WO2001087452A2 - Conception d'ondes stationnaires de lits mobiles simules uniques ou tandem destinee a la resolution de melanges a composants multiples - Google Patents

Conception d'ondes stationnaires de lits mobiles simules uniques ou tandem destinee a la resolution de melanges a composants multiples Download PDF

Info

Publication number
WO2001087452A2
WO2001087452A2 PCT/US2001/015848 US0115848W WO0187452A2 WO 2001087452 A2 WO2001087452 A2 WO 2001087452A2 US 0115848 W US0115848 W US 0115848W WO 0187452 A2 WO0187452 A2 WO 0187452A2
Authority
WO
WIPO (PCT)
Prior art keywords
component
zone
wave
fraction
smb
Prior art date
Application number
PCT/US2001/015848
Other languages
English (en)
Other versions
WO2001087452A3 (fr
Inventor
Nien-Hwa Linda Wang
Yi Xie
Sungyong Mun
Jin-Hyun Kim
Benjamin J. Hritzko
Original Assignee
Purdue Research Foundation
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Purdue Research Foundation filed Critical Purdue Research Foundation
Priority to AU2001263192A priority Critical patent/AU2001263192A1/en
Publication of WO2001087452A2 publication Critical patent/WO2001087452A2/fr
Publication of WO2001087452A3 publication Critical patent/WO2001087452A3/fr

Links

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B01—PHYSICAL OR CHEMICAL PROCESSES OR APPARATUS IN GENERAL
    • B01D—SEPARATION
    • B01D15/00—Separating processes involving the treatment of liquids with solid sorbents; Apparatus therefor
    • B01D15/08—Selective adsorption, e.g. chromatography
    • B01D15/10—Selective adsorption, e.g. chromatography characterised by constructional or operational features
    • B01D15/18—Selective adsorption, e.g. chromatography characterised by constructional or operational features relating to flow patterns
    • B01D15/1814—Selective adsorption, e.g. chromatography characterised by constructional or operational features relating to flow patterns recycling of the fraction to be distributed
    • B01D15/1821—Simulated moving beds
    • B01D15/1828—Simulated moving beds characterized by process features
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B01—PHYSICAL OR CHEMICAL PROCESSES OR APPARATUS IN GENERAL
    • B01D—SEPARATION
    • B01D2215/00—Separating processes involving the treatment of liquids with adsorbents
    • B01D2215/02—Separating processes involving the treatment of liquids with adsorbents with moving adsorbents
    • B01D2215/023—Simulated moving beds

