lactic_rev_SEPPUR8558.pdf

Embed Size (px)

Citation preview

  • 7/28/2019 lactic_rev_SEPPUR8558.pdf

    1/18

    This article was originally published in a journal published by

    Elsevier, and the attached copy is provided by Elsevier for the

    authors benefit and for the benefit of the authors institution, for

    non-commercial research and educational use including without

    limitation use in instruction at your institution, sending it to specific

    colleagues that you know, and providing a copy to your institutions

    administrator.

    All other uses, reproduction and distribution, including without

    limitation commercial reprints, selling or licensing copies or access,

    or posting on open internet sites, your personal or institutions

    website or repository, are prohibited. For exceptions, permission

    may be sought for such use through Elseviers permissions site at:

    http://www.elsevier.com/locate/permissionusematerial

    http://www.elsevier.com/locate/permissionusematerialhttp://www.elsevier.com/locate/permissionusematerial
  • 7/28/2019 lactic_rev_SEPPUR8558.pdf

    2/18

    Autho

    r's

    pe

    rsonal

    copy

    Separation and Purification Technology 52 (2006) 117

    Review

    Comparative assessment of downstream processing options for lactic acid

    H.G. Joglekar , Imran Rahman, Suresh Babu, B.D. Kulkarni, Ajit Joshi

    Chemical Engineering and Process Development Division, National Chemical Laboratory (NCL), Pune-411008, India

    Received 14 October 2005; received in revised form 20 March 2006; accepted 20 March 2006

    Abstract

    The possibility of manufacturing a biodegradable polymer from lactic acid has led to extensive research in recovery of lactic acid produced

    by fermentation, by different downstream processing routes. This paper assesses the suitability of different downstream processing options suchas reactive extraction, adsorption, electrodialysis, esterification and reactive distillation. It compares the costs of different process routes. The

    assessment indicates that the conventional precipitation of calcium lactate, followed by acidification, esterification and hydrolysis will be the most

    economical route although it generates large quantity of gypsum sludge.

    2006 Elsevier B.V. All rights reserved.

    Keywords: Lactic acid; Reactive extraction; Adsorption; Electrodialysis; Esterification; Reactive distillation

    Contents

    1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

    1.1. Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

    1.2. Global scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.3. Polylactic acid/bio-degradable plastics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

    1.4. Lactic acid production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

    1.5. Source of carbohydrates for fermentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

    1.6. Fermentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

    1.7. Downstream processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

    2. Reactive extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

    2.1. Extractant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

    2.2. Diluents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

    2.2.1. Choice of diluents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

    2.3. Effect of extractants, diluents and pH on distribution coefficient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

    2.4. Third phase formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

    2.5. Complex formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

    2.6. Re-extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

    2.6.1. Using NaOH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

    2.6.2. Using HCl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

    2.6.3. Temperature-swing regeneration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

    2.6.4. Using trimethyl amine (TMA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

    2.6.5. Diluent swing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

    2.7. Choice of extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

    3. Adsorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

    3.1. Reported research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

    3.2. Choice of adsorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

    Corresponding author. Tel.: +91 20 25902420; fax: +91 20 25902675.

    E-mail address: [email protected] (H.G. Joglekar).

    1383-5866/$ see front matter 2006 Elsevier B.V. All rights reserved.

    doi:10.1016/j.seppur.2006.03.015

  • 7/28/2019 lactic_rev_SEPPUR8558.pdf

    3/18

    Autho

    r's

    pe

    rsonal

    copy

    2 H.G. Joglekar et al. / Separation and Purification Technology 52 (2006) 117

    4. Electrodialysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

    4.1. Reported research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

    4.2. Choice of electrodialysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

    5. Esterification and reactive distillation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

    5.1. Reported research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

    5.2. Choice for esterification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146. Possible downstream process routes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

    7. Future direction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

    8. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

    References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

    1. Introduction

    1.1. Objective

    Lactic acid is a simple compound containing both the

    hydroxyl and carboxylic acid groups, which permit it to par-ticipate in many interesting and valuable chemical reactions.

    Among the known reactions of lactic acid the dehydration to

    acrylic acid, the polymerization to poly(lactic acid) and conden-

    sation to 2,3-pentanedione are potentially the most profitable

    pathways. The viability of making an inexpensive biodegradable

    polymer poly(lactic acid) from lactic acid has sparked exten-

    sive interest andresearch in the area of producing andrecovering

    relatively pure lactic acid. This paper is to assess various down-

    stream processing options studied by researchers and identify

    affordable process route for manufacturing lactic acid, which is

    sufficiently pure for polymerization.

    1.2. Global scenario

    Lactic acid is currently being produced by number of compa-

    nies around the world, but it is only used in limited applications

    in the food, flavor as preservative, in cosmetics, leather process-

    ing, and pharmaceuticals industries. Lactic acid of higher purity

    or without certain impurities is required for polymerization.

    The possibility of producing a high volume of inexpensive

    lactic acid also has lead to research using lactic acid as an

    alternative feedstock for the production of many speciality and

    commodity chemicals such as acrylic acid, propionic acid, 2,3-

    pentanedione, pyruvic acid and propylene glycol.

    The world market of lactic acid is growing every year. Thelevel of production is around 350 millions kg per year [1] and the

    worldwide growth is believed by some observers to be 1215%

    per year [2].

    1.3. Polylactic acid/bio-degradable plastics

    The main application field of lactic acid polymer has been

    medical applications, fibres, packaging materials and as sol-

    vents. These medical applications include its usage utilizing

    different properties in terms of tensile strength, viscosity, purity

    etc. l-Lactic acid polymer exist in three different forms that can

    be used for filling the gaps in bones, solid with tensile strengthto produce sutures (stitching material), and the glue form that

    is mainly applied in joining membranes or thin skins in humans

    [3]. Another important property of poly lactic acid is its high

    strength against ultraviolet (UV) radiation [4].

    Polylactic acid has a potential to provide a new product plat-

    form to compete with hydrocarbon-based thermoplastic. Poly-

    lactic acid polymers have barrier properties similar to polyethy-

    lene terephthalate (PET) and gloss, clarity and processabilitysimilar to polystyrene (PS). In addition, they provide heat stabil-

    ity at lower temperatures than polyolefins and can be processed

    by most melt fabrication techniques, including fiber spinning.

    The research on lactic acid related materials have attracted

    several universities and institutes in Europe, Asia and USA.

    There are a number of small-scale production facilities of poly-

    lactic acid. Dow Chemicals and Cargill have the largest poly-

    lactide (PLA) producing company with an annual capacity of

    140,000 tonnes located in Blair, USA [5].

    The future of the polylactic acid depends upon the technol-

    ogy efforts of lactic acid manufacturers to develop Polylactic

    acid modified with other co-monomers, so as to compete with

    polymers such as PETand PS,both on cost andapplicationbasis.

    1.4. Lactic acid production

    Lactic acid can be manufactured by either chemical synthe-

    sis or by fermentation of renewable carbohydrates. Both these

    methods are used for commercial production. While chemical

    synthesis route produces only a racemic mixture of lactic acid, a

    fermentation process can produce a stereoisomer of lactic acid.

    Therefore, fermentation process is preferred for the production

    of lactic acid and about 90% of lactic acid is produced by fer-

    mentation route. Currently the fermentation is carried out in

    the batch mode. A large number of carbohydrate materials havebeen used, tested or proposed for manufacture of lactic acid by

    fermentation.

    1.5. Source of carbohydrates for fermentation

    Sucrose from cane and beet, glucose, and whey contain-

    ing lactose, maltose and dextrose from hydrolyzed starch are

    presently used commercially. Molasses and whey are gener-

    ally low-priced sources of sugars for fermentation. Sulfite waste

    liquor and Jerusalem artichokes are potential sources of lactic

    acid.

    Sugar (or dextrose) is the most pure source, but is very costly.Molasses and other wastes are cheap but contain large number

    of impurities in larger quantities, which affect downstream pro-

  • 7/28/2019 lactic_rev_SEPPUR8558.pdf

    4/18

    Autho

    r's

    pe

    rsonal

    copy

    H.G. Joglekar et al. / Separation and Purification Technology 52 (2006) 117 3

    cesses. Sugar cane juice can be a compromise between quality

    and cost. Fermentation conditions also determine the quality of

    broth and suitability of downstream processing options.

    1.6. Fermentation

    Carbohydrates are fermented to l-lactic acid by a number of

    microorganisms. As lactic acid is formed and accumulated, pH

    of the fermentor broth goes on lowering. This affects the produc-

    tivity of microorganisms. Considerable research is being carried

    out to develop strains, which canproduce lactic acid at lowerpH.

    However it is still not a commercially viable preposition. In con-

    ventional process pH of fermentor is continuously maintained

    by addition of lime and thus converting lactic acid (partly) to

    calcium lactate. Downstream processing of this broth includes

    precipitation and acidification of calcium lactate to lactic acid.

