better than those of RO mem-branes.Asexpected,rapidfoulingwas encounteredwhile operatingthe system on either RO or NF membranes.

Liew et al. [35] integrated a polyamide composite membrane

with a stirred cell ( Fig. 6) operating in a dead-end mode. For a250mL reactor, they achieved a ux of 26 L/m 2 h lactic acid alongwitha ux of12.5 L/m 2 h ammoniumlactateand 7.5L/m 2 h sodiumlactate. The optimum operating pressure was 7MPa and optimumagitation was found to be 900rpm. The permeate ux is higherwith the at-sheet module than with the spiralwoundimplies thatspiral wound module is more susceptible to concentration polar-ization than at-sheet module or the whole membrane area of thespiral wound module is not effectively used.

To reap the benet of higher ux of a at sheet cross owmodule, Bouchoux et al. [36] integrated such a module ( Fig. 8)of nanoltration membrane (composite polysulfone type Desal 5DK of GE Osmonics) with a fermentor for production and puri-cation of lactic acid from glucose. It was observed that glucose

retentiondecreasedin presence of sodium lactate.For thisdecrease


7/11


Fig. 10. Typical Spiral wound membrane module.

of retention of glucose, separation was difcult and unachievable.The decrease of a neutral solute (glucose) retention in the pres-ence of charged one(sodium lactate) is probably a general problemin nanoltration. This can be explained by the charge density of nanoltration. Higher membrane charge density in presence of sodium lactate may decrease the glucose retention.

Jeantet et al. [37] integrated a spiral wound nanoltration mem-brane modulewitha fermentoras such a modulehas the advantageof controlling hydrodynamics and concentration polarization tosome extent by changing spacer thickness though membrane sur-face area per volume is low for this module. Spiral wound module(Fig. 10 ) where several at membranes separated by turbulence-promoting mesh separators form a Swiss role, has long been usedforits compact designallowinga numberof modules to be installedin a single pressure tube and permitting a high operating pres-sure. In this spiral wound module, concentration effects of ionswere found to be more important than concentration polarization.Their suggested scheme was for semi continuous production andseparation of lactic acid from the fermentation broth. pH was stillregulated by continuous addition of NaOH. Compared to the pro-ductivity (typically 1 g/Lh) of classical batch production system,they achieved quite high volumetric productivity of 7.1g/L h withspecicproductivityof 3.5h 1 . Butproteinfoulingof thenanoltra-tion membrane was high and they suggested microltration prior

to nanoltration as the used spiral wound modules suffered quickfouling.

In their effort to investigate the effectiveness of mechani-cally strong membranes, Duke et al. [38] used readily available

-alumina nanoltration membrane and the more advancedmolecular sieve silica membranes, to enrich lactic acid for prod-uct use by selectively depleting water through the membrane. Thealumina membranes showed initial ux level at 6 kg/m 2 h but itreduced to 1 kg/m 2 h after 250 min due to pore blocking of lacticacid. The membrane acted to remove water from the 15 wt% feed,with permeate lactic acid concentration at 2 wt% corresponding toa water selectivity factor of 9. Silica membranes on the other handexhibited a water selectivity factor up to 220 (a rejection coef-cient of 0.995) with lactic acid in the permeate as low as 0.08 wt%.

Thestrong surface charge andwider pore size of the alumina mem-brane enabled a slowporeblockingmechanism, withux droppingtowards that of the silica membrane. The silica membrane wastherefore the choice technology as the tight pore spaces inhibitedlactic acid from entering and the charge-neutral surface leadingto a more stable separation not subject to pore blocking. As theux of this kind of membrane is very low, commercial applicationwill need further development in such membranes towards uxenhancement.

To eliminate the need for frequent membrane replacement,Polom and Szaniawska [39] used zirconium (IV) hydrous oxide-polyacrylate dynamically formed NF membranes for productionand purication of lactic acid from waste lactose. Though suchmembranes werefound to retain their mechanical strengthfor longduration, the associated ux was low and cost of such membranes

is still high. In this investigation they observed higher electrolyterejection with increasing pH.