Definitions

  • This invention relates to separations of one or more components from multi- component fluid mixtures using simulated moving bed technology, and to methods of designing such simulated moving beds (SMBs).
  • SMBs designed according to this invention can be used to separate slow and fast moving fractions in multicomponent mixtures, or to separate an intermediate affinity component from the mixture.
  • the invention is applicable to systems exhibiting linear isotherms, and is especially applicable to such systems when separations are inhibited by mass transfer resistances.
  • Liquid chromatography is widely used in industry for separating and purifying components in liquid and gaseous mixtures.
  • most applications of liquid chromatography are batch chromatographic processes, which give low yields, consume high quantities of eluent, utilize adsorbent inefficiently, and are labor intensive.
  • the discontinuous nature of batch chromatography, and the dilution of collected components limits its attractiveness for pilot-plant and process-scale separations.
  • CMBs continuous moving beds
  • the solid phase which moves in a downward direction driven by gravity, exits the column at the bottom and is recycled to the top.
  • the mobile phase enters the column at the bottom, is pumped upward, and exits at the top from which point it is recycled to the bottom of the column.
  • This system has two liquid inlets (a feed inlet and a desorbent inlet) and two liquid outlets (a raffinate outlet and an extract outlet).
  • the feed (containing 2 or more components to be separated) is injected continuously into the middle of the column.
  • This system has two outlet lines: one located beneath the feed where the slower moving component is continuously removed (the extract), and one located above the feed where the faster moving component is removed (the raffinate).
  • Desorbent is added to the system at the bottom to help push the slower component back up into the column (and to make up for desorbent withdrawn through the extract and raffinate streams).
  • Continuous moving bed systems offer a number of potential advantages over batch chromatography, including:
  • the systems can produce products of purity similar to or higher than batch chromatography, at yields substantially greater than batch chromatography.
  • Simulated moving beds are reviewed generally in Ruthven and Ching (1989).
  • Simulated moving bed (SMB) processes realize the counter- cunent movement of solid and liquid phases and the concomitant advantages of continuous moving beds over batch chromatography without the physical movement of solids.
  • a typical SMB process has four zones with two inlet ports (feed and desorbent) and two outlet ports (raffinate and extract), as shown in Figure 1.
  • SMB processes utilize a series of adsorbent columns connected to form a circuit. Each zone typically contains two or more evenly sized columns, and these columns are connected to form a continuous circuit.
  • the solid movement is simulated by periodically moving the inlet and outlet ports one column forward in the direction of flow of the mobile phase, so that the product ports are always near the partially separated concentration waves of products in the system. Similar to the continuous moving bed system, the port switching time, zone length, and zone flow rates are all balanced to attain a desired level of purity of the raffinate or the extract.
  • Zone I This zone is located between the desorbent inlet and extract withdrawal and is used for desorbing the slow fraction. A portion of the stream leaving this zone is withdrawn as the extract while the remainder flows into zone II.
  • Zone II This zone is used for desorbing the fast fraction.
  • Zone HI This zone is used for adsorbing the slow fraction, and separating the slow fraction from the fast fraction. The fast fraction is partially withdrawn at the raffinate port.
  • Zone IV This zone is used for adsorbing the fast fraction as well as desorbent recovery.
  • the stream leaving this zone should ideally contain only pure desorbent which is recirculated to the desorbent inlet for reuse.
  • UOP Universal Oil Products
  • SMB separations were first applied in an industrial plant by Universal Oil Products (UOP) for the purification of hydrocarbons (Broughton and Gerhold, 1961; Broughton, 1968; Broughton et al., 1970).
  • UOP's process eventually evolved into what it now calls its Sorbex process.
  • Sorbex process The principal application for UOP's Sorbex process is in the purification of xylene isomers. Indeed, Parkinson et al. (1994) reports that by 1994, UOP had installed 85 SMB units for use of its Sorbex process, and that most of these had been installed at petroleum refineries for p-xylene purification.
  • UOP's Sorbex process for the purification of p-xylenes has several unique features. First, it separates p-xylene from a mixture of xylene isomers and other phenyl derivatives from the xylene manufacturing process. The process uses a zeolitic solid adsorbent that selectively adsorbs p-xylene. The process also relies upon a displacing component which competes for adsorbent with p-xylene and the other components in a concentration dependent fashion, to prevent cross-contamination of the raffinate and extract.
  • Parkinson et al. (1994) also reports widespread use of SMB processes for the separation of sugars. For example, Parkinson et al. reports that from 1980 to 1994, U.S. Filter had installed approximately 60 SMB units for the separation of sugars. In 1994, U.S. Filter had piloted SMB systems for the separation of lactic and citric acids, glycerine, glycols, amino acids, and herbicides. In 1994, UOP reportedly installed HPLC SMB in pharmaceutical plants.
  • Soken In the same year, Soken (Japan) reportedly finished a one-year pilot test of a continuous HPLC system that can purify three-component mixtures and has six columns.
  • Other applications for SMBs have been reported in the literature (Ching et al., 1993; Pais et al., 1997; Ruthven and Ching, 1989). More recently this technique has been developed for the separation of chiral compounds and the purification of biochemical and pharmaceutical products (Nicoud et al., 1993; Pais et al., 1997; Pedeferri, et al., 1999; Lehoucq et al., 2000). The following discussion summarizes some of the literature references disclosing SMB separations.
  • Protein purification At least four SMB processes have been developed for the purification of proteins.
  • Huang et al. (1986) discloses the separation of trypsin from a porcine pancreas extract using affinity chromatography; Houwing (1996) reports the fractionation of human serum albumin using a two column system; Nicoud (1996) reports the separation of myoglobin from lysozyme using an adsorption process; and Schulte et al. (2000) reports reversed phase and silica gel adsorption processes for separating cyclosporine A, cyclic oligopeptides, and various impurities.
  • Ionic Molecules At least two references disclose the use of SMB processes for separating ionic molecules.
  • Van Walsem et al. (1996) discloses the separation of betaine from molasses using two ion exchange columns; and Maki (1992) reports the separation of L-glutathione from glutamic acid using a cation-exchange resin.
  • Organic solvents SMB processes have also been reported for the separation of organic solvents.
  • Szpepy et al. (1975) reports the SMB separation of various C 16 -C 22 methyl esters of fatty acids using an undisclosed stationary phase; and Blehaut et al. (1996) discloses the separation of stereoisomers of phytol using silica gel.
  • Optical isomers SMB process have found special interest of late in the separation of optical isomers.
  • Fuchs et al. (1992), for example, discloses the separation of threonine isomers using a ligand exchange resin;
  • Balannec et al. (1993) discloses the resolution of racemic la-2,7,7a-tetrahydro-3-methoxynaphthyl-(2,3-b) oxirane using microcrystalline cellulose triacetate;
  • Archibald et al. (1999) reports the separation of epoxide diastereomers using reversed phase, normal phase, or chiral stationary phase SMB chromatography.
  • the successful design and operation of a SMB depends upon the correct selection of operating conditions (zone flow rates and switching time), which cause the slow and fast moving components to travel in opposite directions relative to the feed port.
  • zone flow rates and switching time In a linear isotherm system, the primary operating conditions are the zone flow rates and port switching time in zones JT and HI, because these are the two zones where separation initially occurs, and where the zone flow rates and port switching times must be suitably balanced.
  • the port switching time and the flow rates in zones I and IV are also important in order to prevent migration of fast and slow moving components between zones I and IV.
  • Figure 2 illustrates the triangle theory (Storti et al., 1989) based on the local equilibrium theory, and how it is employed to arrive at zone flow rates and port switching time in zones II and HI.
  • an equilibrium analysis which assumes no mass transfer effects
  • the dimensionless velocities represent net zone flow velocities divided by port movement velocity.
  • the triangular region defines the complete set of operating conditions that will guarantee complete separation (100% purity and yield) of the fast and slow moving components in zones II and ELL
  • the triangle method is useful but it is limited to ideal conditions (i.e. without mass transfer resistances). Therefore, the design does not give any information concerning zone length, mass transfer resistances, or desired product purity and yield.
  • the minimum desorbent consumption for nonideal systems can only be approximated from the triangle theory through a lengthy series of trial and enor computer simulations of operating conditions within the triangular region, to determine the desorbent consumption associated with each combination of flow rates in zones II and III, and hopefully to converge eventually on an approximate minimum desorbent consumption.
  • Azevedo and Rodriguez (1999) proved using computer simulations that the standing wave analysis solution (Ma and Wang, 1997) conesponds to the vertex point of the triangular region. The vertex point gives the highest throughput and the lowest desorbent consumption (see Figure 3).
  • Figure 4 is drawn to show the parameters that can be varied to explore system efficiencies and the parameters that should be estimated in the design.
  • the mobile phase can significantly affect process performance and cost because (1) the mobile phase governs the solubility of solutes in the system, and the maximum throughput on a solid concentration basis, (2) the mobile phase may affect selectivity and effective adsorbent capacity, (3) some mobile phase must continuously be replenished, and (4) some of the mobile phase must be separated from the final product, and separation costs can vary from solvent to solvent.
  • the solid phase, and the size of particles in the solid phase can similarly have substantial effects upon process performance and cost because (1) the solid phase dictates the absorbent capacity of the system, the selectivity, and the attendant maximum throughput, and (2) the size and type of solid phase dictates the allowable pressure drop and maximum mobile phase velocity.
  • the designer typically assumes: (1) a fixed number of columns, and (2) an allocation of columns among the zones, when initiating the trial and enor computer simulation process to determine zone flow rates.
  • the allocation of columns can significantly affect the efficiency of the system, and can even determine whether a desired separation is possible.
  • splitting strategy Yet another parameter that can significantly affect the efficiency of a particular system, which is unique to multicomponent separations in which a component having an intermediate affinity is desired, is splitting strategy.
  • Such separations typically require tandem or parallel SMBs in which the intermediate affinity fraction is successively separated from the fast and slow moving fractions.
  • SMBs tandem or parallel SMBs in which the intermediate affinity fraction is successively separated from the fast and slow moving fractions.
  • Another option (not mentioned in the prior art) is whether to allow the component that is not being split in the first ring to distribute in the first ring, and thereby contaminate both extract and raffinate streams.
  • the standing wave analysis proposed by Ma and Wang (1997) represented a significant advance in the design of binary SMBs, because it eliminated the need to perform numerous computer simulations or experimental trials when designing SMBs. Computer simulations could still be performed, but typically only to validate the results of the standing wave analysis.
  • the standing wave analysis described by Ma and Wang (1997) was limited to binary SMBs, and thus did not show how to design and operate an SMB for multicomponent separations.
  • Ma and Wang (1997) also did not show how to overcome a number of other issues associated with the design of SMBs, especially SMBs for multicomponent separations, including column allocation issues, splitting strategies, and how to account for fronting and extra-column dead volume effects.
  • Another object of the invention is to provide algebraic equations that eliminate the need to perform computer simulations when optimizing design and operating parameters of SMBs for multicomponent separations, such as desorbent consumption, column allocation and number, and desorbent/adsorbent combination.
  • Another object of the present invention is to provide SMBs for separating multicomponent mixtures under linear isotherm conditions wherein non-negligible mass transfer effects are observed, using column configurations that maximize economic efficiency.
  • Still another object of the present invention is to provide SMBs for separating multicomponent mixtures under linear isotherm conditions in two or more rings, and methods of designing such SMBs, using splitting rules that optimize economic efficiency.
  • Yet another object of the invention is to use the design methods of Ma and Wang (1997) in the optimization of design and operating parameters of SMBs for binary separations, such as column allocation and number, and desorbent/adsorbent combination.
  • Another object of the present invention to provide SMBs for separating binary mixtures under linear isotherm conditions wherein non-negligible mass transfer effects are observed, using column configurations that maximize economic efficiency.
  • the inventors have surprisingly discovered the design parameters for purifying a component or fraction from any multicomponent mixture using SMB chromatography, under conditions that minimize desorbent consumption for a given rate of feed when linear isotherms and non-negligible mass transfer resistances are observed.
  • the present invention accounts for mass transfer effects when the initial operating parameters are derived, using a novel extension of the standing wave analysis previously reported by Ma and Wang (1997) for binary separations.
  • the invention provides a method of designing an SMB for separating a desired fraction or component from a first fraction or component, in a mixture of N components: a) providing equations that relate the design, operating, and intrinsic engineering parameters of a SMB that displays linear isotherms, wherein the equations assume standing wave conditions for each of the zones; b) prescribing a first set of design and operating parameters sufficient to determine the intrinsic engineering parameters and to solve the equations for separating the multicomponent mixture in a first SMB; and c) solving the equations.
  • the SMB comprises a first ring A that comprises four zones, and the equations assume the following standing wave conditions in each zone: i) in zone I: the desorption wave of component N; ii) in zone II: the desorption wave of component j; iii) in zone HI: the adsorption wave of component j ' +l; and iv) in zone IV: the adsorption wave of component 1; wherein the components are numbered 1 . . . j, j+l, . . . N in order of increasing affinity for the stationary phase, and a split is desired between components j andj'+l.
  • the method enables for the first time SMBs for separating multicomponent mixtures wherein mass transfer resistances are observed and desorbent consumption is minimized.
  • the invention provides a process for chromatographically separating a desired component or fraction from a multicomponent mixture, under linear isotherm conditions wherein non-negligible mass transfer resistances are observed, comprising: a) providing a simulated moving bed that comprises a first ring and a first desorbent stream; b) providing a first feed stream that comprises the desired component or fraction and a first component or fraction, c) introducing the first desorbent stream and the first feed stream to the first ring under conditions sufficient to separate the desired component or fraction from the first component or fraction, and to minimize the rate of the first desorbent stream; and d) withdrawing the desired component or fraction as a raffinate or extract from the SMB, separated from the first component or fraction.
  • the method can be used to design SMBs in which more than one separation is required, and more than one ring is employed.