    This generates huge quantity of calcium sulphate, which poses

    an environmental problem. Alternatively lactic acid can be con-tinuously removed from the fermentor by adsorption, extraction

    or membrane separation. This is possible if fermentation is in

    continuous mode. Alkalis other than calcium, such as ammo-

    nium hydroxide, or sodium hydroxide (or carbonate) can also

    be used. In that case alkalis need to be recovered and recycled

    back.

    1.7. Downstream processing

    Conventionally two routes are followed: (i) the clarified fer-

    mented liquor is concentrated to 32% well above crystallization

    point and acidified with sulphuric acid to get the crude lactic

    acid. (ii) The crude calcium salt which precipitates from the con-

    centrated fermented liquor is crystallized, filtered, dissolved and

    subsequently acidified with sulphuric acid according to process

    described by Inskeep et al. [6] and Peckham [7]. In this route the

    precipitated calcium lactate is separated from dissolved impu-

    rities by filtration and washing. However, some impurities are

    retained in the cake. There is a loss of calcium lactate along

    with the washes. The conventional route generates huge quan-

    tities of calcium sulphate cake. The cake with retained organic

    impurities is difficult to dispose of.

    Filtered fermented broth mainly contains impurities such as

    residual sugar compounds, colour andother organic acids.These

    impurities can be removed by reactive extraction, adsorption,electro dialysis and esterificationhydrolysis with distillation.

    Research work done by various researchers on these purification

    steps is reviewed and assessed here.

    2. Reactive extraction

    Lactic acid is poorly extractable by common organic solvent

    due to their hydrophilic nature. Therefore, reactive extraction

    has been considered for its recovery from aqueous solutions.

    Reactive extraction of the lactic acid by a suitable extractant

    has been found to be a promising alternative to the conventional

    processes.Reactive extraction uses reaction between the extractant and

    the material being extracted. The extractant in the organic phase

    reacts with a material in aqueous phase and reaction complex is

    solubilized into the organic phase. The lactic acid is recovered

    from the organic layer by stripping. The prerequisite of an eco-

    nomic recovery by extraction is a high distribution coefficient.

    Reactive liquidliquid extraction has the advantage that lactic

    acid can be removed easily from the fermentation broth (extrac-

    tive fermentation), preventing the lowering of pH. Further, the

    lactic acid can be re-extracted and the extractant and diluent

    recycled to the fermentation process. However this implies that

    extractant and diluent must not be toxic to the microorganism

    used in fermentation. This can ideally be done for continuous

    fermentation. Wasewar et al. [8], have reviewed and critically

    analyzed the fermentation of glucose to lactic acid coupled with

    reactive extraction.

    Membrane extraction overcomes many drawbacks of the

    extractive fermentation which have plagued us for long, and

    offers other numerous advantages, such as (i) no fear of

    back mixing, (ii) no direct exposure of microbes to extractionreagents, thereby ensuring biocompatibility, (iii) no need for

    agitation, (iv) potentially high efficiency, etc. For the above rea-

    sons, membrane extraction can be considered a very promising

    alternative to the conventional solvent extraction for separation

    and purification of lactic acid [9].

    Huang et al. [10] and Tong et al. [11] successfully developed

    an energy-efficient hollow-fiber membrane extraction process to

    separate and recover lactic acid produced in fermentation.

    2.1. Extractant

    The extractant should have low water solubility, a high dis-tribution coefficient for the lactic acid and low distribution coef-

    ficient for the impurities such as residual sugars [12].

    The distribution coefficient is defined, as ratio of the concen-

    tration of lactic acid in solvent phase to concentration of lactic

    acid is aqueous phase. Organic bases or amine extractant can

    provide much higher equilibrium distribution coefficient (Kd)

    for the extraction of carboxylic acids than conventional solvents.

    Kd =Cs

    Cw, (1)

    Cs is the total concentration of acid in solvent phase at equi-

    librium, expressed as milliequivalent per milliliter; Cw the total

    concentration of acid in water phase at equilibrium, expressedas milliequivalent per milliliter.

    On account of its highly hydrophilic character, lactic acid

    purification with solvent extraction by carbon bonded oxygen

    donor extractants is not practical. Phosphorus-bonded, oxygen-

    donor extractants contain a phosphoryl group which is a stronger

    Lewis base than the carbon-bonded oxygen group. Further-

    more, extractants belonging to this group coextract less water

    and are less soluble in water than the extractants of the carbon-

    bonded oxygen group.

    It has been reported that long chain tertiary amines are suit-

    able for the recovery of carboxylic acid from aqueous solution

    [13]. Due to physical properties of aliphatic amines they mustalways be used in theform of solutionsin organic diluents.These

    organic diluents affect the basicity of amines thereby, extraction

  • 7/28/2019 lactic_rev_SEPPUR8558.pdf

    5/18

    Autho

    r's

    pe

    rsonal

    copy

    4 H.G. Joglekar et al. / Separation and Purification Technology 52 (2006) 117

    behaviorof amineis significantly varied [14]. Recently, the influ-

    ence of the chain length of tertiary amine on extractability has

    been investigated in polar diluent and nonpolar diluent, respec-

    tively [15]. In polar diluent such as 1-butanol, 1-hexanol, and

    1-octanol, the extractability of tertiary amines increases with the

    chain length of amines. However, the extractability of tertiary

    amines decreases with the chain length of amines in nonpolar

    diluent such as n-heptane [15].

    Tertiary amines seem to be more suitable extractants for

    organic acid removal compared to primary and secondary

    amines [16].

    2.2. Diluents

    Long chain aliphatic amines are effective extractant for sep-

    aration of carboxylic acid from dilute aqueous solution. Gen-

    erally, the amine extractants are dissolved in diluent [17], an

    organic solvent that dilutes the extractants to the desired concen-tration and controls the viscosity and density of solvent phase.

    Solvents that contain functional group that interact strongly with

    complex arecalled as theActive Diluent e.g. 1-octanoland that

    of which interact inertly with complex are called Inert Diluent

    e.g. n-hexane.

    Diluents are classified as active diluents and inert diluents

    depending on the interaction with the complex [18,19].

    2.2.1. Choice of diluents

    The choice of suitable diluents [20] for amine is not easy.

    Polar diluents are more favorable than the nonpolar solvents

    as shown in the extraction of lactic acid [12,18,21]. There areseveral important parameters, which must be taken into consid-

    eration.

    2.2.1.1. Distribution coefficient. The distribution coefficient

    value in the extraction step should be greater than 1 at room

    temperature and below 0.1 in the stripping step in order to obtain

    a high concentration in the stripping raffinate.

    2.2.1.2. Toxicity. Toxicity of the organic solvent and extractant

    to microbes is the critical problem in extractive fermentation.

    The degree of toxicity depends on the microbe and extractant

    solution used [8]. Brink and Tramper [22] reported that theleast toxicity is expected from solvents of low polarity and high

    molecular weight.

    Toxic effect of extractants and diluents on microorganisms

    can also be avoided by immobilization of cells. Yabannavar et

    al. [23], reported that addition of soybean oil in immobilized

    cells protects cells from toxic effects of extractants and diluents.

    Tik et al. [24] obtained a maximum total lactic acid concen-

    tration (2.5 times that without extraction) when 15% Alamine

    336 in oleyl alcohol together with immobilized cell with 15%

    sunflower oil was used.

    Wasewar [8] concluded the Alamine 336 in oleyl alcohol,

    which has the highest distribution coefficient, would serve as anideal extraction system unless the higher phase level toxicity of

    octanol is reduced by using an immobilized enzyme system.

    2.2.1.3. Solubility in water. This is important for several rea-

    sons. One is economy and the diluents have to be recovered

    from the raffinate phase by stripping. The physical properties

    of fermentation broth make it desirable to avoid raffinate strip-

    ping. A low concentration of some cheap hydrocarbon can be

    acceptable for this consideration.

    2.2.1.4. Viscosities and density. Low viscosity and low density

    promote phase separation. These properties are best achieved

    with paraffinic hydrocarbons.

    2.2.1.5. Stability. Hydrocarbons are preferable to alcohol and

    halogenated hydrocarbons as they are more stable.

    2.3. Effect of extractants, diluents and pH on distribution

    coefficient

    Table 1 summarizes the reported results for various extractantin diluents and the Kd values. In general, the extraction power of

    extractant is dictated by the basicity of the amine. The diluents

    also affect the basicity of the amine and thus the stability of the

    ion pair formed and its solvation power.

    The use of tertiary amines including tri-n-octylamine (TOA),

    triisooctylamine, and Alamine 336 for the extraction of

    lactic acid from aqueous streams has been often reported

    [12,1719,21,2527]. Tertiary amine extractants are effective,

    with Kd strongly dependent upon the nature of the diluent [13].

    The lactic acid extracted had to be stripped for the final

    production of lactic acid. The reaction of amine and acid is

    exothermic reaction, so the extraction efficiency of lactic acid

    decreases with increasing temperature for active diluents where

    as it shows little effect for inert diluents [16,18,19,27,28]. The

    stripping efficiency does not change with the variation of inac-

    tive diluent except for polar chlorobenzene [19,29].