A combined process of nanoltration and reverse osmosis wasdeveloped by Li et al. [40] for separation and concentration of lac-tic acid from cheese whey fermentation broth. Five NF membrane(CK, DK, DL, HL, and GE) and two reverse osmosis membrane (DS11 AG and ADF) were tested at different pressures. Highest lac-tose retention (97 1%) was obtained by HL NF membrane witha moderate permeate ux of 33.0 L/m 2 h compared to DK and DL membrane (lactose retention were 88.2% and 69.8% respectively)but the lactic acid retention was also high 43.7% (compared withretention of 38% and 26.1% with other membranes) for this type of membrane which was not desirable. Permeate ux was increasedwith increasing pressure. After nanoltration reverse osmosis wasapplied to concentrate lactic acid. 100% lactic acid retention wasachieved at 5.5MPa pressure for ADF membrane compared toDS11 AG membrane 96% at the same pressure. In this case also,membranes suffered quick fouling in absence of prior ltration of cells.

3.3. Electrodialysis

Fermentation based process requires maintenance of near neu-tral pH for high productivity andthis necessitates addition of alkaliin most of the cases. Alkali addition produces salt of lactic acidinstead of lactic acid itself. To overcome this salt problem, electro-dialysisbasedprocesses thatdo notrequire additionof acidor alkaliagain to convert lactate salts into lactic acid have attracted atten-tion of several researchers. The problem of disposal of by-productgypsum associated with conventional fermentation process can belargely overcome through electrodialysis. Datta et al. [41] studiedthe advantages associated with such electrodialysis process.

Fig. 11. Typical electrodialysis cell conguration.


8/11


Electrodialysis technology a schemeof whichis shown in Fig. 11is based on electromigration of ions through a stack of cation andanion exchange membranes. Basically, it involves two steps. Therst step called conventional electrodialysis (CED) separates andconcentrates lactate salts. The second step called bipolar electro-dialysis (BED) converts lactate salts into lactic acid.

Habova et al. [42] has described such a two-stage electrodialy-sis process for concentration of sodium lactate from fermentationbroth in the rst stage and its subsequent electroconversion intolactic acid in the second stage. They achieved 151 g/L of lactic acidin the nal stage.

Lee et al. [43] integrated an electrodialysis unit consisting of 5 pairs of cation and anion exchange Neosepta membranes witha fermentor (3.7L stirred) for production of lactic acid from cornsteep liquor (CSL) using lactobacillus strain.

Mathieu Bailly [44] studied performance of electrodialysis withbipolar membranes (EDBM) in connection with an industrial unit.He has given the economics of such a plant and claimed that envi-ronmental impact of such a plant is low.

Efcient electrodialysis needs cell free lactic acid. Thus litera-ture shows many such studies that include micro ltration, ultraltration or nanoltration prior to electrodialysis.

Danneret al. [45] developedan ultra-ltration membranebiore-actor coupled with on-line monopolar electrodialysis to recover,pre-purify and concentrate lactate. They started on-line electro-dialysis of the ultraltration permeates after 159 h of cultivationand the fermentation was completed at 1052 h without any micro-bialcontamination.But thevolumetricproductivity(1.38g/L h)waslow and lactic acid concentration was 35 g/L with maximum yieldof 0.80.

Reimann [46] integrated ultraltration, softening, electrodialy-sis, ion exchange andevaporationstepswith a fermentorto achieve99% removal of impurities from lactic acid and a free lactic acidconcentration of 183 g/L. But the process is still cumbersome andinvolves high cost due to so many steps.

Intolerance to multivalent ions is a problem withelectrodialysismembranes.

Bouchoux et al. [36] observed that at sheet cross ow NFmembrane module prior to electrodialysis was quite efcientin separation of calcium and magnesium ions from lactic acidfermentation broth. Magnesium and calcium rejections were64 7 and 72 7%, respectively and lactate recovery rate reached25 2 mol/m/h for P =20bar.