  • SMBs can be designed to separate an intermediate affinity component from a multicomponent mixture using two or more rings connected in series, in which the slower and faster components or fractions are successively separated from the desired end component.
  • the process further comprises: a) introducing a second desorbent stream and the desired component or fraction from the first ring to a second ring, under conditions sufficient to separate the desired component or fraction from a second component or fraction, and to minimize the rate of the second desorbent stream; and b) withdrawing from the second ring an extract or raffinate that comprises the desired component or fraction separated from the second component or fraction.
  • the standing wave analysis enables SMBs to be designed and operated for multicomponent systems in which mass transfer resistances are observed, for the first time without undue trial and error using experiments or computer simulations, by: (1) reducing the degrees of freedom associated with the design, and (2) providing a simplified set of relationships that consider mass transfer resistances, desired purity and yield, zone length, and the relationship between feed rate and desorbent consumption, at the initial design stage when determining operating parameters.
  • N represents the number of components in the system, numbered from low to high affinity as 1, ..., j, j+l, ..., N, and in which a split is desired between component,/ and +1 u 0 is the interstitial velocity
  • S is the column cross sectional area ⁇ K e ⁇ ⁇ p + - K e ⁇ ⁇ p )a t ) (a t is the partition constant between the adsorbed phase and the liquid phase for component i at an infinitesimal concentration, and S p is the porosity of the particle)
  • K el is the size exclusion factor for component i 1- ⁇
  • ⁇ b is the interstitial bed void fraction
  • F feed is the feed flow rate
  • F des is the desorbent flow rate
  • L is the zone length
  • E b is the axial dispersion coefficient
  • K f is the lumped mass-transfer coefficient ⁇ is related to the ratio of the highest concentration to the lowest concentration of the standing wave in a specific zone
  • the standing wave equations can be used to find the port movement velocity and the linear velocities of the mobile phase in the four zones of the SMB.
  • These equations define the standing wave conditions when both E 6 and R ⁇ are significant.
  • they define the maximum linear velocities for zones III and IV and the minimum linear velocities for zones I and II for a system with linear isotherms. Any lower velocities in zones HI and IV and higher velocities in zone I and II result in better than specified product purities and recoveries.
  • Such designs will have lower throughput or higher desorbent consumption than the standing wave design.
  • the equations also define the minimum desorbent consumption for a given rate of feed, because u is the highest and u ⁇ is the lowest among all the feasible designs that guarantee separation at the prescribed yield and purity.
  • the standing wave design has greatly simplified the process of maximizing process efficiency, for binary and multicomponent systems, by enabling a designer to scan the various design and operating parameters that can be employed.
  • the equations allow a designer to quickly scan various desorbent/stationary phase combinations, or particle sizes for the stationary phase, to derive operating parameters from which process economics can quickly be assessed.
  • one can quickly scan numerous column configuration schemes (varying the total number or length of columns and the allocation of columns among zones) to determine which is the most efficient for a particular system.
  • the invention provides a method of optimizing a SMB system that displays linear isotherms comprising: a) providing equations that relate the design, operating, and intrinsic engineering parameters of an SMB that displays linear isotherms; b) prescribing a first set of design and operating parameters sufficient to determine the intrinsic engineering parameters and to resolve the equations for separating a binary or multicomponent mixture in a first SMB; c) prescribing a second set of design and operating parameters sufficient to determine the intrinsic engineering parameters and to resolve the equations for separating the binary or multicomponent mixture in the second SMB; and d) evaluating and comparing the economic efficiency of the first and second SMBs.
  • the ability to scan various design and operating parameters is significant not only for the design of SMBs; it has also, for the first time, led to the development of SMBs which are optimized for economic efficiency.
  • One of the most significant parameters for influencing economic efficiency is zone length, and the present invention allows all of the components of zone length (i.e. column length, column number, or column allocation) to be readily scanned to find the column configuration of optimum economic efficiency.
  • the invention provides four zone SMBs that comprise five or more columns, and five zone SMBs that comprise six or more columns, in which the columns are allocated among zones in a manner that minimizes the desorbent use, or that maximizes the throughput through the SMB, or that balances these two criteria for optimum economic efficiency.
  • the inventors have also surprisingly discovered optimal splitting strategies for separating an intermediate affinity component or fraction from a multicomponent mixture using SMB chromatography, including strategies that had not heretofore been considered.
  • optimal splitting strategies for separating an intermediate affinity component or fraction from a multicomponent mixture using SMB chromatography, including strategies that had not heretofore been considered.
  • the inventors have discovered that when only the intermediate fraction is desired in high purity from a multicomponent fractionation, the rate of desorbent consumed for a given rate of feed can be minimized by allowing one of the other fractions to distribute in the first ring while separating the intermediate fraction from the second fraction.
  • the easiest split should be performed in the first ring, and the more difficult split performed in the second ring.
  • the invention provides methods for designing SMBs for multicomponent separations by observing the foregoing splitting rules.
  • the invention provides SMBs in which when only the intermediate component or fraction is desired in high purity from a multicomponent fractionation, allowing the slower or faster fraction or component to distribute in the first ring while separating the intermediate fraction from the second fraction.
  • the standing wave design of the present invention can be readily applied to convert an industrial or laboratory batch chromatography process to a simulated moving bed process in order to increase the yield, purity, and throughput of the system.
  • Stationary phase utilization (i.e. throughput) and desorbent consumption can be substantially improved over such processes by appropriate designs of zone lengths, flow rates, and port switching time.
  • Figure 1 is a schematic drawing of a four ring simulated moving bed, showing the introduction points of the feed and desorbent lines, the withdrawal points for extract and raffinate, and the direction of mobile phase and port movement.
  • Figure 2 is a hypothetical graph illustrating the triangular region of SMB operating conditions that assures complete separation of two components in an ideal system, where the normalized mobile phase velocity in zone II is plotted against the x-axis, and the normalized mobile phase velocity in zone III is plotted against the y-axis.
  • Figure 3 is a graph showing the triangular region of SMB operating conditions in zones II and El for an ideal system. Several smaller triangular regions are interposed of operating conditions derived from computer simulations of the SMB assuming various mass transfer resistances, wherein the end-purity of the components is held constant at 99%. The graph is taken from Azevedo and Rodriguez (1999). The vertex point of each triangle represents the conditions under which desorbent consumption is minimized, as determined by the standing wave analysis of Ma and Wang (1997).
  • Figure 4 is a list of parameters involved in the design of an SMB process. The relations among these parameters are also shown in this figure.
  • Figure 5 is a concentration profile of the standing concentration waves in a continuous moving bed for separating two components in a non-ideal system. Solute 1 is the fast moving solute; solute 2 is the slow moving solute.
  • Figures 6a and 6b are concentration profiles of the standing concentration waves in an SMB for separating N components without mass transfer resistances (a), and with mass transfer resistances (b), wherein solute 1 is the fastest moving solute, solute N is the slowest moving solute, and a split is desired between components./ andj+1.
  • Figure 7 is a concentration profile showing six strategies for separating components numbered 1, 2, and 3 in the ring of an SMB, in a mixture where a ⁇ ⁇ ⁇ 2 ⁇ ⁇ 3 (i.e. the affinity of a ⁇ for the stationary phase is less than the affinity of ⁇ 2 for the stationary phase, and so forth), and mass-transfer resistances are neglected.
  • Figure 8 is a flow chart of the various parameters which go into the design of an SMB using the standing wave analysis, and how they are used to arrive at the economic efficiency of a particular SMB.
  • Figures 9(a)-(d) are elution chromatograms charting experimental data and computer simulation results of pulse tests with a lab scale column at various concentrations for (a) blue dextran, (b) NaCl, (c) BHI, and (d) ZnCl 2 .
  • Figures 10(a)-(c) are elution chromatograms charting experimental data and computer simulations of long pulse tests employing a saturated BHI saturation solution eluted with 1 N acetic acid with a pilot-scale column, run at (a) 8.1mL/min, (b) 4.0mL/min, and (c) 2.0mL/min.
  • Figures ll(a)-(c) are chromatograms charting experimental data and simulation results of a multiple BELT frontal test with a lab-scale column, including multiple frontal data without a column (a), multiple frontal data with a column (b), and isotherm absorption data imposed against an anti Langmuir isotherm model (c).
  • Figures 12(a) and (b) present experimental data and simulation results of a) the effluent history at the raffinate port of BFfl, HMWP and ZnCl 2 , and (b) the effluent history at the extract port of BHI, HMWP and ZnCl 2 , in an SMB process for separating BHI from HMWP and ZnCl 2 .
  • Figure 12c) shows mid-cycle column concentration profiles at the 45 th cycle.
  • Figures 13(a) and (b) present experimental data and simulation results of (a) the effluent history at the raffinate port of BHI, and (b) the effluent history at the extract port of BHI, in an SMB process for separating BHI from HMWP and ZnCl 2 .
  • Figure 13(c) shows mid-cycle column concentration profiles at the 45 th cycle. Data are presented for five simulations in which E b for zone ⁇ I varies from the value estimated from the Chung and Wen correlation to 200 times the value estimated from the Chung and Wen coreelation.
  • Figures 14(a) and (b) present experimental data and simulation results of a) the effluent history at the raffinate port of BHI, HMWP and ZnCl 2 , and (b) the effluent history at the extract port of BHI, HMWP and ZnCl 2 , in a second SMB process for separating BHI from HMWP and ZnCl 2 .
  • Figure 14(c) shows mid-cycle column concentration profiles at the 45 th cycle.
  • Figures 15(a) and (b) present experimental data and simulation results of a) the effluent history at the raffinate port of BFfl, HMWP and ZnCl 2 , and (b) the effluent history at the extract port of BHI, HMWP and ZnCl 2 , in a third SMB process for separating BHI from HMWP and ZnCl 2 .
  • Figure 15(c) shows mid-cycle column concentration profiles at the 61 st cycle.
  • Figure 16 is an overlay of elution chromatograms of outlet concentration profiles resulting from single component pulse injections of sulfuric acid, glucose, xylose, and acetic acid at 10 mL/min.
  • Figure 17 is an elution chromatogram of the outlet concentration profiles resulting from pulse injections of a four component mixture comprising sulfuric acid, glucose, xylose, and acetic acid at 80 mL/min.
  • Figures 18(a)-(d) are simulated and experimental zone concentration profiles in an SMB experiment for removing sulfuric acid from glucose, xylose, and acetic acid. Zone concentration profiles are provided separately for sulfuric acid (a), glucose (b), xylose (c), and acetic acid (d).
  • Figures 19(a) and (b) are simulated zone concentration profiles in a two ring SMB process in which sulfuric acid is separated from a mixture of glucose, xylose, and acetic acid in the first ring (a), and acetic acid is separated from glucose and xylose in the second ring (b).
  • Figures 20(a) and (b) are simulated zone concentration profiles in a two ring SMB in which acetic acid is separated from glucose, xylose, and sulfuric acid in the first ring (a), glucose and xylose are separated from sulfuric acid in the second ring (b).
  • Figures 21(a) and (b) are simulated zone concentration profiles in a two ring SMB process in which glucose and xylose are separated from sulfuric acid in the first ring, and acetic acid is allowed to distribute between the product ports in the first ring (a), and glucose and xylose are separated from acetic acid in the second ring (b).
  • Figures 22(a) and (b) are simulated zone concentration profiles in a two ring SMB process in which glucose and xylose are separated from acetic acid in the first ring, and sulfuric acid is allowed to distribute between the product ports in the first ring (a), and glucose and xylose are separated from sulfuric acid in the second ring (b).
  • Figures 23(a) and (b) contain column concentration profiles for the estimation of product concentrations and decay factors for components to be enriched at the raffinate port (a) and for components to be enriched at the extract port (b).
  • Figure 23(c) represents the mixing junction at the feed port showing how the mass balance is calculated for a component of the extract in the estimation of product concentrations and decay factors.
  • Chromatography As used herein, the term chromatography re ers to any analytical technique used for the chemical separation of mixtures and components, that relies upon selective attraction among the components of a mixture for a stationary phase. Examples include adsorption chromatography, partition chromatography, ion exchange chromatography, size exclusion chromatography, and affinity chromatography.
  • adsorbent is used herein generically to refer to the stationary phase used in chromatography for which the mobile phase components exhibit a selective affinity. Because such affinity can take a variety of forms other than adsorption (including size exclusion or complexation), the term refers to stationary phases that adsorb the components of a mixture and to stationary phases that do not technically adsorb components from the mobile phase, but which nevertheless behave as an adsorbent by slowing the migration velocity of one component relative to another in a chromatographic system.
  • Insulin refers to a protein having the following amino acid sequence and structure (naturally occurring insulin), and biologically active analogues and derivatives thereof:
  • the term thus includes insulin which is derived from human, porcine, and bovine species, as well as insulin that is chemically synthesized or expressed using recombinant protein expression systems that use, for example, E-coli or yeast as the host.
  • a prefened insulin is human insulin expressed using a protein expression system.
  • Recombinant insulin will sometimes be referred to herein as biosynthetic human insulin or BHI.
  • An analogue of insulin means an insulin that contains one or more amino acid substitutions, deletions, additions, or rearrangements compared with human insulin at sites such that the insulin analogue still participates in the metabolism of carbohydrates when administered in vivo, and can be used in the treatment of diabetes mellitus.
  • insulin analogues include LysB28, ProB29-human insulin, AspB28-human insulin, and GlyA21, ArgB31, ArgB32-human insulin.
  • Insulin derivatives include naturally occurring insulin and insulin analogues that are chemically or enzymatically derivatized at one or more constituent amino acids, including side chain modifications, backbone modifications, and N- and C- terminal modifications, by for example acetylation, acylation, hydroxylation, methylation, amidation, phosphorylation or glycosylation, and that retain the in vivo biological activity of insulin.
  • An example of an insulin derivative is myristoyl-N ⁇ -LysB29-human insulin.
  • a simulated moving bed refers to a modified continuous moving bed configuration in which the movement of the adsorbent phase is simulated by the periodic movement of feed, desorbent, extract, and raffinate lines in the direction of mobile phase flow.
  • the typical simulated moving bed contains four separate zones and at least 4 identically sized columns allocated among the zones.
  • the SMB need not consist of four zones precisely. For example, the number of zones can be reduced to three if the desorbent is not regenerated, and is instead withdrawn completely in the raffinate and extract streams. Similarly, five zones can be employed to allow three streams to be withdrawn from the ring, especially where four or more components are separated in the system.