    The pH of the aqueous phase is an important parameter for

    the reactive extraction of organic acids. Yang et al. [30], reported

    that lower pH values result in good separation of lactic acid by

    long chain tertiary amines. In the intermediate pH range (35),

    Kd decreased with increasing equilibrium pH of the aqueous

    phase. However, in the extremely high and low pH ranges, Kdremained insensitive to pH values. Similarly, Choudhury et al.

    [32], reported that the distribution coefficient of lactic acid was

    high at pH 2.0. In addition, these authors pointed out that theamount of amine in diluents affects the reactive extraction of

    organic acids. Kd increased with increasing amount of aliquat

    336, in their studies. They reported that increasing amine con-

    centration might have a toxic effect on microorganisms in the in

    situ extractive fermentation.

    Temperature and initial organic acid concentration are also

    important parameters for the extraction of organic acids. Yaban-

    navar and Wang [33] investigatedthe extraction of aqueous lactic

    acid by trioctylphosphineoxide in dodecane and Alamine 336

    in oleyl alcohol. In their study, Kd decreased with increasing

    initial lactic acid concentration and increased with increasing

    concentration of Alamine 336 in oleyl alcohol. Martin et al.[21] studied the extraction of lactic acid from aqueous solu-

    tions by Alamine 336 dissolved in toluene. They observed

  • 7/28/2019 lactic_rev_SEPPUR8558.pdf

    6/18

    Autho

    r's

    pe

    rsonal

    copy

    H.G. Joglekar et al. / Separation and Purification Technology 52 (2006) 117 5

    Table 1

    Kd values reported in literature

    Serial number Initial concentration Extractant Diluent Kd Reference

    1 20% w/w TOAtripropylamine (TPA) (2:8w/w) 1 mol/kg 1-Octanol/ n-heptane

    (3:7 w/w)

    0.8 [15]

    10% w/w TOATPA (2:8 w/w) 1 mol/kg 1.2

    7.5% w/w TOATPA (2:8 w/w) 1 mol/kg 4

    5% w/w TOATPA (2:8 w/w) 1 mol/kg 90

    2 20% L.A. Tri-n-butyl phosphate (TBP) 0.95 [34]

    TOA (0.8 M) Chloroform 9.44

    TOA (0.8 M) Tributyl phosphate 7.42

    Tributyl amine (TBA) Methylisobutylketone

    (MIBK)

    0.21

    TOAmethylene chloride (0.6 M) Hexane 2.52

    Chlorobenzene 0.2

    Heptane 2.14

    3 Model media

    containing 90% l+

    lactic acid

    Cynex 923 (0.85 mol/l) Kerosene 2 [35]

    Hostarex-327 (60% w/v) Xylene 4.44

    Trioctylphosphine oxide (TOPO) Xylene 2.2

    4 L.A. (2.3mol/l) TOA (1 M) Chlorobutane 1.4 [36]

    L.A (2.6 mol/l) TOA (1.2 M) Methylene

    chloriden-hexane

    1.45

    5 1.3 mol/l L.A. MIBK 0.13 [17]

    Alamine 336 (0.3 mol/l) MIBK 8.3

    Alamine 336 (0.3 mol/l) Chloroform 5.3

    Trilauryl amine (1 mol/l) Xylene 8.3

    6 0.0051.5 mol/l Alamine 336 940% MIBK decanol,

    octanol

    4.24 [37]

    24.0

    26.0

    7 10 g/l L.A. 50% Alamine 336 Oleyl alcohol 13 [38]100 g/l L.A. 50% Alamine 336 Oleyl alcohol 4

    8 1.6 mol/l L.A. Alamine 336 (40% v/v) MIBK 4.24 [39]

    9 Calcium lactate Alamine (48%v/v) Octanol (30% v/v)

    Isopar K Exxon

    (22%) CO2 (300psig)

    0.62 [40]

    10 79 g/l lactate TOPO (30 wt.%) MIBK (70 wt.%) 1.05 [41]

    Alamine 336 (82 wt.%) 10% Octanol 8%

    isopar K

    3.35

    Alamine (100%) 0.36/5.9

    Alamine 336 (89 wt.%) 9% Dodecanol 2%

    isopar K

    2.6

    that the extraction of lactic acid decreased with increasing

    temperature.

    2.4. Third phase formation

    In extraction of lactic acid with long chain tertiary amines

    dissolved in organic diluents a third phase may form between

    aqueous and organic phase. This phase is not broken by cen-

    trifugation at high rpm. The third phase represents limiting

    conditions for extraction, which must be avoided [42]. For best

    separation of the emulsion that results from the third phase, a

    phase modifier is being added to the organic solvent, represent-ing a long chain aliphatic alcohol, a component assuring a higher

    solubility of the acid complex in the extract phase. Sometimes

    mixed tertiary amine of short chain amine and long chain can

    overcome the third phase formation. No cloudy, bubbly looking

    phase was observed in this study. The prevention of third phase

    makes the phase separation easy, thereby, the settling time gets

    shortened.

    2.5. Complex formation

    Long-chain aliphatic amines are effective extractants for the

    separation of carboxylic acids from dilute aqueous solution. The

    specific chemical interactions between the amines and the acid

    molecules to form acidamine complexesin the extractant phaseallow more acid to be extracted from the aqueous phase. Gen-

    erally, the amine extractants are dissolved in a diluent such as

  • 7/28/2019 lactic_rev_SEPPUR8558.pdf

    7/18

    Autho

    r's

    pe

    rsonal

    copy

    6 H.G. Joglekar et al. / Separation and Purification Technology 52 (2006) 117

    a ketone, an alcohol, chloroform, etc., that is, an organic sol-

    vent that dilutes the extractant to the desired concentration and

    controls the viscosity and density of the solvent phase.

    Many factors have been found to influence the equilibrium

    extraction characteristicsof these systems. Three important vari-

    ables are the nature of the acid extracted, the concentration of

    the extractant, and the type of diluent. These factors have been

    extensively studied for the extractions by tertiary amines of sev-

    eral carboxylic acids, including lactic acid [31,43].

    Eyal et al. [44], summarized the extraction of acid by amines

    and proposed thatfour majormechanisms determine acid extrac-

    tion by amine-based extractants: anion exchange, ion-pair for-

    mation, H-bond formation, and solvation. These four mecha-

    nisms depend on the strength of the amine-based extractant. For

    amines of basicity similar to or higher than that of the anion

    of the extracted acid, extraction is governed by ion-pair forma-

    tion. H-bonding of undissociated acid becomes important only

    for extractions of weak bases. Also, the specific basicity of theextractant (pKa,B), defined as follows, is suggested to express

    the nature of extractant

    R3NH+Ka,BR3N+H

    +, (2)

    Ka,B =[R3N][H

    +]

    [R3NH+]

    , (3)

    where R3N represents TOA. The species in the organic phase

    are marked with an overbar, and all of the concentrations are

    expressed in molar terms. pKa,B

    depends on the amine concen-

    tration and diluent type, representing the structure of the amine,

    multiplied by coefficients for diluent solvation properties and

    the properties of the anion of the extracted acid (hydrophobic

    properties and steric hindrance).

    The reactive extraction equilibrium for polar dilute solu-

    tions of organic solutes can be described by the mass action

    law in which the equilibrium behavior is modeled by postulat-

    ing the formation of various stoichiometric complexes of acid

    and amine. In addition, it is assumed that the carboxylic acid

    is dissociated in water and extracted into the extractant phase

    through the physical solubility of the solute in the diluent of

    the extractant phase and the formation of 1:1, 2:1, 3:1, and

    1:2 acidamine complexes, where the 2:1 and 3:1 complexesresult from the dimer and trimer, respectively, of the acid in the

    extractant phase and the 1:2 complex results from hydrogen-

    bond association between TOA and the hydroxyl and aldehyde

    group on carboxylic acid. The physical extraction with diluent

    and reactive extraction with TOA fit the simple additive model.

    Although the stoichiometry of lactic acidamine extraction

    systems has been studied there is no general agreement. The

    formation of (1:1) acidamine complex has been commonly

    suggested. Prochazka et al. [45] have indicated the forma-

    tion of (1:1), (2:1) and (2:2) complexes for the extraction

    of lactic acid by trialkylamine in mixtures of 1-octanol-n-

    hexane. The quantitative description of the effect of amine andacid concentration on the change in complex composition is

    unclear.

    2.6. Re-extraction

    In reactive extraction, the amine extractant recovers the acid

    by reacting with it to form an acidamine complex that is sol-

    ubilized into the extractant phase. A Second step, referred to

    as regeneration, reverses this reaction to recover the acid into a

    product phase and the acid free extractant, available for recycle.

    The reported regeneration methods for lactic acid from a

    loaded organic phase are described individually in the following:

    Tamada and King [17] considered two approaches for regen-

    eration through backextraction into an aqueous phase. These

    involve changes in the equilibrium relationship through a swing

    of temperature and a swing of diluent composition. Yabannavar

    and Wang [23,33] suggested two methods for recovery of lactic

    acid from a loaded solvent phase: using NaOH and using HCl.