Energy consumption associated with ED processes if normallyhigh and to make the economically attractive attempts have beenmade to increase energy efciency. A modied two-phase eletro-electrodialysys (MTPEED) process developed by Yi et al. [47]

showed a strong potential in the recovery and concentration of lactic acid without generating salt waste compared to EED pro-cess. The process was more effective and more economic. Thecurrent efciency was 100% for MTPEED process compared to EEDprocess having 68.2% and the specic energy consumption waslower 1.404 kWh/kg than EED process 1.956kWh/kg at currentdensity 160.7A/m 2 . The cell voltage was depending on phase ratiooforganicphaseto waterphase andlower phase ratio was desirablefor MTPEED. The electro-electrodialyzer made of PTFE is dividedinto one cathode compartment and one anode compartment by ananion- exchange membrane of 0.005 m thickness and 0.0056m 2

surface area.Though several studies have been undertaken to establish the

potential of bipolar electrodialysis as an efcient and eco-friendlymethod of lactic acid production, concentration and purication,commercialization of bipolar membrane itself has been done invery limited cases and electrodialysis fermentation (EDF) for lac-tic acid has hardly been commercialized. Due to poor conductivityof the organic phase power consumption in electrodialysis processis always high.

4. Integrated membrane processes

The situation improved when UF of the broth was done priorto nanoltration or reverse osmosis. Gonzalez et al. [48] recoveredlacticacidfromultralteredwheyby twotypes of membranes- spi-ral wound DK 2540C and tubular AFC80nanoltration membranes.Lactic acid transport was strongly affected by pH. At increasing pH,lactate retention was higher but the permeate ux decreased. Fora constant permeate ux of 60L/m 2 h in the pH range of 6.0, DKmembranesachieveda lactate rejectionof 1091% whereas for AFCmembranes, rejection was 4582% as the molecular weight cut-off was higher in DK membranes.

US Patent [49] claimed to develop a membrane-integrated pro-cess that integrates one ceramic tubular UF module in the rststage cell separation from fermentation broth and NF in the secondstage for lactic acid separation. The second stage module is a spiralwound type. The process claims to achieve high cell concentration5 10 9 CFU/mL. Ammonium lactate produced does not precipitateallowing ltration and continuous removal. Lactic acid concentra-tion of 12 wt% is attained using lactobacillus strain on sacchariedmash of grain as carbon source.

To ensure long term operation without membrane foulingTejayadi and Cheryan [50] integrated hollow-ber ( Fig. 5) ultra-ltration module for clarication and cell recycling and a reverseosmosis system for preconcentrating whey permeates with a fer-mentor. Thus productivity was 22.5 g/Lh with a yield of 0.89 in this

Table 3Comparative cost factors of some lactic acid production processes.

Process Major cost factors Major merits/demerits Reference

Conventional fermentat ion mult iple operat ion xed capita l, labor,energy, maintenance and chemical costOperation cost is 50% of the total cost

low raw material cost, gives highly pureproduct form but not a clean technology,generates 1 ton waste/ton acid

[1]

Chemical synthesi s p rocess High feed and Equipment cost fails to yi eld pu re produc t and not a cleantechnology

[2]

RO-integrated fermentation RO membrane cost 28% of xed cost high productivi ty of 22.5g/L/h Cleantechnology

[26]

Electrodialysis-integrated fermentation multiple operation involving 80% of total cost for 70% conc. acid, cost US$1.15/ton (in 2000)

Clean technology pure product [3]

UF, RO, Ion-exchange-Integrated fermentation Concentration and fermentationinvolved 40% and 47% of total costrespectively for 50% conc. acidCost US$ 1.25/kg (in 2007)

Clean technology, pure product [51]

RO, NF-integrated fermentat ion Except membrane cost , o ther costfactors on energy, labor, equipment arevery low

Clean technology, high productivity and purity [52,53]


9/11


membrane recycle fermentor and the concentration of lactate was89g/L obtained in 150 h of operation. But the membrane units con-tributed about 28% of the total xed capital costs. However thereare some adverse effects of this lactate on the performance of RO.