  • the SMB can comprise one or more regeneration zones.
  • Ring is used to describe how the zones are configured in relation to one another in an SMB because the output of each zone comprises the input for the successive zone, in a circular fashion.
  • the term “ring” should thus not be understood to be limited to a circular configuration of the zones and columns within the zones.
  • Linear Isotherms ⁇ A system that exhibits linear isotherms refers to a system in which the partition coefficient of component i (i.e. the ratio of the concentration of component i in the mobile phase divided by the concentration of component i in the stationary phase) is constant.
  • the migration velocity of component i is governed by the following equation:
  • Weak adsorption (with values for ; approaching zero) can be represented using an apparent Ke that has a higher value than the intrinsic Ke.
  • Many types of chromatographic systems can be operated within a region of linear isotherms. However, some systems observe nonlinear isotherms over a larger range of operating conditions, or a larger number of feed/adsorbent combinations. Systems typically depart from linear isotherms as component concentration is increased (or absorbent capacity is decreased), and competition among the components for available adsorbent capacity becomes a factor.
  • Multicomponent mixture refers to a fluid mixture that comprises three or more components or fractions which can be separated using a prescribed chromatographic process, because each component or fraction displays a different affinity for the adsorbent employed.
  • a component or fraction is said to be purified when its relative concentration (weight of component or fraction divided by the weight of all components or fractions in the mixture) is increased by at least 20%. In one series of embodiments, the relative concentration is increased by at least 40%, 50%), 60%, 75%, 100%), 150%, or 200%).
  • a component or fraction can also be said to be purified when the relative concentration of components from which it is purified (weight of component or fraction from which it is purified divided by the weight of all components or fractions in the mixture) is decreased by at least 20%, 40%, 50%, 60%, 75%, 85%, 95%, 98%, or 100%.
  • the component or fraction is purified to a relative concentration of at least 50%), 65%, 75%, 85%, 90%, 95%, 97%, 98%), or 99%).
  • a component or fraction in one embodiment is “separated” from other components or fractions, it will be understood that in other embodiments the component or fraction is “purified” at the levels provided herein.
  • the intermediate component or fraction of a multicomponent mixture refers to the component (or fraction of components) which, in a particular chromatographic system, displays an affinity for the stationary phase that is intermediate of the affinities of at least two other components or fractions for the stationary phase.
  • the "components" or “fractions” of a mixture do not include solvents and other materials that do not exhibit a significant affinity for the adsorbent, and are separated from the desired components after SMB processing through conventional chemical operations such as distillation.
  • Mass transfer effects refer generally to those physical phenomena which cause components of a mixture to display distinct dispersion behavior from the mixture in a given system, and to depart from the ideal system. Mass transfer effects thus include those effects modeled using axial dispersion coefficients, intraparticle diffusion coefficients, and film mass transfer coefficients. Mass transfer effects thus also include fronting and dispersion due to extra-column dead volume. A separation is hindered by non-negligible mass transfer effects if the mass transfer correction term, discussed in more detail below, is more than 2% of the mobile phase velocity in any of the zones as prescribed by equations 3-6 (the mobile phase velocities for an ideal system), to achieve a prescribed purity and yield.
  • the designs of the present invention can also extend to systems in which the mass transfer correction velocity increases or decreases the mobile phase velocity for the ideal system by more than 1, 3, 5, 7.5, 10, 15, 20, 30, 50, 75, 100%), 200%), 400%, 600%, 1000%, or more.
  • Fronting and/or extra-column effects - Fronting refers to a physical phenomenon that concentration waves spread in the direction of fluid flow as a result of anti-Langmuir type of isotherm (Karger et al., 1973) or deviation from plug flow. The deviation is usually caused by non-uniform packing, viscosity effects (Flodin, 1961; Collins, 1961), etc.
  • Extra- column effects refer to the effects that are resulted from extra-column dead space including connecting tubing, sampling lines, void space of column heads. The extra-column effects result in a delay in solutes travelling through a chromatographic system and also result in concentration wave spreading in addition to the spreading caused by the aforementioned intra-column mass transfer effects.
  • a first component or fraction is said to be separated from a second component or fraction when the weight ratio of the second component or fraction to the first component or fraction is reduced by at least 50%.
  • the ratio is preferably reduced by at least 65%, 75%, 85%, 90%, 95%, 98%, or 99%.
  • first or second is used to describe a particular element of a system or method. It should be understood that the numeric identifier is used simply for drafting convenience, and should not be interpreted to imply the existence of a second from the presence of a first. Thus, reference to a first ring allows for the presence of a second ring, but does not imply such presence. Similarly, where first and second things are prescribed, such as first and second stationary phases of two tandem rings, it is to be understood that these two things may be the same (such as the same type of absorbent), but that the things are distinct in a spatial or other separative sense.
  • the throughput or desorbent use for a particular system is "optimized,” it will be understood that the throughput approaches or equals the maximum level obtainable under standing wave conditions, or that the desorbent use approaches or equals the minimum desorbent use obtainable under standing wave conditions, for the particular system being investigated.
  • throughput is maximized if it exceeds 85%, 90%, 95%, 97%, 98%, or 99% of the standing wave maximum.
  • Desorbent use is minimized if it is less than 125%, 115%, 110%, 105%, 103%, 102%), or 101% of the standing wave minimum.
  • Desorbent means the mobile phase which is added to a SMB ring between the raffinate and extract ports.
  • desorbent is used because, when combined with the total mobile phase, the desorbent contributes to the desorption waves of the various components in their respective zones, whether by increasing the mobile phase velocity in a size exclusion system, or by physically displacing or removing the solution from the solid phase.
  • the term “desorbent” is not meant to imply that the desorbent composition has any particular desorbent capacity, or that it desorbs through any particular physical mechanism.
  • Desorbent use refers to the volumetric flow rate of desorbent into a SMB, and can thus be measured directly at the desorbent inlet line. Desorbent use is minimized if, for a given separation in a defined system, desorbent use is less than 125%, 115%, 110%, or 105% of the theoretical desorbent use defined by the standing wave equations. The desorbent use is preferably less than 102% or 101% of the theoretical use.
  • the inventors have surprisingly discovered the standing wave design parameters for separating slow and fast moving components or fractions from any multicomponent mixture (containing three or more components) using SMB chromatography, under conditions that minimize desorbent consumption for a given rate of feed when linear isotherms and non-negligible mass transfer resistances are observed.
  • the methods can be used to achieve pseudo-binary separations of multicomponent mixtures, as well as separations of intermediate affinity components and/or fractions from multicomponent mixtures.
  • a pseudo-binary split the desired component or fraction of a mixture will be the fastest or the slowest of a multicomponent mixture, and will be separated from the other component or fraction in one separation step, using only one SMB ring.
  • the intermediate component or fraction will need to be separated from the faster and slower components or fractions in two successive separations, using two SMB rings.
  • the invention provides a process for chromatographically separating a desired component or fraction from a multicomponent mixture, under linear isotherm conditions wherem non-negligible mass transfer resistances are observed, comprising: a) providing a simulated moving bed that comprises a first ring and a first desorbent stream; b) providing a first feed stream that comprises the desired component or fraction and a first component or fraction; c) introducing the first desorbent stream and the first feed stream to the first ring under conditions sufficient to separate the desired component or fraction from the first component or fraction, and to minimize desorbent use; and d) withdrawing the desired component or fraction as a raffinate or extract from the SMB, separated from the first fraction or component.
  • the method will further comprise: a) introducing a second desorbent stream and the desired component or fraction from the first ring to a second ring, under conditions sufficient to separate the desired component or fraction from the second component or fraction, and to minimize the second desorbent use; and b) withdrawing from the second ring an extract or raffinate that comprises the desired component or fraction separated from the second component or fraction.
  • the SMBs designed according to the present invention can be optimized in a number of respects according to the various optimization strategies discussed herein.
  • the SMBs can be designed to assure that columns are allocated among zones in a manner that maximizes economic efficiency, or to assure that the splitting sequence employed is the most economical.
  • the methods can be readily adapted to accommodate extra-column effects and/or non-ideal flow effects such as fronting.
  • a particularly important feature of the SMBs of the present invention is their optimal balance of desorbent use and raw material feed for a given system. In other words, desorbent use is minimized for a given rate of feed, and the rate of feed is maximized for a given rate of desorbent use.
  • desorbent use is minimized for a given rate of feed, and the rate of feed is maximized for a given rate of desorbent use.
  • the design methods of the present invention are preferably performed in a computer environment.
  • the system is preferably designed by inputting one or more design or operating parameters into a machine in which equations that govern the SMB are encoded, and obtaining from the machine an output that defines one or more other design or operating parameters.
  • the design method is typically followed by the step of performing an SMB separation (either experimentally or using a computer simulation) under the conditions determined by the design.
  • an SMB separation either experimentally or using a computer simulation
  • it may be beneficial to adjust the flow rates determined by the design in which case one would relax the standing wave conditions determined by the design in one or more of the zones, before performing the separation.
  • the following is an explanation of the method and its theoretical underpinnings.
  • Zone I This zone is located between the desorbent inlet and extract withdrawal and used for desorbing the slow fraction. A portion of the stream leaving this zone is withdrawn as the extract while the remainder flows into zone E.
  • Zone E This zone is used for desorption of the fast fraction and separation from the slow fraction.
  • Zone IE This zone is used for the separation of the slow fraction from the fast fraction. The fast fraction is partially withdrawn at the raffinate port.
  • Zone IN This zone is used for adsorption of the fast fraction as well as desorbent recovery. The stream leaving this zone should ideally contain only pure desorbent which is recirculated to the desorbent inlet for reuse.
  • Equations 13 and 14 indicate that the migration velocity of the slow fraction (component 2) should be less than vin zone IE and that of the fast fraction (component 1) should be less than in zone IV.
  • Equations 15 and 16 indicate that the velocity of component 2 should be greater than vin zone I and that of component 1 should be greater than vin zone E.
  • Equations 13- 16 define all the feasible zone flow rates and the port movement velocities needed to guarantee separation. These equations must be satisfied in order to have separation for systems with or without mass-transfer resistances.
  • FIG. 13-16 there are infinite combinations of zone flow rates and port movement velocities (switching times) that guarantee separation in a given system.
  • the optimal zone flow rates and port movement velocity that give the highest throughput and the lowest desorbent consumption for a multicomponent mixture can be found by selecting the proper concentration waves to remain standing in each zone (i.e. the standing wave analysis).
  • Figure 6a illustrates, for a system without mass transfer effects, which waves must remain standing in each zone to assure complete separation of two fractions in a multicomponent system.
  • Figure 6a shows that:
  • N represents the number of components in the system, numbered from low to high affinity as
  • u 0 is the interstitial velocity of the mobile phase, and the u 0 determined by equations 17-21 represents what is termed herein the "ideal mobile phase velocity.”
  • S is the column cross sectional area
  • ⁇ t ( ⁇ K d ⁇ p + (1 - K ei ⁇ p )a t ) is the retention factor for linear systems
  • a i is the partition constant between the adsorbed phase and the liquid phase for component i at an infinitesimal concentration, and ⁇ is the porosity of the particle
  • K ei is the size exclusion factor for component i for the packing material used in the SMB.
  • the variables a E p , K ei , and£ & are intrinsic engineering parameters that can be estimated from elution chromatograms or, for some systems and packing materials, may be taken from the literature or manufacturer's specifications.
  • equations 17-20 would need to be appropriately modified.
  • the desired component was the fastest component, then the retention factor of component 1 would be used to determine the mobile phase in zone E.
  • the retention factor of component N would be used to determine the mobile phase velocity in zone IE.
  • desorbent consumption and feed rate are two of the principle factors that determine system cost and efficiency
  • the ability to quickly identify maximum feed rates for a given desorbent use, or minimum desorbent use for a given feed rate allows one to quickly and efficiently scan various desorbent/feed ratios and determine system parameters that optimize the economic efficiency of the system. This is especially true in the ideal systems governed by the above operating parameters, in which 100% product yield is assumed.
  • the waves can be confined to their respective zones, and pure products obtained at the extract and raffinate ports, by increasing the zone lengths, or altering the velocities of port movement and of the mobile phase.
  • some degree of contamination is typically acceptable and even desirable from an economic standpoint.
  • the SMB can also be designed to allow for a particular degree of contamination, and to allow a more economical operation of the SMB.
  • the SMBs of the present invention can be designed to accommodate systems in which separation is hindered by non-negligible mass transfer resistances, regardless of which mass transfer mechanism or mechanisms predominate, by an extension of the foregoing standing wave concepts to non-ideal systems.
  • equations 17-21 are modified by correction terms to account for the wave spreading of the components whose concentration waves are kept standing in each zone, attributable to the mass transfer effects.
  • the complete solutions of the Unear velocities for multicomponent separation in the four zones in a SMB, corrected for the effects of mass transfer, can be found from equations 3-6, based on the concept of standing waves (Figure 6b).
  • the right side of equations 3-6 is the mass transfer correction term for each zone, which establishes the extent to which the "ideal mobile phase velocity" must be altered in each zone to accomplish a desired level of separation.
  • the mass transfer correction factors account for wave spreading by the components retained (or released) in each zone.
  • the mobile phase velocity which is allowable in each zone is a function of zone lengths, product purities desired, isotherms and mass-transfer parameters. Isotherms are accounted for through the retention factor for the mass center velocity for each concentration wave that remains standing in the four zones. The remaining factors are associated exclusively with mass transfer, and are taken into account by the mass transfer correction term for each concentration wave that must remain standing in the four zones.