    2.6.1. Using NaOH

    In the first recovery method, Yabannavar and Wang [23] sug-gested the backextraction of lactic acid from a loaded organic

    phase (lactic acid + Alamine 336 + oleyl alcohol) with small vol-

    ume of a sodium hydroxide solution [1:10 (v/v) NaOH-solvent].

    NaOH in excess of stoichiometric amounts can be used to

    ensure complete lactic acid recovery. Yabannavar and Wang [23]

    obtained 100% recovery of lactate. The resultant high product

    concentration is certainly desirable from the point of economic

    product recovery.

    However, the acid is then recovered as sodium lactate. One

    must add an appropriate acid (e.g., sulfuric acid) to return it to

    the free acid form. This approach has the same drawbacks as the

    classical calcium precipitation process for direct recovery from

    the aqueous feed. Both sulfuric acid and NaOH are consumed,

    and a waste salt is formed, which requires disposal. The degree

    of reextraction of lactate with 0.1 N NaOH is high (90%). The

    degree of reextraction with water amounts to only 55%.

    2.6.2. Using HCl

    In the another recovery method suggested by Yabannavar and

    Wang [33], concentrated HCl is used to essentially displace the

    lactic acid from the loaded organic phase (lactic acid + Alamine

    336 + oleyl alcohol). More than stoichiometric amounts of HCl

    were necessary to recover most of the product from the solvent.

    The lactic acid recovered through backextraction with HCl is in

    the undissociated form. It is possible to regenerate the solvent bydistilling diluent in thedispersed organic phase. This methodhas

    the drawbacks that aqueous HCl is highly corrosive and requires

    special material of construction (glass lined/graphite).

    Yabannavar and Wang [33], recommended the reextraction

    of the free acid from Alamine 336/oleyl alcohol solution with

    cone. HCI. The obtained yield was fairly high (83%), however,

    the emulsion formed during the reextraction was very stable.

    2.6.3. Temperature-swing regeneration

    In a temperature-swing extraction/regeneration scheme [46],

    the extraction is carried out at relatively low temperature, pro-

    ducing an acid-loaded organic extract and an aqueous raffinatewaste stream containing the unwanted feed components. Dur-

    ing regeneration, the extract is contacted with a fresh aqueous

  • 7/28/2019 lactic_rev_SEPPUR8558.pdf

    8/18

    Autho

    r's

    pe

    rsonal

    copy

    H.G. Joglekar et al. / Separation and Purification Technology 52 (2006) 117 7

    stream at a higher temperature to produce an acid-laden aqueous

    product stream and an acid-free organic phase. The concentra-

    tion of the acid achievable in this stream depends on the amount

    of change in the extraction equilibrium between temperatures

    and can be higher than that in the original aqueous feed stream

    [46].

    At low acid concentrations, a strong diluent effect can be

    recognized. The Kd value is a function of the polarity of the

    diluent and its ability to form hydrogen bonds. At higher acid

    concentrations, however, this diluent effect diminishes consid-

    erably. For highly complexed amines, the Kd value depends only

    slightly on the acid concentration. The Kd value is considerably

    reduced with increasingtemperature. This behavior ofKd is used

    to reextract the citric acid into the aqueous phase by temperature

    increase. Because of the lower enthalpy change, the extraction

    efficiency of lactic acid by temperature-swing regeneration is

    low [47].

    2.6.4. Using trimethyl amine (TMA)

    Lactic acid extracted in the organic phase can be back

    extracted with a stronger volatile amine like TMA in aqueous

    phase. It avoids consumption of chemicals and creation of salt

    byproduct. The aqueous base, which is volatile, enables thermal

    decomposition of the acidbase complex in the aqueous back-

    extract. The decomposition forms lactic acid as a product and

    freebase as a vapor that can be reabsorbed in water and recycled

    for reuse in backextraction. The most obvious water-soluble,

    volatile base is ammonia. However, ammonia and both primary

    and secondary aminesform amides when they are heated in mix-

    tures with carboxylic acids [4850]. The amides are sufficiently

    stable so that it is difficult to reverse the process and recover the

    amine.

    The theory of extraction accompanied by chemical reaction

    has been used to obtain the kinetics of back extraction of lactic

    acid by aqueous TMA. In thermal regeneration of TMA, 99%

    of TMA was removed at 200 mmHg from trimethylammonium

    lactate aqueous solution.

    An approach for regeneration and product recovery from

    organic extracts (alamine, TPATOA in diluents) is to back

    extract lactic acid with a water soluble, volatile tertiary

    amine such as trimethyl amine. Equilibrium data presented by

    researchers shows near stoichiometric recovery of lactic acid

    from amine extract.The organic phase, which is recycled to the fermentor, may

    contain residual dissolvedTMA,whichcan affect the bioactivity

    of the enzyme.

    2.6.5. Diluent swing

    This is based on a shift of the equilibrium distribution of acid

    from the aqueous phase to the organic phase between forward

    and back extraction caused by a change in composition of the

    diluent with which the extractant is mixed. This diluent compo-

    sition swing facilitates back extraction of acid into an aqueous

    product phase. The composition of the acid laden organic phase

    leaving the extractor is altered, by either removal of the diluentor addition of another diluent, to produce a solvent system that

    promotes distribution of the acid to the aqueous phase.

    This altered organic phase is contacted with a fresh aque-

    ous stream in the regenerator to produce the acid laden aqueous

    product andthe acid-freesolvent for recycle to theextractor [47].

    Adjustment of the diluent composition can also occur before

    this solvent reenters the extractor. This approach involving more

    than one diluent appears to be more complicated than the TMA

    approach, where an easily removable volatile amine (TMA) is

    the only externally introduced component. This process has the

    disadvantage of diluting the extract stream and requiring dis-

    tillation of large amounts of solvent (after the regeneration) to

    obtain the same shift in the active/inert diluent ratio.

    Tamada and King [47], suggested the diluent-swing regen-

    eration process for lactic acid. Baniel et al. [40], also described

    regeneration by back extraction following a changein thediluent

    composition. In the diluent-swing process, extraction is carried

    out in a solvent composed of the amine and a diluent that pro-

    motes distribution of the acid in the organic phase.

    To date, diluent systems have been composed of organic sol-vents that are liquids at room temperature. A drawback of such

    systems is that changes in extractant-phase composition gen-

    erally involve a distillation step to separate the active and inert

    diluents.To avoid this energyexpense, a new process is proposed

    that will replace the inert liquid diluent with a gas anti solvent.

    Here, anti solvent is used to connote a substance that has a low

    capacity to solubilize the extracted acid. In this process, the dilu-

    ent composition change will be effected by pressurizing it with

    a gaseous antisolvent (e.g., propane).

    For carboxylic acids that exist as solids at room temperature,

    this composition change might induce precipitation of the acid

    product from the extractant phase. Alternatively, for acids that

    are not solids at room temperature, this composition change will

    be carried out while the extractant is in contact with an aqueous

    product phase. Here, the antisolvent will shift the equilibrium

    distribution of the acid to favor the aqueous product phase.

    The idea for this process stems from previous work [51] on

    extraction of carboxylic acids with amine extractants that indi-

    cated that the chain length of liquid-phase aliphatic hydrocarbon

    diluents (C7C12) has little effect on extractant loading. This

    result is not surprising, since such diluents provide essentially no

    stabilization of the acidamine complex. It is therefore expected

    that gaseous aliphatic hydrocarbons (e.g., propane) will behave

    similarly as inert diluents. Acid extraction power is expected

    to decrease with increasing concentrations (i.e., higher partialpressures) of propane in the organic extractant phase. Thus,

    a diluent-swing regeneration process will be devised whereby

    pressurizing the system with propane gas will effect changes in

    extractant composition. A benefit of this process over conven-

    tional recovery techniques is that the diluent components can be

    easily separated (e.g., by a flash operation) without involving a

    distillation step.

    2.7. Choice of extraction

    The main parameters for selection of a diluent-extractant

    system for extraction of lactic acid are distribution coefficient,toxicity, complexation constant and feasibility for back extrac-

    tion. From the literature it is noted that Alamine 336 in proper

  • 7/28/2019 lactic_rev_SEPPUR8558.pdf

    9/18

    Autho

    r's

    pe

    rsonal

    copy

    8 H.G. Joglekar et al. / Separation and Purification Technology 52 (2006) 117

    diluent is the best single extractant in terms of distribution coef-

    ficient, toxicity and feasibility for back extraction. The oleyl

    alcohol is a better diluent than octanol. However this system

    forms a third layer, which is undesirable.

    In reactive extraction by using mixed diluents such as 1-

    octanol/n-heptane, the third layer formation is avoided, as well

    as optimum extractability can be obtained by proper ratio of

    short chain amine and long chain amine. Extraction efficiency

    increases with decreasing initial lactic acid concentration in

    aqueous phase. For 5 wt.% lactic acid concentration, the maxi-

    mum extraction efficiency was obtained for TPA/TOA ratio in

    the range of 6:48:2.