Majority of the research efforts have been directed towardsproduction of high purity lactic acid in fermentative process withautomatic or manual control of pH through addition of NaOH,Ca(OH)2 or NH 4 OH resulting in production of lactate salts insteadof lactic acid directly. This necessitated concentration and conver-sion of lactate salts into lactic acid through further processing likeextraction,evaporation, precipitationfollowed by acidication, andelectrodialysis. Suchdownstreamprocessing stepsnot onlyinvolvehigh energy requirement but add to overall production cost. Com-parison of major cost factors associated with a traditional processvis--visthe membranebased processesas shown in Table 3 clearlyestablishes edge of the latter over the former mainly on accountof downstream processing costs and raw material cost. Table 3also shows that membrane based processes represent clean tech-nology as need for multiple steps along with many chemicals of the conventional processes is eliminated. High productivity in acontinuous membrane based process adds to the economy of theprocess. As membranes with high selectivity can be manufacturedand used in compact modular designs, high purity of the product isensured in a membrane integrated process. Membrane integratedprocesses are also better than conventional processes in terms of energy consumption and labour costs as the plants are compactinvolving only a very few steps and no phase change is involved.However, Table 3 shows that multiple operations are involved inelectrodialysis based process and 80% of the total cost is due theseoperations. In UF, RO and ion-exchange processes, very dilute lac-tic acid is produced and hence concentration involves 4050% of the total cost. Though the major cost contributor in a membranebased process is the membrane, these costs are likely to comedown with development of membranes with better mechanicalstrength and antifouling characteristics. This indicates potentialscope of process intensication through integration of membraneswith conventional fermentative process. Integration of micro or

ultraltration membranes mostly in tubular, hollow ber or spiralwound modules with fermentor have been done for cell separa-tion and recycle. Direct reverse osmosis or nanoltration of cellfree broth for separation and purication of lactic acid have beeninvestigated in very few studies. Economic evaluation has beendone for certain categories of schemes which have unnecessarilyrelied heavily on some specic feedstock involving cumbersomeprocessing. An exhaustive evaluation considering both fermenta-tive conventional as well as fully membrane-integrated processesfor separation of both cells and lactic acid is desired to help choosethe best production scheme. Efforts should be directed towardsdeveloping a fully membrane-integrated fermentative process foroperation in continuous mode with continuous cell separation andrecycle and continuous separation of lactic acid directly as perme-

ate of a properlyselectednanoltrationmodule thatovercomes theproblem membrane fouling. Thus the scheme presented in Fig. 12could be one such membrane-integrated continuous productionschemehavingthe potentialof achieving themajorgoals of processintensication in lactic acid manufacturing.

5. Economic evaluation of lactic acid production inmembrane-integrated processes

The conventional fermentative process of lactic acid productioninvolves a series of steps like precipitation, ltration, acidication,carbon adsorption, evaporation and crystallization and 50% of thecost of lactic acid manufacture is consumed in these processingsteps. Moreover, 1tonne CaSO 4 is generated as solid waste per ton

of lactic acid produced. Thus the conventional process is neither

economicnor environmentallybenign.Manyintegrated membranesystems have been studied for the production of lactic acid but theeconomic analysis has been done in few studies.

While studying the techno-economic feasibility of a membranerecycle fermentor, Tejayadi and Cheryan [50] have given the costbreak up which shows that for a 89 g/L concentrated lactic acidwith a productivity of 22.5 g/L h and yield of 0.89, cost componentfor the RO membrane unit constitutes about 28% of the xed costand 5% of the total operating cost. The two largest costs were wheytransportation cost (35%) and yeastextract cost (38%). To minimizethewhey transportation cost, theysuggested locatingthe lactic acidplant within the cheese plant itself.