  • L is the zone length
  • E b is the axial dispersion coefficient
  • K f is the lumped mass-transfer coefficient
  • ? is related to the ratio of the highest concentration to the lowest concentration of the standing wave in a specific zone ( Figure 6b), and the remaining terms are as defined above.
  • the intrinsic engineering parameters E b and K f can be determined experimentally by small pulse or large pulse elution chromatograms, or in some systems by reference to the literature.
  • the lumped mass-transfer coefficient, KfJ for component i in zone m can be calculated from the following equation:
  • the film mass transfer coefficient (kf) can be estimated from literature correlations (e.g. Wilson and Geankoplis, 1966).
  • the pore diffusivity (D p ) is estimated from pulse tests (Wooley et al., 1998; Wu et al., 1998).
  • the ⁇ terms are the index of product purity and yield, and ultimately determine the overall purities and yields of products at the extract and raffinate ports. Because of their relationship to concentration gradients within a particular zone, the ⁇ terms are sometimes referred to herein as "decay coefficients.”
  • the ⁇ terms are derived solely as a function of component concentrations within particular zones, ⁇ 1 " , for example, is the natural logarithm of the ratio of the concentration at the inlet to zone III ( _f max ) to that at the raffinate port for component 2 ( Figures 5 and 24).
  • ⁇ 1 value the higher the product purity of component 1 in the raffinate and the higher the yield of component 2 in the extract.
  • To achieve high yield of component 1 in the raffinate and high purity of component 2 in the extract requires a high ⁇ " value.
  • both high ⁇ " 1 and ⁇ " values are needed.
  • ⁇ values can be derived with knowledge of the purity and yield of one of the components to be separated in a two component system, or with knowledge of the yield of all components (or fractions) in an N component system.
  • ⁇ values can be derived with knowledge of the purity of the two components to be separated through mass balance analyses, because knowledge of the purity of the two components to be separated confers knowledge of purity and yield of both of the components to be separated, and vice versa.
  • a detailed method of determining the decay coefficients for each component is provided in the examples hereto. For many systems, however, an approximate ⁇ value will suffice.
  • equations 3-6 can be used to find the four linear velocities and the port movement velocity. These equations define the standing wave conditions when both E b and K f are significant.
  • Desorbent consumption (27) where BN is the total bed volume, cf eed is the concentration of component i (product) in the feed, and /is the ratio of desorbent flow rate to feed flow rate. The higher the feed flow rate, the higher the throughput. The higher the ratio of desorbent flow rate to feed flow rate, the higher the desorbent consumption.
  • equations 1-6 would need to be appropriately modified.
  • the mass transfer correction term for component 1 would be used to determine the mobile phase in zone II.
  • the mass transfer correction term for component ⁇ would be used to determine the mobile phase velocity in zone EL
  • the invention provides a method of designing an SMB for separating a desired fraction or component from a first fraction or component, in a multicomponent mixture of components numbered 1 . . . j, j+ 1, . . . N, from lowest to highest affinity, comprising: a) providing equations that relate the design, operating, and intrinsic engineering parameters of a SMB that displays linear isotherms, wherein the equations assume standing wave conditions for the SMB; b) prescribing a first set of design and operating parameters sufficient to determine the intrinsic engineering parameters and to solve the equations for resolving the multicomponent mixture in a first SMB; and c) resolving the equations.
  • the equations preferably assume the following standing wave conditions: in zone I: the desorption wave of component N, in zone E: the desorption wave of component j; in zone El: the adsorption wave of component +1; and in zone IN: the adsorption wave of component 1; wherein the components are numbered 1 . . . j, j+1, . . . N in order of increasing affinity for the stationary phase, and a split is desired between components j and j+1.
  • the standing wave conditions for the first ring described in strategy S6 in Tables 1 and 2 is preferably observed (when a split is desired between components/ and ' +l is desired).
  • the standing wave conditions set forth for strategies S1-S6 in Tables 1 and 2 are preferably observed.
  • the concentration waves remain standing in the sense that the adsorption waves and desorption waves in the system, when plotted in a two dimensional graph that relates component concentrations to bed position (see, Figure 6b), are assumed to remain standing as if in a continuous moving bed. In reality, the waves will migrate at a continuous rate into the successive zone until brought back to their respective zones by a port movement. Generally speaking, however, it has been determined experimentally and by using computer simulations that increasing the frequency of movement, and increasing the number of columns in the separation zones (zones I and E), overcomes this problem.
  • the waves remain standing in the sense that the equations define an ideal mobile phase velocity and port movement velocity for a given rate of feed, modified to account for wave spreading due to mass transfer effects.
  • the mobile phase velocity is increased to compensate for the wave spreading
  • the mobile phase velocity is decreased to compensate for the wave spreading.
  • the invention provides a method of designing an SMB for separating a desired fraction or component from a first fraction or component, in a multicomponent mixture, comprising: a) providing equations that relate the design, operating, and intrinsic engineering parameters of a SMB that displays linear isotherms, b) prescribing a first set of design and operating parameters sufficient to determine the intrinsic engineering parameters and to solve the equations for resolving the multicomponent mixture in a first SMB; and c) resolving the equations; wherein the equations define the minimum rate of desorbent use for a given rate of feed.
  • the SMBs of the present invention can also be designed to accommodate fronting and extra-column effects, in addition to intra-column mass transfer effects.
  • the separations achieved in the various methods of this invention are hindered by non-ideal flow effects such as non-negligible fronting and/or extra-column effects (see Definitions section for explanation). Separation is hindered by fronting and/or extra- column effects if such effects would cause significant contamination among zones, but for the design of the system to account for the effects.
  • extra-column effects are considered non-negligible if, when assessed according to the methods prescribed below, they cause ⁇ ⁇ for a standing wave in any zone to change by more than 1%, 2%, 5%, 10%, or 20% from the value it would have assumed but for the extra-column effects.
  • Fronting and/or extra-column effects are also considered non-negligible if, when assessed according to the methods prescribed below, they cause the mass transfer correction term for the standing wave of component i in any zone to change by more than 1%, 5%, 20%, 50%, 200%, or 1000% from the value it would have assumed but for the fronting and/or extra-column effects.
  • equations 1-6 can be suitably modified to take into account the increase in wave retention time and additional wave spreading caused by these phenomena.
  • the increase in retention time in linear systems can be taken into account by adding the time for a wave to go through the extra-column dead volume to the retention factor ⁇ .
  • a modified retention factor ⁇ j can be used as follows:
  • V CSTR is the extra- column dead volume of each column (which can be modeled as a continuous stirred tank
  • P( ⁇ -) is the bed phase ratio
  • ⁇ h is the interstitial bed void fraction
  • ⁇ p is the
  • the additional wave spreading due to dispersion in extra-column dead volume or fronting can be taken into account by multiplying by a factor f the intra-column dispersion coefficient E in each zone, which is estimated from a literature correlation of Chung and Wen (1968) for flow in a uniformly packed column.
  • This factor can be estimated from batch chromatography data or from SMB data. This factor can range from unity in a pilot scale SMB from U.S. Filter, to as high as 200 in some SMBs from AST (Advanced Separation Technologies).
  • the factor/ can take into account the dispersion due to fronting, tailing, or other non-ideal flow effects.
  • the splitting strategy used for a particular system must be selected before the design of a multicomponent SMB separation for an intermediate component can be completed.
  • the process of trial and error selection of operating parameters is substantially complicated, because each of the various designs must be simulated for each of the splitting strategies.
  • the design parameters of the present invention have lead to (1) novel splitting strategies for multicomponent systems, and (2) a set of rules for determining optimal splitting strategies, based upon the relative ease of splitting the various components, and the number of pure components that are desired from a particular system.
  • the methods of the present invention also encompass various splitting strategies which can be employed in the purification of insulin when it is the intermediate affinity component of a multicomponent mixture.
  • various splitting strategies which can be employed in the purification of insulin when it is the intermediate affinity component of a multicomponent mixture.
  • two 4-zone SMB units are operated in series, there are several splitting strategies that can be taken to obtain purified products.
  • four such strategies (S1-S4) are shown for separating components numbered 1, 2, and 3, in a mixture where the order of retention is as follows: a ⁇ ⁇ 2 ⁇ 3 .
  • component 1 is split from 2 and 3 in the first SMB unit (ring A), and the extract product (which contains unresolved components 2 and 3) is fed to a second SMB unit (ring B), in which component 2 is split from component 3.
  • Figure 7b is similar: component 3 is split from 1 and 2 in ring A, and 1 is split from 2 in ring B.
  • An N component system in which purified component j was desired would observe similar splitting strategies, as set forth in Table lb.
  • Table lb Standing waves for various splitting strategies for N components.
  • components 1 . . . j-1 would be withdrawn in the extract of the first SMB unit, and the raffinate product (which contains unresolved components j . . . ⁇ ) would be fed to the second SMB unit in which component j would be separated from components j+1 . . . ⁇ .
  • Strategy S2 is similar: components 1 . . . j would be split from components j+1 . . . ⁇ in the first ring, and components 1 . . . j-1 would be separated from component j in the second ring.
  • Figures 7a-b show how to obtain three pure products from a mixture of three components. In most cases, however, it is not necessary to obtain every component in high purity and high yield. When only the middle product is required in high purity and high yield, the inventors have unexpectedly determined that it may be preferable to allow one component to distribute throughout the entire SMB unit. For example, in Figure 7c, components 2 and 3 are separated from 1 in the first ring, so component 1 should be kept from the extract and collected in the raffinate; however, if components 1 and 3 are not required in high purity, then the standing wave condition in zone I can be relaxed to allow part of component 3 to "wrap around" (that is, to migrate from zone I into zone IV) and be taken with component 1 in the raffinate.
  • the tandem SMB for multicomponent fractionation is not limited to two 4-zone rings. When five zones are employed in the first ring of the tandem SMB system, there is increased flexibility in the design.
  • Figures 7e-f show schematic representations of two such processes for the separation of three components.
  • component 1 is completely split from components 2 and 3 in the first SMB unit (ring A). Some of component 3 is withdrawn purified in an extract stream, and the remainder contaminates an intermediate extract stream that also contains component 2.
  • the intermediate extract stream component (which contains unresolved components 2 and 3) is fed to a second SMB unit (ring B), in which component 2 is completely split from component 3.
  • Figure 7f is similar: component 3 is completely split from 1 and 2 in ring A, some of component 1 is withdrawn purified in a raffinate stream, and the remainder contaminates an intermediate raffinate stream that also contains component 2.
  • the intermediate raffinate stream is fed to the second SMB unit for separating components 1 and 2.
  • A" ⁇ are positive numbers, the first subscript indicates the zone number and the second subscript indicates the solute.
  • E A PS ⁇ b v A ( ⁇ 3 - ⁇ 1 ) (40) where D A is the desorbent flow rate for ring A and EA is the extract flow rate. Combining equations 38 and 39 leads to the following:
  • the extract from ring A is the feed to ring B, so that
  • Ring A is a five-zone SMB and Ring B is a four-zone SMB; whereas in other strategies, both Ring A and Ring B are four-zone SMBs. **Component7 is the desired product.
  • strategies S3 and S4 lead to lower solvent consumption and lower product dilution. This is achieved at the price of obtaining only one purified product.
  • Strategies SI and S2 lead to higher desorbent usage and higher product dilution, but three purified products are obtained. It should also be noted that in choosing among strategies SI and S2, lower desorbent consumption and lower product dilution are preferred. From Table 3a, one can see that if ( ⁇ - > ( ⁇ s - ⁇ z), then strategy SI should be chosen because the desorbent requirement and the product dilution are lower. If ( ⁇ - ⁇ ) > ( ⁇ - ⁇ ), then strategy S2 should be chosen. In other words, the easier separation should be performed first if three purified products are desired.
  • the tandem SMB for multicomponent fractionation is not limited to two 4-zone rings. When five zones are employed in the first ring of the tandem SMB system, there is increased flexibility in the design.
  • Figures 7e-f shows schematic representations of two such processes for the separation of three components.
  • Table 3a (strategies S5 and S6) shows a comparison of the system performance under the operating conditions shown in Figures 7e-f. An inspection of Table 3 a reveals that the easier separation should be performed in the first ring to minimize solvent consumption. Neither configuration in strategies S5 and S6 has an advantage in terms of product dilution because both result in an undiluted center product.
  • Ring A is a five-zone SMB and Ring B is a four-zone SMB; whereas in other strategies, both Ring A and Ring B are four-zone SMBs. **Component is the desired product.
  • the invention provides the multicomponent SMB separation processes of the present invention wherein: (a) the intermediate fraction is withdrawn in the extract of the first SMB; and (b) the difference between the affinities of the intermediate fraction and the slow fraction for the stationary phase in the first SMB is less than the difference between the affinities of the intermediate fraction and the fast fraction for the stationary phase in the first SMB.
  • the invention provides the multicomponent SMB processes of the present invention wherein: (a) the intermediate fraction is withdrawn in the raffinate of the first SMB; and (b) the difference between the affinities of the intermediate fraction and the slow fraction for the stationary phase of the first SMB is greater than the difference between the affinities of the intermediate fraction and the fast fraction for the stationary phase for the first SMB.
  • Embodiments Premised Upon Allowing One Component to Distribute in First Ring Using the standing wave design of the present invention, it has also been proven theoretically that an alternative splitting strategy should be observed when there is no need to obtain the slow or fast fraction in purified form. In this situation it does not matter whether the easier separation is performed first. However, it has proven beneficial to design the system without taking into consideration one of the components in the design of the first ring, and to allow one of the components or fractions to distribute between the raffinate and extract during separation.
  • the standing waves conditions for each zone are preferably specified as set forth in tables 1 and 2 for strategies S3 and S4.
  • the invention provides a method for chromatographically purifying an intermediate affinity fraction from a multicomponent mixture, under linear isotherms, wherein non-negligible mass transfer effects are observed, comprising: a) providing a simulated moving bed that comprises first and second rings and first and second stationary phases; b) providing first and second desorbent streams; c) providing a first feed stream that comprises a fast fraction or component, an intermediate fraction or component, and a slow fraction or component; d) introducing the first desorbent stream and the first feed stream to the first ring under conditions sufficient to separate the intermediate fraction from the fast or slow fraction or component; e) allowing the fast or slow fraction or component to distribute in the first ring; f) withdrawing from the first ring the intermediate fraction or component separated from the fast or slow fraction or component; g) introducing the second desorbent stream and the intermediate fraction or component from the first ring