    Regeneration of lactic acid-laden amine extractants by addi-

    tion of a gas antisolvent appears to have advantages over more

    traditional diluent swing or temperature swing methods. A

    potential secondary benefit of regeneration by gas antisolvents

    might be the ability to fractionate a mixture of carboxylic acids.

    However this needs to be evaluated.Thus the best results can be obtained by employing reactive

    extraction using a mixture of TPA and TOA as extractant, 1-

    octanol and n-heptane as diluent and gas-antisolvent-induced

    regeneration for reextraction. It is to be used advantageously

    with continuous or semi continuous fermentation. Reactive

    extraction does not totally separate other organic acids from

    lactic acid and further purification is necessary.

    3. Adsorption

    Recovery of carboxylic acids from fermentation broths

    presents a challenging separation problem, because of the dilute,

    complex natureof fermentation broths. Methods of recovery that

    utilize separatingagents, such as solidsorbents, that are selective

    for carboxylic acids are attractive and reported by researchers

    such as King [52], Tung and King [53]. Important characteristics

    of extractants and solid sorbents are a high capacity for the acid,

    a high selectivity for the acid as opposed to water and substrate

    (e.g., glucose), regenerability, and, depending upon the process

    configuration, the biocompatibility with microorganisms. Many

    fermentations, such as that for lactic acid production, are subject

    to end-product inhibition [23,54]. If a solid sorbent can be used

    in situ or in an external recycle loop, higher overall yields can

    be achieved.

    3.1. Reported research

    Many fermentations to produce carboxylicacids operate most

    effectively at pH abovepKa of theacid product. Forexample, lac-

    tic acid (pKa = 3.86) is typicallyproducedat pH 56[55,56].One

    approach for recovering carboxylic acids from such solutions

    is to use agents that are sufficiently basic to retain a substan-

    tial capacity several pH units above the pKa of the carboxylic

    acid. Tung and King [53] investigated extraction and sorption

    of lactic and succinic acids using different basic extractants and

    polymeric sorbents. The results showed that the uptake in the

    pH range 56 varied substantially from one agent to another andwas strongly dependent upon the basicity and capacity of the

    agent. Agents to be used in fermentation processes should pro-

    vide high selectivity between the product carboxylic acid and

    substrate sugars. In addition to conservation of substrate, one

    reason why very high selectivitys are often sought is the ten-

    dency for small amounts of sugars in a product such as lactic

    acid to cause discoloration. Reported measurements of uptake

    capacities for sugars and for selectivity achieved between car-

    boxylic acids and sugars with extractants and/or solid sorbents

    are few and fragmentary.

    Kaufman et al. [57], screened a series of solid sorbents

    preliminarily for possible utilization in biparticle fluidized-bed

    fermentation to produce lactic acid from immobilized Lacto-

    bacillus delbreuckii, measuring selectivity between acid and

    sugar as well as other properties. Kaufman et al. [58], presented

    selectivities between lactic acid and glucose obtained with a

    fixed bed of the sorbent Amberlite IRA-35 under a particu-

    lar set of operating conditions. Evangelista et al. [59], reported

    breakthrough curves for lactic acid and glucose during fixed-bed

    adsorption with actual fermentation broths.For typical lactic acid fermentations, the sorbent should

    demonstrate substantial uptake in the pH 56 range [53,60].

    Ion exchange technique is widely used in bio-separation

    [61,62] and several different ion exchangers such as poly(4-

    vinylpyridine) resin (PVP) [63], IRA-420 [64], IRA-400 [65,66]

    have been studied on the recovery of lactic acid in the past

    few years. However, little systematic research has been done

    on the effect of different operating conditions on the separa-

    tion efficiency of lactic acid and especially on the purification

    performances of lactic acid.

    Karl-Heinz et al. [67], discloses the separation of lactic acid

    from a fermentation medium with an adsorbent comprising a

    polymer with tertiary amino groups as described in US Patent

    4,552,905 [68].

    Kawabata et al. [69], separated carboxylic acid by using a

    polymer adsorbent of pyridine skeletal structure and a cross-

    linked structure.The polymer adsorbentshowed goodselectivity

    and high adsorption capacity for carboxylic acids even in the

    presence of inorganic salts. The selected eluants were aliphatic

    alcohol, aliphatic ketone, and carboxylic ester.

    Evangelista and Nikolov [70], recovered lactic acid from

    fermentation broth by weak base polymer adsorbents MWA-

    1, IRA-35, and VI-15. The pH for the adsorption of lactic acid

    was below its pKa and fermentation broth was acidified by using

    cation exchange resin instead of using inorganic acid to elim-inate possible competition between inorganic acid and lactate

    in the subsequent adsorption steps. Methanol and 5% NH4OH

    were used as elutants. Though 1.5 times of bed volume of 5%

    NH4OH could recover all the adsorbed lactic acid from MWA-1

    column, product purity was not high. However, 6.8 times of bed

    volume of methanol could completely desorb lactic acid from

    VI to 15 anion exchange resin with higher purity.

    Simulated moving bed (SMB) chromatography is a continu-

    ous separation process that has many important industrial appli-

    cations [71]. An SMB consists of a circle of chromatographic

    columns. The circle is typically divided into four zones by two

    inlet ports and two outlet ports. The four ports move periodicallyby one column length along the mobile phase flow direction to

    follow themigrating solute bands.The port movement maintains

  • 7/28/2019 lactic_rev_SEPPUR8558.pdf

    10/18

    Autho

    r's

    pe

    rsonal

    copy

    H.G. Joglekar et al. / Separation and Purification Technology 52 (2006) 117 9

    product purity and yield, while achieving simulated counter-

    current movement between the adsorbent and the fluid [72].

    In comparison to batch chromatography, SMB processes have

    higher yield and product purity as well as lower solvent con-

    sumption. SMB has been used, for instance, for hydrocarbon

    and high fructose corn syrup purification [71]. Recently, SMB

    has been developed for the separation of racemic mixtures or

    isomers [73,74].

    Lee et al. [75], explored the feasibility of developing a SMB

    process for the separation of lactic acid from other organic

    acids present in a fermentation broth. Poly(4-vinylpyridine)

    resin (PVP), which was found to have a high capacity and selec-

    tivity for organic acids in a previous study [76], was chosen as

    the adsorbent. A systematic design method, which is based on

    the standing wave analysis for Langmuir isotherm systems, was

    tested for this separation [74]. Adsorption isotherms and mass

    transfer parameters for the two organic acids were determined

    from frontal chromatograms. These parameters were first vali-dated by comparing the single-column frontal chromatograms

    with simulations based on a rate model [77]. The validated

    parameters were applied in the standing wave design to find the

    four zone flow rates and step time for a laboratory SMB unit.

    High product purity (99.9%) and relatively high yield (>93%)

    were achieved in the SMB experiments. SMB column profiles

    and effluent histories can be better explained if the adsorption

    isotherm of lactic acid is represented by a modified Lang-

    muir equation. The standing wave design method was extended

    to nonlinear systems with modified Langmuir isotherms. Rate

    model simulations show that this method can achieve higher

    purity and higher yield for this separation than the design based

    on the Langmuir isotherms.

    Chen et al. [78], examined polyvinyl pyridine (PVP) and

    activated carbon for the adsorption characteristics pertinent to

    their application in lactic acid fermentation. For PVP the lin-

    ear adsorption constant, Kad, was between 0.7 and 1.0 for an

    equilibrium pH range of 39. The pH was adjusted by acid/base

    addition,similar to pH control in fermentation. Thevalues ofKadin the pH-adjusted systems were much lower than that reported

    for pure lactic acid solutions, i.e., about 9.7. Furthermore, no

    clear effect of pH was observed. These are attributed mainly to

    the competition of anions (Cland lactate) for the adsorption

    sites of protonated pyridinal N. Its adsorption capacity was also

    found to decrease with the base regeneration (by about 14%

    each time) after being contacted with the culture broth. These

    factors limit its potential application in lactic acid fermentation.

    Activated carbon was much more effective in lactic acid/lactate

    adsorption than PVP. The adsorption further favored lower pH

    under acid (HCI) addition. Activated carbon has been reported

    to adsorb glucose. However, the presence of glucose in 070 g/l

    was found in this study to have an insignificant effect on lac-

    tate adsorption. Cells ofL. delbrueckii also adsorbed rapidly on

    activated carbon. This cell adsorption had a negative effect on

    lactate adsorption.