Akerberg et al. [3] have done an economic evaluation of lac-tic acid production from whole-wheat our consisting of starchand brancontaining nutrients. They compared production costs for70% concentrated lactic acid in a 3X10 4 tonnes/year capacity plantunder different production schemes with the purpose of identify-ing the bottlenecks of production. They observed that operationalcost contributed80% of the total cost andthemajoroperationalcostcomponents were raw material, saccharication, fermentation andelectrodialysis. Lactic acid production cost per kg was found to bearound US$ 1.101.15 in 2000. But most of the cost componentsare very much raw material-specic and scheme-specic whereasa number of better production schemes and better raw materialoptions are available.

An integrated process for production and purication of lacticacid by fermentation was economically evaluated by Gonzalez etal. [51] . They arrived at a calculated annual cost of US$ 1.25/kgfor 50% concentrated lactic acid. The integrated process consistedof fermentation, ultraltration, ion-exchange, reverse osmosis andvacuum evaporation steps. It was observed that 40% of the totalcost was contributed by concentration step alone whereas thelargest cost component (47% of the total cost) was fermentationcost. Concentration step involved high cost due to depreciationof the equipment (41%) and the main cost contribution in clari-cation by ultraltration came from replacement cost of ceramicmembrane. The total annual cost was high in fermentation pro-

cess 52% followed by ion exchange process 27%. In fermentationstep, cost of the yeast extract used as supplement was high (54%)and it could be reduced by using other supplement like whey pro-teins butthe productivitywas lowin that case. Cost of yeast extractcould be reduced by operating in continuous mode and recyclingthe biomass and protein to the reactor so that productivity would

Fig. 12. A promising continuous fermentative lactic acid production scheme with

two-stage integration of membranes representing high process intensication


10/11


also increase. In the ion exchange process, resin regeneration cost(91%) wasthe highest. Vick Roy [52] , Timmer etal. [53] haveshownthat continuous production of lactic acid in cell recycle fermentorwere economically favourable.

Preliminary economic evaluation for a 1000 tonnes/year plantfor four different production schemes- reactive extraction (non-membrane and batch fermentation), ion exchange and adsorption(non-membrane, continuous fermentation), conventional batchfermentation (non-membrane and calcium lactate precipitationtype) and monopolar and bipolar electrodialysis can also be foundin Jogelekar et al. [54] . But they largely bypass the membrane-integrated processes that represent high process intensication.

6. Conclusion

In view of the soaring demand of lactic acid in the market, mul-titude of research activities have been carried out for production of lactic acid in fermentative process using cheap, renewable carbonsources. In several membrane based studies, successful separationof the components of fermentationbrothby micro, ultra, nano,andreverse osmosis and electrodialysis membranes has been demon-strated. However, integration of suchmembranes witha fermentorin a singlestage has failedto achieve allthe goals of process intensi-cation in lactic acidmanufacturing.A judiciouscombinationof thetypes of membranesand themoduleshas thepotential of achievingthesegoals. It appears thatintegrationof microltration membranein the rst stage followed by a nanoltration membrane in thesecondstage ina atsheet cross owmodulecanhelpreachthetar-gets of commercial production of monomer grade lactic acid withhigh productivity and concentration in an environmentally benignprocess. Such a system should be studied in detail for scale up.

References

[1] R. Datta, M. Henry, Lactic acid: recent advances in products, processes andtechnologiesa review, J. Chem. Technol. Biotechnol. 81 (2006) 11191129.

[2] Niju Narayanan, Pradip K. Roychoudhury, Aradhana Srivastava, L(+)lactic acidfermentation and its product polymerization, Microb. Biotechnol. 7 (2) (2004)

115.[3] C. Akerberg, G. Hofvendahl, B. Halm-Hagerdal, Modeling the inuence of pH,temperature,glucoseand lacticacid concentrationson thekineticsof lacticacidproduction by Lactobactococcus lactis ssp Lactis ATCC 19435 in whole-wheatour, Appl. Microbiol. Biotechnol. 49 (1998) 682690.