  • the invention provides a method of designing an SMB for separating an intermediate fraction from a fast fraction and a slow fraction in a multicomponent mixture, under linear isotherms, in ideal and non-ideal systems, wherein the SMB comprises first and second rings, comprising: a) providing splitting rules which provide: i) when only the intermediate fraction or component is purified, allowing the fast or slow fraction or component to distribute in the first ring; ii) when the fast and/or slow fractions and/or components are purified, perform the easiest split in the first ring; and b) selecting a splitting strategy based upon whether one or more of the fast and/or slow fractions and/or components is purified.
  • the foregoing splitting rules and methods can be integrated into each of the various methods of design and operation set forth herein, for separating an intermediate component using tandem rings.
  • One of the most significant advantages of the design method of the present invention is the ability to screen with relative ease design and operating variables to arrive at a SMB that is most economically efficient.
  • the method can be used to optimize SMB separations of binary and multicomponent mixtures.
  • the prior art only provided equations from which one could derive the triangular region of separations for an ideal system (which in turn defined the relative velocities of port movement and mobile phase in zones E and El of the SMB). Any further design of the system to account for mass transfer effects, or to identify the minimum desorbent consumption, required one to perform lengthy and complex computer simulations or expensive experimental trials.
  • the design equations of the present invention simplify this process by several orders of magnitude, by reducing all of the variables associated with the design of an SMB to a simplified set of five equations 1 and 3-6. By resolving equations 1 and 3-6 one is able to arrive at the complete set of operating conditions for an SMB for separating multicomponent mixtures, in a non-ideal system under linear adsorption isotherms.
  • FIG. 8 A schematic drawing that shows the interrelationship of these various parameters is contained in Figure 8.
  • one first assumes a number of design and operating parameters, performs some initial tests and calculations to determine retention and mass transfer parameters, and performs some initial calculations to determine maximum pressure drop allowable by the adsorbent (and hence maximum interstitial velocity). From these initial assumptions and determinations, one is able to perform the standing wave analysis to determine the maximum feed flow rate, maximum feed concentrations, the zone flow rates, and the port movement velocity of the SMB which will achieve a prescribed separation.
  • any of the design or operating parameters can be varied to determine whether other systems might be more economically efficient.
  • the standing wave design eliminates the need to perform complex computer simulations each time any of the design or operating parameters is varied, and allows variations to be evaluated simply by iteratively resolving the standing wave equations for each choice of variables (after performing any calculations or experiments needed to derive a new set of engineering parameters). Once these iterations have been performed, one is able to compare the throughput of the system per bed volume and the rate of solvent consumption for each SMB system being evaluated. Because throughput and solvent consumption are the dominant considerations when evaluating the economic efficiency of a particular SMB system, on both an operating and capital cost basis, they can be used to screen the various systems for economic efficiency.
  • the invention provides several embodiments for optimizing SMBs using the design equations of the present invention, varying a number of the design parameters.
  • the invention provides a method of optimizing a non-ideal SMB system that displays linear isotherms comprising: a) providing equations that relate the design, operating, and intrinsic engineering parameters of an SMB; b) prescribing a first set of design and operating parameters sufficient to determine the intrinsic engineering parameters and to resolve the equations for a first SMB; c) prescribing a second set of design and operating parameters sufficient to determine the intrinsic engineering parameters and to resolve the equations for a second SMB; and d) evaluating and comparing the economic efficiency of the two or more SMBs.
  • the standing wave conditions are preferably observed in the equations used in the foregoing design.
  • the equations preferably define the minimum desorbent use for a given feed flow rate.
  • the design parameters are preferably selected from: zone length, desorbent type, adsorbent type, adsorbent particle size, and splitting strategy, and the foregoing method can be used to screen variations in any of these design parameters.
  • zone length desorbent type
  • adsorbent type adsorbent particle size
  • splitting strategy a method that can be used to screen variations in any of these design parameters.
  • alternative splitting strategies need not be explored, because of the splitting rules already established by the present invention.
  • the standing wave equations define the maximum feed flow rate for a given desorbent flow rate, or minimum desorbent flow rate for a given feed flow rate, the equations eliminate the need to vary the feed flow rate or the desorbent flow rate when performing the design.
  • zone length has three components: the length of columns used in the SMB, the number of columns employed, and the allocation of columns among the zones. Variability in any of these components can be explored using the design method of the present invention, as discussed more fully below.
  • the operating parameters integrated into the above equations are preferably selected from: zone flow rates, port movement velocity, feed flow rate, desorbent flow rate. Again, any of these parameters can be varied in the screening process of this invention to optimize the SMB.
  • the intrinsic engineering parameters for a nonideal system that displays linear adsorption isotherms are the mass transfer parameters and the retention parameters. As discussed above, these parameters can be well modeled and accounted for in the equations by the axial dispersion coefficient E b and the lumped mass transfer coefficient K f .
  • the retention parameters can be well represented in the equations by the term ⁇ , because this term accounts for all retention phenomena (size exclusion and adsorption) in a linear relationship.
  • ⁇ the retention parameters
  • various equations and variables are defined in the field of chemical engineering to model the effects of mass transfer, or the relative retention of components in a linear chromatographic process, and the invention is intended to encompass all such equations and variables.
  • a straightforward approach to optimizing the economic efficiency of a system is set forth in the examples hereto.
  • One of the principal features of this approach is to first determine maximum mobile phase velocity allowed by the adsorbent or the SMB equipment, because the SMB must be operated at the maximum allowable mobile phase velocity in order to maximize throughput for the system (i.e. within about 85, 90, 95, 97, 98%, or 99%, of the maximum allowable mobile phase velocity).
  • Most adsorbents and SMB equipment valves, columns, connectors, and pumps
  • have a maximum pressure limit or a maximum flow rate limit One can calculate from the standing wave equations (equations 25 and 1-6) the max feed flow rate and the mobile phase velocities that can meet the purity requirement.
  • the velocity in each zone should not exceed the corresponding maximum velocity calculated from the pressure or flow rate limit. If this condition is not satisfied, the feed flow rate in the standing wave equations (equations 1-6) is reduced until this condition is satisfied. In a like manner, one can start with a mobile phase velocity of a lesser percentage of the maximum flow rate limit, such as 50%, to arrive at the operating parameters.
  • feed flow rate may be fixed because of the output requirements or raw material production constraints of the manufacturer.
  • cost of the raw material will be so high, and the purity requirements so demanding (such as in pharmaceutical applications), that product purity and yield will essentially be dictated at the outset.
  • the invention provides SMBs for binary and multicomponent separations, in nonideal separations wherein linear isotherms are observed, in which column configuration is optimized for economic efficiency.
  • the invention provides a method for chromatographically separating two or more components or fractions using a SMB, under linear isotherm conditions wherein non- negligible mass transfer resistances are observed, comprising: a) providing a simulated moving bed that comprises a stationary phase, a first desorbent stream, and a first ring that comprises 4 zones and 5 or more columns, or 5 zones and 6 or more columns; b) providing a feed stream that comprises a fast and a slow fraction and/or component; c) introducing the desorbent stream and the feed stream to the SMB under conditions sufficient to separate the fast fraction or component from the slow fraction or component; and d) withdrawing the fast fraction or component as a raffinate and the slow fraction or component as an extract from the SMB; wherein: the columns are allocated among the zones in a manner that optimizes
  • the first ring comprises 4, 6, 8, 10, 12, 14, 16, 18, or 20 or more identically sized columns.
  • economic efficiency is optimized by maximizing the throughput of the SMB, or by minimizing the desorbent use, for a desired yield and purity of the fast or slow fraction.
  • the method is employed to resolve a multicomponent mixture, and the columns are similarly configured for optimum economics.
  • the number of columns used in an SMB is optimized for a particular column allocation.
  • the number of columns for defined zone lengths can always be increased to improve the efficiency of the system. Therefore, in this context, an optimum design will provide greater than about 85%, 90%, 95%, 91%, 98%, or 99% of the theoretical throughput, or less than 125%, 110%), 105%, 103%), 102%, or 101% of the minimum desorbent consumption, each as defined by the standing wave design, at an infinite column number.
  • any chromatographic process can be used in the methods of the present invention, as long as linear isotherms are observed.
  • the chromatographic method can be based upon differences in adsorption, partition, ion exchange, molecular exclusion, or affinity, within the linear isotherm regime.
  • the method is generally applicable to multicomponent separations, and especially multicomponent separations in which the desired component exhibits an intermediate affinity for the selected adsorbent.
  • some of the methods can also be employed in binary systems (such as the method in which column configuration is optimized).
  • the method is also generally applicable to systems in which non-negligible mass transfer resistances are observed.
  • the chromatographic process of the present invention is size exclusion chromatography (also known as gel permeation or gel filtration).
  • size exclusion chromatography the process lacks any substantial attractive interaction between the stationary phase and component.
  • the liquid or gaseous phase passes through a porous gel which separates the molecules according to size.
  • the pores are normally small and exclude the larger component molecules, but allow smaller molecules to enter the gel, causing them to pass through a larger volume. This causes the larger molecules to pass through the column at a faster rate than the smaller ones.
  • SEC is typically referred to as gel filtration chromatography (GFC)
  • GFC gel filtration chromatography
  • GPC gel permeation chromatography
  • suitable gels include, among others, starch (including maize starch), crosslinked galactomannan, crosslinked dextran, agar or agarose, polyacylamides, copolymers of acylamide and methylene bis-acrylamide, copolymers of methylene bis- acrylamide with vinylethyl carbitol and with vinyl pyrrolidone, and the like.
  • starch including maize starch
  • crosslinked galactomannan crosslinked dextran
  • agar or agarose polyacylamides
  • copolymers of acylamide and methylene bis-acrylamide copolymers of methylene bis- acrylamide with vinylethyl carbitol and with vinyl pyrrolidone
  • a rigid polymer-based gel made from a silica-based or polymer-based material.
  • the preferred gels are crosslinked dextrans, such as the Sephadex series from Pharmacia Fine Chemicals, Inc., Piscataway, NJ.
  • Size exclusion methods can be used to purify various biological macromolecules including proteins and peptides, carbohydrates (i.e. polysaccharides), poly(nucleic acids) (including DNA and RNA fragments, and vaccines), lipids (including fatty acids, triacylglycerols, phospholipids, glycolipids, steroids, and terpenes), and blood plasma components. Size exclusion methods can also be used to separate various synthetic polymers based upon length, molecular weight, or other measure of size.
  • Proteins and peptides that can be purified using the methods of the present invention include insulin, human serum albumin, erythropoeitin, interleukin, interferon, human growth hormone, and bovine growth hormone.
  • the process can be used to purify any protein or peptide that comprises at least 2, 5, 10, or 25 amino acid (or synthetic amino acid) residues.
  • the peptide or protein can be derived from natural sources, made using synthetic chemical techniques, or made from protein expression systems using recombinant DNA techniques.
  • the methods and processes of the invention are practiced using mixtures that do not contain insulin.
  • the size exclusion processes can be used to separate carbohydrates greater than 3, 5, 10, or 20 saccharide units in length.
  • carbohydrate generally refers to a compound of carbon, hydrogen, and oxygen that contains the saccharose unit or its first reaction product and in which the ratio of hydrogen to oxygen is the same as in water.
  • Carbohydrates which can be separated by the present invention may be substituted or deoxygenated at one or more positions, in which case the ratio of hydrogen to oxygen will be different than water.
  • Carbohydrates thus include substituted and unsubstituted oligosaccharides, and polysaccharides.
  • the saccharide units can be an aldose or ketose, and may comprise 3, 4, 5, 6, or 7 carbons.
  • Size exclusion processes can also be used to separate synthetic polymers and copolymers based upon size or molecular weight.
  • exemplary polymers include polyolefins (such as polyethylene, poly(ethylene glycol), and polypropylene), polystyrene, poly(methyl methacrylate), and poly(vinyl alcohol). Generally, the polymer will comprise greater than 20, 50, or 100 monomeric units.
  • the method also has been found particularly useful in the purification of sugars from extraction media (especially acids) used in the sugar processing industry, using cation exchange resins such as the Dowex resins produced by Dow Chemical, made of sulfonated polystyrene.
  • cation exchange resins such as the Dowex resins produced by Dow Chemical, made of sulfonated polystyrene.
  • a preferred resin for separating glucose and xylose from sulfuric acid and acetic acid is Dowex 99, made from sulfonated polystyrene with approximately 6% divinylbenzene or a crosslinker.
  • a preferred particle size is about 320 um.
  • the process can be used generally to recover sugar from any acidic extraction medium, including acetic acid, citric acid, sulfuric acid, carbonic acid, hydrochloric acid, phosphoric, or sulfuric acid.
  • the invention can be used to purify any sugar, using a cation exchange resin under linear isotherms.
  • sucrose includes the monosaccharides glyceraldehyde, erythrose, threose, ribose, arabinose xylose, lyxose, allose, altrose, glucose, mannose, gulose, idose, talose, galactose, psicose, fructose, sorbose, tagatose, ribulose, xylulose, erythrulose, dihydroxyacetone, and oligosaccharides based thereon, including but not limited to sucrose, trehalose, maltose, cellobiose, gentiobiose, lactose, and raffinose.
  • Another preferred application of the methods of the present invention is in the purification of synthetic or naturally occurring pharmaceutical compounds. These compounds are particularly well suited for SMB purification because of their tremendous expense, and the ability of SMB purifications to deliver higher yields than batch chromatography.
  • the SMBs are used to purify paclitaxel from a mixture of components such as cephalomannin, using a selective adsorbent (preferably a high capacity polystyrene divinyl-benzene adsorbent).
  • a selective adsorbent preferably a high capacity polystyrene divinyl-benzene adsorbent.
  • a preferred solvent for such a system is ethanol, and its concentration preferably is between 30% and 100%.