    Kulprathipanja et al. [79], studied the separation of lactic acid

    from a fermentation broth by using an adsorbent comprising

    a water-insoluble macro-reticular gel or weakly basic anionic-

    exchange resin possessing tertiary amine or pyridine functional

    groups or a strongly basic anionic exchange resin possessing

    quaternary functional groups. The resins are in sulphate form

    and have a cross-linked acrylic or styrene resin matrix. Theadsorption mechanism is shown Fig. 1. For tertiary amine and

    pyridine-function-containing ion exchange resins, the lone elec-

    tron pair of the nitrogen atom enables nitrogen atom to form

    hydrogen bond with sulfate ion. IRA-400, strongly basic qua-

    ternary ammonium ion exchange resin has positive charge and

    can form ionic bond with sulfate ion. The sulfate form of qua-

    ternary ammonium of anion exchange resin has a weakly basic

    property and can adsorb lactic acid through acidbase interac-

    tion. Consequently, the adsorption of lactic acid is not affected

    by inorganic salt in fermentation broth. Lactic acid is desorbed

    with water or dilute inorganic acid. The pH of the feed is main-

    tained below the ionization constant (pKa

    ) of the lactic acid to

    obtain high selectivity.

    Several different ion exchangers such as IRA-400, PVP,

    and IRA-92, Amberlite were tested. IRA-92, a weakly basic

    exchanger, is selected as the ion exchanger for the purification

    of lactic acid because of its fair purity and recovery yield of

    lactic acid comparable with other resins results.

    Tong et al. [80], examined the capacity of IRA-92 to lactic

    acid in the fermentation broth, and then, optimized the operating

    conditions, and finally gained a fair product of lactic acid with a

    higher purity, recovery yield and productivity with only one-step

    chromatographic procedure from the supernatant of the fermen-

    tation broth. The experimental results demonstrate that when the

    Fig. 1. The mechanism of adsorption of lactic acid on the anion exchange resin.

  • 7/28/2019 lactic_rev_SEPPUR8558.pdf

    11/18

    Autho

    r's

    pe

    rsonal

    copy

    10 H.G. Joglekar et al. / Separation and Purification Technology 52 (2006) 117

    pH of the fermentation broth is 6.0, the recovery yield, purity

    and productivity in the lactic acid purification are the highest.

    With the decrease of the sample volume loaded and of flow rate,

    the recovery yield and the purity are improved but the productiv-

    ity apparently reduced. In the experimental scale, the scale-up

    of purification process exhibits little influence on the recovery

    yield and the purity. Under the preliminary optimal conditions,

    the yield, purity and specific productivity are 82.6%, 96.2% and

    1.16 g LA/(g-resin day), respectively.

    Raya-Tonetti et al. [81], studied the use of strong anionic

    exchange resin (IRA-400) to recover lactic acid directly from

    fermentation in an upflow fluidized bed column, resulting in

    0.18 g lactic acid/g resin bound with a subsequent elution of

    94%. When the culture broth was heated and adjusted pH to 8.0,

    0.4 g lactic acid was bound per gram of resin, with a subsequent

    elution of 90%. l (+) and d () lactic acid isomers distribution

    was analyzed in the elution product resulting in an increase of

    l (+) isomer concentration. The resin did not alter its bindingcapacity even after 23 cycles. The lactic acid recovery attained

    was not dependent on the increase of the resin poured to the

    columns but due to the increase in the hydraulic residence time

    and back mixing which resulted in a high contact time between

    lactate ions and resin particles.

    Srivastava et al. [64], separated lactic acid by using IRA-

    400 column coupled with fermentor. This study was focused on

    improving fermentation yield, and the separation performance

    of IRA-400 was not studied. The Amberlite IRA-400 resin has

    proper pore size and high adsorption capacity for recovery of

    lactic acid and it can adsorb lactic acid in wide pH range.

    Cao, et al. [65], studied the application of Amberlite IRA-

    400 anion exchange resin for the recovery ofl-(+)-lactic acid

    from fermentation broth. Adsorption isotherm and breakthrough

    curves for the separation of (l+)-lactic acid were obtained at pH

    5.0 and 2.0, respectively. Recovery experiment coupled with

    fermentor was carried out successfully by using a column with-

    out autoclaving. Different types of adsorption isotherms were

    found at pH above and below the pKa (3.86) of lactic acid. The

    isotherm was found to be a Langmuir type at pH 5.0, whereas the

    isotherm was type II (multilayer adsorption) at pH 2.0. At pH 5.0,

    the maximum adsorption capacity of the resin, and dissociation

    constant, were 222.46 mg/g wet resin and 60.7 mg/ml, respec-

    tively. Breakthrough curve for the separation of lactic acid from

    fermentation broth was also obtained. The maximum adsorptioncapacity (197.09 mg/g wet resin) at pH 5.0 was much higher

    than that at pH 2.0 (106 mg/g wet resin). Proper elution and

    washing conditions were sought by using H2SO4, methanol,

    ammonia or their mixtures as eluant. When column separa-

    tion was performed at pH 5.0 by using 50% (v/v) methanol

    as washing solvent and 1.0 M H2SO4 as eluant, the total yield

    was 86.21%. However, the total yield was 92.11% when the col-

    umn separation was performed at pH 2.0 and water was used as

    eluant.

    IRA-400 was successfully applied for the separation of (l+)-

    lactic acid from fermentation broth at pH above and below the

    pKa (3.86). It shows different adsorption mechanisms in the sep-aration of lactic acid at pH above and below pKa. 1.0 M H2SO4could be used for the elution of lactic acid at pH 5.0 with high

    recovery. Product loss in column washing step could be reduced

    using methanol. The process is simple and feasible in lactic acid

    industry and could be coupled with fermentation.

    Salts ions in fermentation broth and inorganic acids used for

    adjustment of pH did not affect the adsorption performance of

    resin for lactic acid at pH 2.0. The elution of the IRA-400 anion

    exchange resin could be easily performed using water as eluant.

    Ionic exchange resins have been utilized by Vaccari, et al.

    [82], Senthuran et al. [83] and Kaufman [84], obtaining 0.1, 0.18

    and 0.2 g lactic acid/g resin, respectively. All purification tech-

    niques mentioned earlier, need prior cell removal by filtration

    or by centrifugation. Those steps mean increase in equipment

    and energy costs, increase in total process time, and reduction

    in total yield.

    Lactic acid was recovered and purified by using a fluidized

    bed column refilled with a strong anionic exchange resin:

    Amberlite IRA-400 achieving 0.126 g LA/g resin [85]. This

    methodology is more useful considering the fact that only con-trolling the ascensional velocity is sufficient to washout the

    biological solids and to bind the lactate to the resin in only

    one step. A very important advantage that this technique counts

    with is the use of simple equipment, such as a glass column,

    an anionic exchange resin, a peristaltic pump, and commercial

    grade reagents, to carry out the early recovery process of lactic

    acid from the culture [86].

    Sosa, et al. [87], studied the influence of two design parame-

    ters, bed-diameter (D) and bed-height (H), on lactic acid binding

    capacity utilizing fluidized bed columns refilled with anionic

    exchange resin. The binding capacity of the resin was constant

    during the lactic acid downstream process when the columns

    were scaled-up on diameter, keeping constant the bed height.

    Substantial increase (50%) in lactic acid adsorption to Amberlite

    IRA-400 was achieved in the case of scaling up the H parameter

    maintaining the diameter of the columns constant.

    The same ratio of grams of lactic acid to grams of resin was

    fed to both columns in all experiments. Thus, the lactic acid

    recovery attained was not dependent on the increase of the resin

    poured to the columns but due to the increase in the hydraulic

    residence time and back mixing which resulted in a high contact

    time between lactate ions and resin particles. By changing the

    settled bed height from 2.5 to 5cm for each diameter of column

    analyzed, it was possible to obtain 50% increase in the binding

    capacity of the resin in all experiments. This fact was attributedto a higher contact time between theculture broth andthe anionic

    resin produced by the increase in back mixing and lactic acid

    residence time.

    There is a further need to study the effect of other parameters

    on the performance of lactic acid purification process utilizing

    fluidized bed columns, taking into account the developing of this

    procedure on larger scale.

    The uptake data for lactic acid with all three sorbents (Dowex

    MWA-1, Reillex 425, Amberlite IRA-35) are well described by

    1:1 complexation [88], which gives a functional form equivalent

    to the Langmuir equation.

    Qca

    Qm=

    KCa,f

    (1+KCa,f), (4)

  • 7/28/2019 lactic_rev_SEPPUR8558.pdf

    12/18

    Autho

    r's

    pe

    rsonal

    copy

    H.G. Joglekar et al. / Separation and Purification Technology 52 (2006) 117 11

    where Qca is thecomposite uptake of lacticacid (g of lacticacid/g

    of sorbent); Qm the maximum value of sorbed lactic acid at high

    concentration (g of lactic acid/g of sorbent); Ca,f the weight

    fraction of lactic acid in the final solution; K the complexation

    constant for sorption (g of resin/g of lactic acid).

    Daietal. [89], has found that the uptakeof lacticacidfor sorp-

    tion by Dowex MWA-1 is much higher than that for extraction

    by Alamine 336. This advantage is more significant in the higher

    pH range. On the other hand, the uptake capacity for water is

    much higher for sorption than for extraction, and this results in a

    substantially higher uptake of glucose onto the sorbent. Marked

    swelling of the sorbent in the solution reduces the selectivities

    for acid over water and glucose.