[4] P.M.Yin, N. Nishina, Y. Kosakai, K. Yahiro, Y. Park, M. Okabe,Enhanced produc-tion of L (+)-lactic acid from corn starch in a culture of Rhizopus oryzae usingan air-lift bioreactor, J. Ferment. Bioeng. 84 (1997) 249253.

[5] Frederico V. Pasos, Henry P. Fleming, David F. Ollis, Kinetics and modeling of lactic acid production by Lactobacillus plantarum , Appl. Environ. Microbiol. 7(1994) 26272636.

[6] A.E. Ghaly, M.S.A. Tango, M.A. Adams, Enhanced lactic acid production fromcheese whey with nutrient supplement addition, Agricult. Eng. Int.: the CIGR J.Sci. Res. Dev. 5 (2003) 0209.

[7] M.G. Adsul, A.J. Varma, D.V. Gokhale, Lactic acid production from waste sugar-cane bagasse derived cellulose, Green Chem. 9 (2007) 5862.

[8] Arti Dumbrepatil, Mukund Adsul, Shivani Chaudhari, Jayant Khire, DigambarGokhale, Utilization of molasses sugar for lactic acid production by Lactobacil-lus delbrueckii subsp. Delbrueckii Mutant Uc-3 in batch fermentation, Appl.Environ. Microbiol. 74 (2008) 333335.

[9] Walaiporn Timbuntam, Klanarong Sriroth, Yutaka Tokiwa, Lactic acid produc-tion from sugarcanejuiceby a newly isolated Lactobacillus sp, Biotechnol. Lett.28 (2006) 811814.

[10] R. Chen, Y.Y. Lee, Membrane-mediated extractive fermentation for lactic acidproduction from cellulosic biomass, Appl. Biochemist. Biotechnol. 63 (1997)435448.

[11] C.R. Soccol, V.L. Stonoga, M. Raimbault, Production of L-lactic acid by Rhizopusspecies, World J. Microbiol. Biotechnol. 10 (1994) 433435.

[12] E.N. Kaufman, S.P. Cooper, S.L. Clement, M.H. Little, Appl. Use of a biparticleuidized-bed bioreactor for the continuous and simultaneous fermentationand purication of lactic acid, Appl. Biochem. Biotechnol. 5152 (1995)605620.

[13] R.L. Evangeliesta, A.J. Mangold, Z.I. Nikolov, Recovery and purication of lacticacid fromfermentation broth by adsorption, Appl. Biochem. Biotechnol. 4546(1994) 131144.

[14] L.A. Tung, C.J. King, Sorption and extraction of lactic and succinic acids atpH>pK a1 . I: Factors governing equilibria, Ind. Eng. Chem. Res. 33 (1994)

32173223.

[15] K.L. Wasewar, Separation of lactic acid: recent advances, Chem. Biochem Eng.Q. 19 (2005) 159172.

[16] Y. Tong, M. Hirata, H. Takanashi, T. Hano, F. Kubota, M. Goto, F. Nakashio, M.Matsumoto, Extraction of lactic acid from fermented broth with microporoushollow ber membranes, J. Membr. Sci. 143 (1998) 81.

[17] S.L.Matson,Membraneseparationstechnology, in: R.D.Noble, S.A.Stern (Eds.),Principles and Applications, Elsevier, Amsterdam, 1995 (Chapter 8).

[18] L.M.D. Goncalves, A.M.R.B. Xavier, J.S. Almeida, Concomitant substrate andproduct inhibitionin lacticacid production, EnzymeMicrob. Technol.13 (1991)311319.

[19] I. Ohara, Keiichiro Hiyama, Toshiomi Yoshida, Kinetics of growth and lactic

acid production in continuous and batch culture, Appl. Microbiol. Biotechnol.37 (1992) 544548.[20] W. Fu, A.P. Mathew, Lactic acid production from lactose by Lactobacillus Plan-

tarum : kinetic model and effects of pH, substrate, and oxygen, Biochem. Eng. J.3 (1999) 163170.

[21] Stephen Milcent, Helene Carrere, Clarication of lactic acid fermentationbroths, Sep. Purif. Technol. 2223 (2001) 393401.