  • the SMBs are used in hydrophobic adsorption processes, or in ion exchange processes at low concentrations.
  • Insulin feed used in the single column and the SMB experiments was provided in a broth containing HMWP/HPI, insulin, ZnCl 2 , and water.
  • Glacial acetic acid was purchased from Mallinckrodt Baker Inc. (Paris, KY) and pure ethanol from Pharmaco Products Inc. (Brookfield, CT).
  • Distilled deionized water (DDW) was obtained through a Milli-Q system by Millipore (Bedford, MA).
  • HPLC-grade acetonitrile was purchased from Fisher Scientific (Fairlawn, NJ).
  • Sulfate buffer was prepared from sodium sulfate and 85% phosphoric acid, which were purchased from Mallinckrodt Baker Inc. Sodium chloride and zinc chloride were also purchased from Mallinckrodt Baker Inc.
  • Blue dextran used in bed porosity determination was purchased from Sigma Chemical Co. (St. Louis, MO).
  • the gel particles have an average radius of 54 ⁇ m .
  • All the columns and fittings were purchased from Ace Glass Inc. (Louisville, KY). The columns are 5.10 cm in diameter and 13.7 cm in packing length.
  • a Vydac C-18 HPLC column (25cm x 4.6 mm) for protein and peptide analysis was purchased from Vydac Co. (Hesperia, CA) and used to determine the concentration of BHI (Biosynthetic Human Insulin).
  • a Waters insulin HMWP HPLC column (30cm x 7.8 mm) was purchased from Millipore Corp. (Milford, MA) and used to determine the concentration of HMWP.
  • Equipment The HPLC system consists of two pumps (Waters 510), a tunable single-wavelength detector (Waters 486) and an injector (Waters U6K).
  • a Pharmacia (Piscataway, NJ) fast protein liquid chromatography (FPLC) system was used in the single column experiments.
  • This system consists of two pumps (Pharmacia P-500), a liquid chromatography controller (Pharmacia LCC-500), and an injection valve (Pharmacia MV- 7).
  • a lab-scale SMB unit was manufactured by Advanced Separation Technology (Lakeland, FL). It consists of a controller for the adjustment of the switching time and a frame that supports a rotation gear, a drive assembly, and a column rack.
  • Two single-piston pumps (Model RHV) purchased from Fluid Metering Inc. (Syosset, NY) were used to control the flow rates for Zone E and Zone IV.
  • the flow rates of the feed and desorbent were controlled by two FPLC pumps. The flow rates were reproducible and controlled to within + 1%.
  • the HETP (Height Equivalent to Theoretical Plate) test was performed in order to check the column efficiency for each column.
  • a 2-mL pulse of sodium chloride (5g/L in IN acetic acid) was injected into the column, followed by elution with IN acetic acid at the maximum flow rate. Since the conductivity detector has a relatively high flow resistance, the column effluent was split into two streams in a ratio of 8 to 1. The split is needed to keep the pressure drop in the column below 100 psi. The smaller stream was passed through the conductivity detector.
  • N Number of Theoretical Plates
  • a 2-mL loop was connected to the injection valve (Pharmacia MN-7). In the load position, the loop was filled with a feed solution. The eluent flow rate was controlled by a controller (Pharmacia LCC-500). Then the injection valve was switched to the inject position by the controller in order to start the injection. Data recording was started simultaneously.
  • Three pulse tests were carried out: ⁇ aCl, ZnCl 2 and BHI. The concentrations of BHI were detected by the UV detector at 294 nm. All pulse tests were performed at a flow rate of 6 mlJmin.
  • SMB experiments The 10 columns in the SMB were packed with Sephadex G50 resin using the aforementioned method.
  • the column configuration was either 2-3-3-2 (there were 2, 3, 3, and 2 columns in Zones I, E, El, and IN, respectively) or 2-2-4-2.
  • the extra- column dead volume was determined by measuring the amount of water that filled in the tubing and rotary valve.
  • the flow rates of the four pumps were set manually and were checked later periodically during the experiment to ensure accuracy.
  • the switching time was set using the SMB controller.
  • the experiment was started by turning on the pumps simultaneously. Feed and desorbent were continuously pumped into the columns.
  • the feed for Ring I was prepared by dissolving crude insulin powder (containing ZnCl 2 ) in an HMWP broth.
  • the feed for Ring II was a binary mixture of BHI and ZnCl 2 , which was made by dissolving the crude insulin powder in 1 ⁇ acetic acid.
  • the desorbent was 1 ⁇ acetic acid.
  • Samples were collected from the extract port and the raffinate port over an entire switching period (between two switches). Therefore, the concentration of each sample represents the average concentration over a switching period.
  • the Ring I experiments were stopped in the middle of the 45 th step and the Ring E experiment was stopped in the middle of the 61 st step. Zone profiles were then obtained by collecting samples at the sampling ports, which were located at each column outlet.
  • the weight refers to the packing weight for an intermediate packing volume of 304 mL, which is larger than the final bed volume (280 mL).
  • Pulse tests The chromatograms of the pulse tests are shown in Figure 9. The elution volume of each pulse is estimated from the mass center of the chromatogram and is used to determine the column properties and size exclusion factors. Note that the extra- column dead volume is subtracted to obtain the net elution volume through the column.
  • the elution volume of a blue dextran pulse is the inter-particle volume or void volume (Vo), because the blue dextran molecule is much larger than the average resin pore size and is totally excluded from the particle.
  • the bed voidage, ⁇ %, is calculated from the void volume as follows:
  • RV the bed volume (mL).
  • the NaCl molecule is smaller than the molecule of HPI, BHI, or ZnCl 2 , and is chosen as the tracer to estimate intra-particle volume.
  • the total void volume (V t ) including inter-particle and intra-particle volume is obtained from the NaCl pulse. Combined with the void volume, the total void volume can be used to calculate the particle porosity, £ p .
  • the size exclusion factors of BHI and ZnCl 2 can be estimated as follows:
  • Ke estimated from the lab-scale column with a small pulse, in which the peak concentration is very low, is close to intrinsic Re.
  • the Re for the plant-scale column is an apparent Re, which is concentration-dependent.
  • Figure 10 shows the chromatograms of the saturation and elution tests at three different flow rates and the results of the simulations (dashed lines) with the intrinsic Re estimated from the small pulse test (Table 5).
  • the simulated chromatograms are ahead of the data for both loading and elution processes at all three flow rates.
  • a multiple frontal test was first conducted without the column. This test is to determine the extra-column dead volume and to obtain the calibration curve of UV absorbance as a function of concentration.
  • the chromatogram shown in Figure 11a is used to estimate the dead volume present in the mixer and the connecting tubing. The dead time of each frontal obtained without the column is subtracted from the breakthrough time of the corresponding frontal obtained with the column.
  • the chromatogram of the multiple frontal test with the column is shown in Figure 1 lb.
  • the anti-Langmuir isotherm parameters were used in a simulation to predict the multiple frontal test data. Good agreement between the model prediction and data is shown in Figure 1 lb. Since the curvature of the isotherm is quite small, the isotherm can be well represented by a linear adsorption isotherm or an apparent Ke, which has a higher value than the intrinsic Re. For this reason, the multiple frontal test data can be simulated with an apparent Re. The simulated chromatograms with adsorption (anti-Langmuir) and with the apparent Re are almost identical (Figure lib). The value of the apparent Re applied in the simulations is 0.74 (>0.68, the value of the intrinsic Re of BFE).
  • the maximum interstitial velocity for the Sephadex G50 resin cannot exceed 1.28 cm/min.
  • the maximum zone flow rate is estimated to be 9 mL/min. This flow rate limit was applied in the design of both Ring I and Ring II with the lab-scale columns. Since the same size columns were used in both Ring I and Ring II experiments, the extract flow rate of Ring I is not matched by the feed flow rate of Ring E as shown in Table 7. This mismatch, which is considered in calculating the overall throughput, however, does not affect the validation of the tandem SMB design and the model.
  • the SMB experiment Run 1 (Ring I) lasted for 21 hours or 45 cycles.
  • the simulation and experimental data of the effluent histories at the raffinate and extract ports and the mid-cycle column profiles at the final cycle are shown in Figure 12. Close agreement between the model prediction and the experimental data indicates that the column properties and size exclusion factors used in the SMB designs are accurate and the design method is robust, even though the simulations do not accurately predict the data of the batch frontal processes.
  • the fronting of BEE in Zone El of Ring I can be simulated by using a larger axial dispersion coefficient (E b ).
  • E b axial dispersion coefficient
  • Figure 13a the loss of BHI in the raffinate increases
  • Figure 13c the adsorption wave of BFfl in Zone ⁇ I spreads out
  • Figure 13 shows that the breakthrough curve of BHI in the raffinate can be well predicted when E value is 40 times as large as that estimated from the Chung and Wen correlation (1968).
  • the fronting should be considered and the fitted E b value should be used in the standing wave design.
  • Table 9 shows the results of Ring I design, in which the fronting of BHI is considered by using the fitted E b (40 times as large as that estimated from the Chung and Wen correlation).
  • Different zone configuration was investigated in the standing wave design. To reduce the difference between the true moving bed and the simulated moving bed, at least 2 columns are needed for each zone. For some zone configurations, there are no feasible operating conditions that can meet the yield requirement (99.99% for BHI). Among the three zone configurations that can meet the high yield requirement, 2-2-4-2 gives the lowest solvent consumption (desorbent to feed ratio) as shown in Table 9. To verify the design results, N ⁇ RS ⁇ simulations were conducted. The simulation results are listed in Table 10.
  • Zone Zone flow rate (mL/min) Switching Desorbent configuration E El IV time /Feed (min)
  • Zone flow rates are less than 8.5 mL/min, which is 5% lower than the theoretical maximum value.
  • E F is the feed flow rate (mL/min)
  • F is the extract flow rate (mlJmin)
  • BV stands for the bed volume (mL) of a single column in the experiments.
  • Eo is the desorbent flow rate (mlJmin)
  • F R is the raffinate flow rate (mL/min)
  • subscript " ⁇ xt” stands for extract and "Raff for raffinate.
  • the unit of the BHI concentration is g/L.
  • V c (l- ⁇ b )(l- ⁇ p ) where ⁇ , is the partition coefficient in ml/ml solid volume (SN.), F is the volumetric flow rate of the feed, t garnish- is the solute retention time (first moment), and t p is the time of the pulse injection.
  • the term to ( - is the touring time of solute i (neglecting adsorption), and it is calculated as
  • Table 11 lists the partition coefficients for the four components in this system. Solid lines in
  • Standing wave designs To compare designs for different column configurations and different splitting strategies, it is necessary to define a set of standard operating parameters.
  • two 20-column SMB systems in series are used to purify the center cut (which in this case consists of the two sugars glucose and xylose).
  • the feed to the first SMB has a flow rate of 33.3 ml/min and a composition given in Table 13.
  • the feed flow rate to the second SMB is taken as the flow rate of that product from the first SMB (raffinate or extract) that contains the unresolved components.
  • the feed to the second SMB in each pathway is taken as the time-averaged composition at steady state; in reality, the concentration of each component changes with time even when the system has reached a cyclic steady-state.
  • the yield is set to 0.90 for each component at each SMB step for the cases in which complete resolution of all components is desired; the yield is set to 0.50 for components that are allowed to distribute among the two product ports.
  • the yield of 0.90 is approximately the value that was obtained from the SMB experiment for parameter validation ( Figure 18).
  • Figures 19 and 21 show column profiles for two strategies for isolating glucose and xylose from a feed mixture that contains both sulfuric acid and acetic acid, via complete separation of the two sugars from one of the acids in each step.
  • the numerical parameters used in the simulations are listed in Table 14.
  • the operating parameters and product compositions are listed in Table 15.
  • the column configuration for each ring in each strategy is determined by doing a computer search of the standing wave designs of all possible configurations that satisfy the desired purity and yield, and selecting the configuration for which the standing wave design requires the lowest desorbent consumption.
  • the column profiles are taken after 400 switch times; at this point, the concentration of each component has nearly approached a limiting value.
  • the mass-balance is checked for each component over one switch interval; for each component, if the difference between the mass introduced during the switch interval and the mass leaving the system through the two product ports is less than 0.1%, then the system is defined here to be at cyclic steady-state. (This same criterion is used in the subsequent experiments.)
  • Figures 21 and 23 show column profiles for two strategies for isolating the two sugars via complete separation from one acid and partial separation from the other in the first step and complete separation from the other acid in the second step.
  • the operating parameters and product compositions are listed in Table 16.
  • Tables 16 and 17 show the comparison between the mass-balance estimates of product concentrations and the concentrations obtained after detailed simulations based on the given operating conditions.
  • the mass-balance approach is generally conservative in that the desired product concentrations are underestimated by 3 to 15% while the impurity concentrations are overestimated up to fifty fold.
  • Figures. 20 to 21 that within a given zone some waves are not standing but are traveling either faster or slower than the selected standing wave.
  • the adsorption waves of xylose and acetic acid downstream from the feed port do not migrate as far through that zone as glucose (which is the standing wave there).
  • Xylose and acetic acid are effectively "pinched" toward the feed port because the migration velocity of these solutes is relatively slow in this zone.
  • Figure 7 illustrates the strategies without mass-transfer resistance
  • Figures 20 to 24 illustrate the strategies with mass-transfer resistance.
  • the feed flow rate to the first SMB is 33.3 ml/min.
  • the desorbent port is treated as a mixing junction between zones I and IV and the desorbent inlet.
  • Yi the desired product yield of each component in the mixture. This can be denoted by Yi for each component .
  • Figure 23 shows schematic column profiles for the standing components in a four-zone SMB. For each component that is desired to be recovered in the raffinate,
  • Equations 62 to 65 can be applied to solve for the concentration of each component at each port, which in turn can be used to calculate the decay coefficient for each zone:
  • the solution of the four zone flow rates and the switching time (Eqs. 3 to 6 and 21) is iterative.
  • the initial guess of the zone flow rates and switching time come from the equilibrium equations (Eqs. 17 to 20). These parameters are used to determine mass- transfer parameters for each component in each zone, which are in turn used to re-calculate the zone flow rates and switching time. The calculation and re-substitution is continued until the flow rates and switching time change by less than 0.01% at each step.
  • Nomenclature ⁇ ( - linear equilibrium distribution coefficient of component , ml/ml S.N.
  • C ⁇ mobile-phase concentration of component i (a function of z and t), mg/ml
  • P i pore-phase concentration of component i (a function of r and t), mg/ml
  • ⁇ ,m ⁇ n minimum time-averaged concentration of component i in zone m, mg/ml
  • EA extract flow rate from ring A in a two-ring SMB, ml/min
  • F A feed flow rate to ring A in a two-ring SMB, ml/min
  • E ⁇ feed flow rate to ring B in a two-ring SMB
  • ml/min j component index kff - film mass-transfer coefficient of component i in zone m, cm/min
  • Kfl lumped mass-transfer coefficient of component i in zone m, min D1
  • N numbering index of highest-affinity component
  • N c m number of columns in zone m
  • V c total column volume (particles + voids), ml
  • A" 1 mass-transfer correction term for standing component i in zone m
  • ⁇ intraparticle void fraction
  • ⁇ p interparticle void fraction
  • v solid movement velocity, cm/min
  • VA solid movement velocity for ring A in a two-ring SMB
  • cm/min solid movement velocity for ring B in a two-ring SMB