    Sun et al. [90] demonstrated that the in situ product removal

    process using the OH-form anion-exchange resin can be suc-

    cessfully carried out for lactic acid fermentation. Both in the

    shaking flasks and in the airlift bioreactor (ALB), the broth pH

    was controlled at values higher than 6, and over 90% of the lacticacid produced by the bioconversion was recovered by the anion

    exchanger. In addition, the mathematical model developed for

    the in situ product removal (ISPR) process agreed well with the

    experimental results.

    3.2. Choice of adsorption

    Adsorption or theion exchangeis a reliable technology. How-

    ever, the process requires regeneration of an ion exchange resin

    and adjustment of the feed pH to increase the sorption efficiency,

    requiring large amounts of chemicals and needs to treat and dis-

    pose of large quantities of salts/effluents.

    Adsorption on ion exchange resin has the advantage that it

    can be coupled with fermentation process. For lactic acid fer-

    mentations, the sorbent should demonstrate substantial uptake

    in the pH range of 56.

    The uptake of lactic acid for sorption by Dowex MWA-

    1 is much higher than that for extraction by Alamine 336.

    This advantage is more significant in the higher pH range.

    The capacity of IRA-92 to sorb lactic acid in the fermenta-

    tion broth, at optimized operation conditions, is high with a

    higher purity and recovery yield with only one-step chromato-

    graphic procedure from the supernatant of the fermentation

    broth.

    The isotherm was found to be a Langmuir type at pH 5.0,whereas the isotherm was type II (multilayer adsorption) at pH

    2.0. At pH 5.0, the maximum adsorption capacity of the resin,

    qm and dissociation constant, Kd were 222.46 mg/g wet resin

    and 60.7 mg/ml, respectively for recovery of l (+)-lactic acid

    from fermentation broth using Amberlite IRA-400. The maxi-

    mum adsorption capacity (197.09mg/g wet resin) at pH 5.0 was

    much higher than that at pH 2.0 (106 mg/g wet resin). Proper

    elution and washing conditions were sought by using H2SO4,

    methanol, ammonia or their mixtures as eluant. When column

    separation was performedat pH 5.0by using 50%(v/v) methanol

    as washing solvent and 1.0 M H2SO4 as eluant, the total yield

    was 86.21%. However, the total yield was 92.11% when the col-umn separation was performed at pH 2.0 and water was used as

    eluant.

    The Amberlite IRA-400 resin has proper pore size and high

    adsorption capacity for recovery of lactic acid and it can adsorb

    lactic acid in wide pH range.

    To recover lactic acid from fermentation broth the most suit-

    able adsorbent is found to be Amberlite IRA-400 resin with

    methanol and 1.0 M H2SO4 as eluant. It has high adsorption

    capacity to adsorb lactic acid in wide pH range. The elution

    by methanol and 1.0 M H2SO4 is also advantageous to directly

    esterify the recovered lactic acid in methanol. Thereby avoiding

    the addition of alcohol and acid catalyst during esterification

    to purify the lactic acid. Further purification by esterification

    is necessary after adsorption, as it does not remove all other

    organic acids.

    4. Electrodialysis

    Electrodialysis is one of very promising and perspective

    methods provided by rapid development of the membrane pro-cesses, especially the membranes in the 80s and 90s [91].

    Electrodialysis fermentation EDF is also studied by several

    authors [9295]. Electrodialysis fermentation (EDF) is promis-

    ing because it can remove product lactic acid continuously from

    the fermentation and maintain the pH of the broth.

    Most of the literature [9599], shows the feasible produc-

    tion of lactic acid from lactate salts in two steps: conven-

    tional electrodialysis (ED) for concentration and purification

    and bipolar electrodialysis for conversion of lactate salts into

    lactic acids with recovery of alkali. Bipolar membrane elec-

    trodialysis also referred to as water splitting electrodialysis,

    can convert aqueous salt solutions into acids. A water split-

    ting stack is similar to a conventional (mono-polar) electro-

    dialysis stack but incorporates a third type of membrane, the

    bi-polar membrane, which is composed of a cation and anion

    membrane layers laminated together. To meet the feed require-

    ments of EDBM, clarification of the fermentation broth is first

    performed by microfiltration. After the conversion of sodium

    lactate into lactic acid, the product is purified by ion exchange

    resins.

    4.1. Reported research

    Hongo et al. [91], proposed to use electrodialysis for in

    situ lactate recovery to reduce the product inhibition. The mainobservation was fouling of anion exchange membranes by cells

    in the electrodialysis fermentation(EDF). To solve this problem,

    Nomura et al. [100], used immobilized growing cells entrapped

    in calcium alginate. Theamount of lactic acid produced by semi-

    continuous electrodialysis fermentation using immobilized cells

    was eight-times higher than that produced by non-pH controlled

    fermentation, but some free cells were found in the solution.

    Czytko et al. [93], found that the electrodialysis unit could only

    be operated with cell free solutions in order to prevent deposi-

    tion of bacteria on the membranes. To increase the productivity

    of the lactic acid fermentation and to reduce the amounts of

    effluents, Boyaval et al. [94] combined the cell recycle systemwith electrodialysis and obtained a lactic acid concentration of

    85 g l h and a productivity of 22g l h. Siebold et al. [95], showed

  • 7/28/2019 lactic_rev_SEPPUR8558.pdf

    13/18

    Autho

    r's

    pe

    rsonal

    copy

    12 H.G. Joglekar et al. / Separation and Purification Technology 52 (2006) 117

    that lactic acid production with in situ electrodialysis without

    ultrafiltratioon is also possible.

    The two-stage electrodialysis method was described by

    Glassner et al. [101]. Madzingaidzo et al. [96], presented a pro-

    cess for sodium lactate purification by mono-polar electrodialy-

    sis and recoveryof lactic acidby bi-polar electrodialysis. Sodium

    lactate of feed concentration 125 g/l was concentrated by mono-

    polar electrodialysis to a maximum of 150 g/l. A low incidence

    of impurities was observed in the concentrated solutions with

    less than 2 g/l glucose and 1.5g/l acetate detected. Bipolar elec-

    trodialysis unit yielded 160 g/l free lactic acid while color and

    otherchemicalimpuritieswere significantly reduced. The lactate

    solution is passed through the bipolar electrodialysis unit to get

    free lactic acid. Additional bleaching and de-ionization process

    steps are followed to polish the free lactic acid for high-grade

    applications.

    Rathin Datta et al. [97], studied the various advantages of

    bipolar electrodialysis to eliminate the salt or gypsum wasteproduced in the conventional processes in large scale lactic acid

    production processes.

    M. Bailly, et al. [102] suggested a complete production

    scheme, targeted around membrane operations for clarification,

    concentration and conversion. The fermentation broth is first

    clarified by cross flow microfiltration. Divalent cations, that con-

    stitute a poison in the configuration used, are then removed from

    the clarified broth. Conventional electrodialysis (ED) is further

    used prior to EDBM in order to increase the concentration of

    ionic species, comprising the organic acid salt. In this manner,

    the area of bipolar membrane required for the conversion is low-

    ered so that the economical profitability of the EDBM step is

    improved.

    As described by Mathieu Bailly [98], electrodialysis with

    bipolar membranes (EDBM) is used to prepare polymer grade

    lactic acid. The process has to be simultaneously economically

    attractive and have a low impact on the environment. In this con-

    text, this article details the performance and cost related to an

    industrial EDBM plant operated for the production of organic

    acids. Over the past 15 years, not more than 12 commercial

    plants totaling about 2400 m2 of Aqualytics and Tokuyamas

    bipolar membranes have been installed throughout the world

    for different chemicals. A complete downstream (from the fer-

    mentation) process based on combined electrodialysis steps for

    purification/concentrationand conversion is proposed in the caseof lactic acid.

    Finally Bailly describes performances of a plant with a pro-

    duction capacity 5000 tonnesof lactic acid peryear which makes

    use of a two-stage electrodialysis process. An electricity con-

    sumption of 1.8 kW/kg of produced acid is reported.

    Habova et al. [99], confirmed that the two-stage electrodial-

    ysis is a suitable and efficient technique for recovery of lactate

    ions from the pretreated fermentation broth and subsequent

    conversion into lactic acid. The pretreament consisted of ultra-

    filtration, decolourisation and removing of multivalent metal

    ions.

    Several authors have carried out studies to improve the eletro-dialysis fermentation (EDF) method. [103105]. But commer-

    cialization of EDF process has not been reported.

    Kang, et al. [106] described nanofiltration (NF) as an alter-

    native to desalting electrodialysis (ED) and ion exchange for the

    recovery of ammonium lactate from fermentation broth.

    Li et al. [107], developed a laboratory scale bioreactor com-

    bining conventional electrodialysis and bipolar membrane elec-

    trodialysis for in situ product removal and pH control in lactic

    acid fermentation. The electrokinetic process enabled removal

    of the biocatalytic product (lactic acid) directly from the biore-

    actor system, in a concentrated form, as well as enabling good

    pH control without generation of troublesome salts.