[22] Massayuki Taniguchi, Nobuharu Kotani, Takeshi Kobayashi, High concentra-tion cultivation of lactic acid bacteria in fermentor with cross-ow ltration, J.Ferment. Technol. 65 (2) (1987) 179184.

[23] H. Moueddeb,J. Sanchez,C. Bardot, M. Fick,Membrane bioreactorfor lactic acidproduction, J. Membr. Sci. 114 (1996) 5971.

[24] Lidietta Giorno, Katarzyna Chojnacka, Laura Donato, Enrico Drioli, Study of acell-recycle membrane fermentor for the production of lactic acid by Lacto-bacillus bulgaricus , Ind. Eng. Chem. Res. 41 (2002) 433440.

[25] A.M.R.B. Xavier, L.M.D. Goncalves, J.L. Moreira, M.J.T. Carrondo, Labaratorio deEngenhariaBioquimica,Faculdadee de Cienciase Technologia, Operational pat-terns affecting lactic acid production in ultraltration cell recycle bioreactor,Biotechnol. Bioeng. 45 (4) (1995) 320327.

[26] S. Kwon, I.K. Yoo, W.G. Lee, H.N. Chang, Y.K. Chang, High rate continuousproduction of lactic acid by Lactobacillus bulgaricus in two-stage membranecell-recycle bioreactor, Biotechnol. Bioeng. 73 (2001) 2534.

[27] Levente L. Diosady, Taya Puzanov, Membrane Fermentation of Lactic acid, Int. J. Appl. Sci. Eng. 3 (1) (2005) 1925.

[28] Anna Torang, Ann-So Jonsson, Guido Zacchi, Separation of Cells and ProteinsfromFermentationbrothin a shear-enhancedcross-owultraltrationmoduleas rst step in renement of Lactic acid, Appl. Biochem. Biotechnol. 76 (1999)144157.

[29] J.P.S.G. Crespo, A.M.R.B. Xavier, M.T.O. Bareto, L.M.D. Goncalves, J.S. Almeida,M.J.T. Carrondo,Tangentialow ltration for continuouscell recycle culture of acidogenic bacteria, Chem. Eng. Sci. 47 (1992) 205214.

[30] A. Nishiwaki, I.J. Dunn, Performance of a two stage fermentor with cell recyclefor continuous production of lactic acid, Bioprocess Eng. 21 (1999) 299305.

[31] Guo-qian Xu, Ju Chu,Yong-hong Wang,Ying-ping Zhuang, Si-liang Zhang, Hua-qiongpeng, Development of a continuouscell-recyclefermentation system forproduction of lactic acid by Lactobacillus paracasi , Process Biochem. 41 (2006)24582463.

[32] J. Sikder, C. Pereira, S. Palchoudhury, K. Vohra, D. Basumatary, P. Pal, Synthe-sis and characterization of cellulose acetate- polysulfone blend microltrationmembranefor separation ofmicrobialcells fromlactic acidfermentationbroth,Desalination, Elsevier Science, doi:10.1016/j.desal.2008.11.024 .

[33] J.M.K. Timmer, H.C. van der Horst, T. Robbertsen, Transport of lactic acidthrough reverse osmosis and nanoltration membranes, J. Membr. Sci. 85(1993) 205216.

[34] J.M.K. Timmer, H.C. van der Horst, T. Robbertsen, Lactic acid separation fromfermentation broths by reverse osmosis and nanoltration, J. Membr. Sci. 92(1994) 185197.

[35] M.K.H. Liew, S. Tanaka, M. Morita, Separation and purication of lactic acid:Fundamental studies on the reverse osmosis down-stream process, Desalina-tion 101 (1995) 269277.

[36] Antoine Bouchoux, Helene Roux-de Balmann, Florence Lutin, Nanoltration of glucose and sodium lactate solutions Variations of retention between single-and mixed-solute solutions, J. Membr. Sci. 258 (2005) 123132.

[37] R. Jeantet, J.L. Maubois, P. Boyaval, Semi continuous production of lacticacid ina bioreactorcoupledwith nanoltrationmembranes, Enzyme Microb. Technol.16 (1996) 614619.

[38] Mikel C. Duke, Agnes Limb, Sheila Castro da Luz, Lars K.AT Lactic acid enrich-ment with inorganic nanoltration and molecular sieving membranes bypervaporation Nielsen, Food Bioproducts Process. 86 (4) (2008) 290295.

[39] Ewa Polom, Daniela Szaniawska, Rejection of lactic acid solutions by dynam-ically formed nanoltration membranes using a statistical design method,Desalination 198 (2006) 208214.

[40] Yebo Li, Abolghasem Shabazi, Karen Williams, Caixia Wan, Separate and con-centrate lactic acid using combination of nanoltration and reverse osmosismembranes, Appl. Biochem. Biotechnol. 147 (2008) 19.

[41] R. Datta, S.P. Tsai, P. Bonsignore, S.H. Moon, J.R. Frank, Technological and eco-nomic potential of poly(lactic acid)and lactic acid derivatives, FEMS Microbiol.Rev. 16 (1995) 221231.

[42] V. Habova, K. Melzoch, M. Rychtera, B. Sekavova, Electrodialysis as a usefultechnique for lactic acid separation from a model solution and a fermentationbroth, Desalination 162 (2004) 361372.

[43] KiBeom Lee, A media design program for lactic acid production coupled withextraction by electrodialysis, Bioresour. Technol. 96 (2005) 15051510.

[44] Mathieu Bailly, Productionof lacticacidsby bipolar electrodialysis: realizationsand perspectives, Desalination 144 (13) (2002) 157162.


11/11


[45] H. Danner, L. Madzingaidzo, C. Thomasser, M. Neureiter,R. Barun, Thermophilicproduction of lactic acid using integrated membrane bioreactor systems cou-pled with monopolar electrodialysis, Appl. Microbiol. Biotechnol. 59 (2004)160169.

[46] W. Reimann, Down streaming of lactic acid from hydrolysate of barley afterfermentation, Agricult. Eng. Int: the CGIR Ejournal, Manuscript, FP 05 004 VII,2005.

[47] S.S. Yi, Y.C. Lu, G.S. Luo, Separation and concentration of lactic acid by electro-electrodialysis, Sep. Purif. Technol. 60 (2008) 308314.

[48] M.IsabelGonzalez, SilviaAlvarez, FranciscoA. Riera, Ricardo Alvarez,Lactic acidrecoveryfrom wheyultraltrate fermentationbroths and articialsolutionsby

nanoltration, Desalination 228 (2008) 8496.[49] Russo L.J., Kim, H, Membrane-based process for the recovery of lactic acidby fermentation of carbohydrate substrates containing sugars, US Patent No.5,503,750.

[50] S. Tejayadi, M. Cheryan, Lactic acid from cheese whey permeates, produc-tivity and economics of a continuous membrane bioreactor, Appl. Microbiol.Biotechnol. 43 (1995) 242248.

[51] M.I. Gonzalez, S. Alvarez, F. Riera, R. Alvatez, Economic evaluation of an inte-grated process for lactic acid production from ultraltered whey, J. Food Eng.80 (2007) 553561.

[52] Vick Roy, T.B. Lactic acid production in membrane fermentors, Ph.D. Thesis,University of California, Berkeley, CA, 1983.

[53] J.M.K. Timmer, J. Kromkamp,Efciencyof lacticacid productionby LactobacillusHelvetius in a membrane cell recycle reactor, FEMS Microbiol. Rev. 14 (1994)2938.

[54] H.G. Joglekar, Imran Rahman, Suresh Babu, B.D. Kulkarni, Ajit Joshi, Compar-ative assessment of downstream processing options for lactic acid, Sep. Purif.Technol. 52 (2006) 117.

Process intensification in lactic acid production

Documents

Transcript of Process intensification in lactic acid production