    4.2. Choice of electrodialysis

    Electrodialysis has been reported for sodium or ammo-

    nium lactate broth. Ammonium lactate is preferred as it uses a

    cheaper alkali. An electrodialysis fermentation(EDF) is promis-

    ing because it can remove produced lactic acid continuously

    from the fermentation and maintain the pH of the broth [103].Commercial exploitation of electrodialysis with bipolar

    membranes (EDBM) is reported by Mathieu Bailly [98]. Broth

    needs to be micro filtered prior to electrodialysis. Monopolar

    electrodialysis is used to concentrate and purify the broth. How-

    ever the output is in the form of ammonium lactate. Further

    bipolar electrodialyis can convert ammonium lactate to lactic

    acid. Recovered ammonium hydroxide can be reused in fer-

    mentation. Addition of ammonium hydroxide in fermentation

    is advantageous as it improves productivity of microorganisms.

    Thus downstream process route will involve microfiltration,

    monopolar electrodialysis and bipolar electrodialysis. However

    this process also does not separate out other organic acids from

    lactic acid.

    5. Esterification and reactive distillation

    High purity lactic acid can be prepared by esterification of

    crude lactic acid with alcohols, distillation of ester, hydroly-

    sis of the distillated lactate ester to yield the alcohol and lactic

    acid. Esterification is the only downstream process, which sep-

    arates other organic acids from lactic acid. Esterification gives

    esters of lactic acid and further hydrolysis of esters is neces-

    sary to get the product as lactic acid. Simultaneous distillation

    with esterificationhydrolysis is called reactive distillation. Fer-

    mented broth containing lactic acid needs to be pretreated toremove some impurities before reactive distillation using resin

    beds as catalyst as well as bed for distillation column.

    5.1. Reported research

    Smith and Claborn [108], described methods for the prepara-

    tion of purified lactic esters from crude lactic acid obtained from

    fermentation. To obtain high yields of the pure lactic esters with

    alcohol containing less than four carbons, it is necessary to use

    a large excess of alcohol during esterification and then rapidly

    remove water with excess alcohol from the ester at low temper-

    ature, preferably in vacuum with aid of an efficient fractionatingcolumn. Calcium lactate is dissolved in methanol and equiva-

    lent amount of sulphuric acid is added to liberate the lactic acid

  • 7/28/2019 lactic_rev_SEPPUR8558.pdf

    14/18

    Autho

    r's

    pe

    rsonal

    copy

    H.G. Joglekar et al. / Separation and Purification Technology 52 (2006) 117 13

    to precipitate and separate calcium sulphate. Lactic acid solu-

    tion is heated for 48 h at refluxing temperature to complete the

    esterification. Similarly method for preparation of lactic esters

    insoluble in water was also presented.

    Filachione and Fisher [109], described a method for purifi-

    cation of lactic acid by preparation of methyl lactate from crude

    aqueous lactic acid. The method comprises of passing methanol

    vapor through aqueous lactic acid and condensing the efflu-

    ent vapors. The condensate, a mixture of methanol, water and

    methyl lactate, can be distilled to recover the methyl lactate and

    hydrolyzed to obtain purified lactic acid.

    Schopmeyer et al. [110], suggested heating an esterification

    mixture of lactic acid (4060 wt.%), a relatively low boiling

    aliphatic alcohol, and an acidic catalyst (sulphuricacid) for ester-

    ifiation in a jacketed kettle. The vapors from the esterification

    kettle are continuously fed to a fractionating column, which may

    contain alcohol, ester and water. Continuous hydrolysis of ester

    in a fractionating column is done at atmospheric pressure. Alco-hol produced by hydrolysis is removed fromthe top and returned

    to esterification kettle, and lactic acid along with water is drawn

    from the bottom.

    Recently reactive distillation has drawn considerable atten-

    tion because of its striking advantages, especially for

    equilibrium-limited reactions. Purification of lactic acid through

    reactive batch distillation was investigated by several investi-

    gators [111113]. The experimental setup used by the authors

    consisted of two columns for separation of reactants from the

    product and two reboilers for esterification reaction and hydrol-

    ysis reaction. The feed consisted of only 1020% lactic acid by

    weight.

    Choi et al. [114], studied batch distillation of lactic acid with

    simultaneous reactions. Lactic acid was reacted with methanol,

    and methyl lactate was produced by the esterification reaction in

    presence of cation exchange resin (Dowex 50W). The volatile

    methyl lactate was distilledsimultaneouslywith hydrolysisreac-

    tion forming lactic acid. To recover pure lactic acid through two

    reactions and distillation, batch distillation system consisting of

    two condensers, feed vessel, and reboiler was used. The yield of

    recovered lactic acid was as high as 95%. When impure lactic

    acid solution obtained from bacterial fermentation was used as

    feed, highly pure lactic solution was obtained with yield of 92%

    in reboiler.

    Seo et al. [111], studied the feasibility of recovery of lacticacid by two batch reactive distillations using cation exchange

    resin (Dowex 50 W) as a catalyst and glass packed column.

    For the recovery of lactic acid, two reactions, esterification and

    hydrolysis, are involved and hence, an apparatus with two dis-

    tillation columns was developed and operated in a batch mode

    to ensure enough residence time in the reboiler and column. The

    effects of operating variables such as catalyst loading, mole ratio

    of lactic acid to methanol, feed concentration, type of alcohols

    and partial condenser temperature on yield were studied. The

    reaction products of the esterification (methyl lactate and water)

    were distilled to the hydrolysis part to recover into pure lactic

    acid. The yield of lactic acid increased as catalyst loading in theesterification part was increased. The decrease in mole ratio of

    lactic acid to methanol, and lactic acid feed also improved the

    yield of lactic acid. Methanol as a reactant gave higher yield

    than any other alcohol. The yield of lactic acid was as high as

    90%.

    Kim et al. [112], also carried out similar studies using Old-

    ershaw columns and reboilers for fractionation and reactions.

    Concentration and temperature profiles in the reboilers and on

    each stage were investigated throughout the operation. Six hour

    operating time was required for getting high purity lactic acid.

    The effect of columns on the recovery yield was investigated.

    The column improved the fractionation of the boilups from the

    reboilers. More effective fractionation in columns allowed vapor

    streamin columns to contain more methanol and liquid streamto

    contain less methanol. Thereby, methanol-concentrated recycle

    flow was obtained more effectively and methanol was prevented

    to remain in hydrolysis part. That resulted in the more effective

    reaction in both reboilers and improved the yield.

    Kim et al. [113], studied reactive batch distillation similar to

    Kim et al. [112]. Methanol recycle and feeding method wereinvestigated as the factors, which could control the component

    boilup rate of each species and rate of esterification reaction.

    The temperature of partial condenser controlled the flow rate

    and composition of methanol recycle stream. Semi batch oper-

    ation was compared with batch operation. Continuous feeding

    of methanol enhanced the recovery system performance while

    continuous feeding of lactic acid aqueous solution, deteriorated

    the recovery compared with batch operation.

    In continuous esterification, the mixture of an aqueous solu-

    tion of lactic acid to be purified, a relatively low boiling aliphatic

    alcohol, and an acidic catalyst of esterification is heated to pro-

    duce mixture of vapors. Use of sulphuric acid as catalyst may

    result in traces of acid in the product. In presence of impurities

    with cation exchange resin catalyst it may be difficult to main-

    tain the operation at steady state. Therefore it may be necessary

    to purify the crude lactic acid from impurities like residual sugar

    and protein before hand.

    The two column reactive distillation scheme studied by Seo

    et al. [111], Kim et al. [112], Kim et al. [113], is for upto 30%

    crude lactic acid feed. And the catalyst used for the studies for

    both esterification and hydrolysis is cation exchange resin i.e.

    DOWEX 50W. This scheme with feed of 5060% lactic acid

    concentration may reduce the size of equipment.

    The reported yield of the ester vary from about 60 to 100%,

    depending largely upon the method used for removing the wateras it is formed during the esterification reaction. It is more dif-

    ficult to obtain a high yield of ester with methyl or isopropyl

    alcohol than with a long chain alcohol such as butyl, amyl,

    acetyl, lauryl, and stearyl, because the lower alcohols and the

    corresponding lactic esters are soluble in water. It is necessary,

    therefore, to resort to other methods for removing this water or

    reducing the ratio of water to alcohol in the reacting mixture by

    using a large excess of alcohol.

    Recently, Sun at al.[115], studied conversion of LA or ammo-

    nium lactate (NH4LA) into esters and subsequent hydrolysis

    of the purified ester into LA to obtain highly pure LA. In this

    study, using two reactors with a rectifying column were used torecover LA from the fermentation broth. NH4LA obtained by

    fermentation was used directly to produce butyl lactate by react-

  • 7/28/2019 lactic_rev_SEPPUR8558.pdf

    15/18

    Autho

    r