Landscapes

  • Chemical & Material Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Sustainable Development (AREA)
  • Analytical Chemistry (AREA)
  • Chemical Kinetics & Catalysis (AREA)
  • Treatment Of Liquids With Adsorbents In General (AREA)

Abstract

L'invention concerne des séparations d'un ou plusieurs composants issus de mélanges de fluides à composants multiples, au moyen de la technologie du lit mobile simulé, ainsi que des procédés permettant de concevoir de tels lits (SMB). Ces lits, conçus selon la présente invention, peuvent être utilisés pour séparer des fractions mobiles lentes et rapides dans des mélanges à composants multiples ou pour séparer un composé à affinité intermédiaire du mélange. L'invention peut être appliquée à des systèmes présentant des isothermes linéaires et elle est particulièrement appliquée à ces systèmes quand des séparations sont inhibées par les résistances de transfert de masse.
PCT/US2001/015848 2000-05-16 2001-05-16 Conception d'ondes stationnaires de lits mobiles simules uniques ou tandem destinee a la resolution de melanges a composants multiples WO2001087452A2 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
AU2001263192A AU2001263192A1 (en) 2000-05-16 2001-05-16 Standing wave design of single and tandem simulated moving beds for resolving multicomponent mixtures

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
US20470000P 2000-05-16 2000-05-16
US20470100P 2000-05-16 2000-05-16
US60/204,700 2000-05-16
US60/204,701 2000-05-16

Publications (2)

Publication Number Publication Date
WO2001087452A2 true WO2001087452A2 (fr) 2001-11-22
WO2001087452A3 WO2001087452A3 (fr) 2003-07-17

Family

ID=26899729

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2001/015848 WO2001087452A2 (fr) 2000-05-16 2001-05-16 Conception d'ondes stationnaires de lits mobiles simules uniques ou tandem destinee a la resolution de melanges a composants multiples

Country Status (2)

Country Link
AU (1) AU2001263192A1 (fr)
WO (1) WO2001087452A2 (fr)

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
FR2843893A1 (fr) * 2002-08-28 2004-03-05 Inst Francais Du Petrole Methode pour optimiser le fonctionnement d'une unite de separation de xylenes par contre courant simule
US7141172B2 (en) 2001-09-27 2006-11-28 Purdue Research Foundation Versatile simulated moving bed systems
WO2013005048A1 (fr) * 2011-07-06 2013-01-10 Basf Pharma (Callanish) Limited Procédé smb pour produire epa très pur à partir d'huile de poisson
WO2013005052A1 (fr) * 2011-07-06 2013-01-10 Basf Pharma (Callanish) Limited Procédé smb amélioré
WO2013005051A1 (fr) * 2011-07-06 2013-01-10 Basf Pharma (Callanish) Limited Nouveau procédé smb
JP2013516398A (ja) * 2009-12-30 2013-05-13 ビーエイエスエフ ファーマ（コーラニッシュ）リミテッド 擬似移動床式クロマトグラフ分離方法
JP2014518312A (ja) * 2011-07-06 2014-07-28 ビーエイエスエフ ファーマ（コーラニッシュ）リミテッド Ｓｍｂ方法
US9150816B2 (en) 2013-12-11 2015-10-06 Novasep Process Sas Chromatographic method for the production of polyunsaturated fatty acids
US9428711B2 (en) 2013-05-07 2016-08-30 Groupe Novasep Chromatographic process for the production of highly purified polyunsaturated fatty acids
US9694302B2 (en) 2013-01-09 2017-07-04 Basf Pharma (Callanish) Limited Multi-step separation process
US9771542B2 (en) 2011-07-06 2017-09-26 Basf Pharma Callanish Ltd. Heated chromatographic separation process
US10975031B2 (en) 2014-01-07 2021-04-13 Novasep Process Method for purifying aromatic amino acids
WO2021226072A1 (fr) * 2020-05-05 2021-11-11 Amalgamated Research Llc Systèmes comprenant des séparateurs à lit mobile simulé pour la production de fructose de haute pureté et procédés associés

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0875268A1 (fr) * 1997-04-30 1998-11-04 Institut Francais Du Petrole Procédé pour contrÔler un système de séparation de constituants à lits mobiles simulés
GB2326357A (en) * 1997-06-19 1998-12-23 Inst Francais Du Petrole Optimum operation of loops in a moving bed

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0875268A1 (fr) * 1997-04-30 1998-11-04 Institut Francais Du Petrole Procédé pour contrÔler un système de séparation de constituants à lits mobiles simulés
GB2326357A (en) * 1997-06-19 1998-12-23 Inst Francais Du Petrole Optimum operation of loops in a moving bed

Non-Patent Citations (9)

* Cited by examiner, † Cited by third party
Title
BIRESSI G ET AL: "Design and optimisation of a simulated moving bed unit: role of deviations from equilibrium theory" JOURNAL OF CHROMATOGRAPHY, ELSEVIER SCIENCE PUBLISHERS B.V. AMSTERDAM, NL, vol. 876, no. 1-2, April 2000 (2000-04), pages 3-15, XP004196981 ISSN: 0021-9673 *
G. D]NNEBIER ET AL.: "Modelling and simulation of nonlienear chromatographic seapration processes: a comparision of different modelling approaches" CHEMICAL ENGINEERING SCIENCE, vol. 55, 23 August 1998 (1998-08-23), pages 373-380, XP001148498 Pergamon, Elsevier Science Ltd. *
MA Z ET AL: "STANDING WAVE ANALYSIS OF SMB CHROMATOGRAPHY: LINEAR SYSTEMS" AICHE JOURNAL, NEW YORK, NY, US, vol. 43, no. 10, October 1997 (1997-10), pages 2488-2508, XP008015141 ISSN: 0001-1541 cited in the application *
MAZZOTTI M ET AL: "Optimal operation of simulated moving bed units for nonlinear chromatographic separations" JOURNAL OF CHROMATOGRAPHY A, ELSEVIER SCIENCE, NL, vol. 769, no. 1, 2 May 1997 (1997-05-02), pages 3-24, XP004064202 ISSN: 0021-9673 *
PAIS L S ET AL: "Modeling, simulation and operation of a simulated moving bed for continuous chromatographic separation of 1,1Ð?-bi-2-naphthol enantiomers" JOURNAL OF CHROMATOGRAPHY A, ELSEVIER SCIENCE, NL, vol. 769, no. 1, 2 May 1997 (1997-05-02), pages 25-35, XP004064203 ISSN: 0021-9673 *
PROLL T ET AL: "Optimization strategy for simulated moving bed systems" JOURNAL OF CHROMATOGRAPHY A, ELSEVIER SCIENCE, NL, vol. 800, no. 2, 27 March 1998 (1998-03-27), pages 135-150, XP004113589 ISSN: 0021-9673 *
WU D-J ET AL: "DESIGN OF SIMULATED MOVING BED CHROMATOGRAPHY FOR AMINO ACID SEPARATIONS" INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH, AMERICAN CHEMICAL SOCIETY. WASHINGTON, US, vol. 37, no. 10, October 1998 (1998-10), pages 4023-4035, XP001149344 ISSN: 0888-5885 *
WU D-J ET AL: "Optimization of throughput and desorbent consumption in simulated moving-bed chromatography for paclitaxel purification" JOURNAL OF CHROMATOGRAPHY A, ELSEVIER SCIENCE, NL, vol. 855, no. 1, 3 September 1999 (1999-09-03), pages 71-89, XP004180016 ISSN: 0021-9673 *
XIE Y ET AL: "EXTENDED STANDING WAVE DESIGN METHOD FOR SIMULATED MOVING BED CHROMATOGRAPHY: LINEAR SYSTEMS" INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH, AMERICAN CHEMICAL SOCIETY. WASHINGTON, US, vol. 39, no. 6, 2000, pages 1993-2005, XP001149345 ISSN: 0888-5885 *

Cited By (35)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7141172B2 (en) 2001-09-27 2006-11-28 Purdue Research Foundation Versatile simulated moving bed systems
WO2004020068A2 (fr) * 2002-08-28 2004-03-11 Institut Francais Du Petrole Methode pour optimiser le fonctionnement d'une unite de separation de xylenes par contre courant simule
WO2004020068A3 (fr) * 2002-08-28 2004-04-08 Inst Francais Du Petrole Methode pour optimiser le fonctionnement d'une unite de separation de xylenes par contre courant simule
JP2005536344A (ja) * 2002-08-28 2005-12-02 アンスティテュ　フランセ　デュ　ペトロール シミュレートされた向流を用いたキシレン分離装置の運転を最適化する方法
US7192526B2 (en) 2002-08-28 2007-03-20 Institut Francais Du Petrole Method of optimizing the operation of a xylene separation unit using simulated countercurrent
KR101047081B1 (ko) * 2002-08-28 2011-07-06 아이에프피 에너지스 누벨 모의 역류를 사용하여 크실렌 분리 유닛의 조작을 최적화하는 방법
FR2843893A1 (fr) * 2002-08-28 2004-03-05 Inst Francais Du Petrole Methode pour optimiser le fonctionnement d'une unite de separation de xylenes par contre courant simule
JP2013516398A (ja) * 2009-12-30 2013-05-13 ビーエイエスエフ ファーマ（コーラニッシュ）リミテッド 擬似移動床式クロマトグラフ分離方法
US9790162B2 (en) 2009-12-30 2017-10-17 Basf Pharma (Callanish) Limited Simulated moving bed chromatographic separation process
JP2014518313A (ja) * 2011-07-06 2014-07-28 ビーエイエスエフ ファーマ（コーラニッシュ）リミテッド 魚油から高純度のｅｐａを生成するためのｓｍｂ方法
US9771542B2 (en) 2011-07-06 2017-09-26 Basf Pharma Callanish Ltd. Heated chromatographic separation process
KR20140034924A (ko) * 2011-07-06 2014-03-20 바스프 파마 (칼라니쉬) 리미티드 개선된 smb 공정
CN103764241A (zh) * 2011-07-06 2014-04-30 巴斯夫制药（卡兰尼什）公司 改进的smb方法
CN103826714A (zh) * 2011-07-06 2014-05-28 巴斯夫制药（卡兰尼什）公司 用于从鱼油中生产高纯度epa的smb方法
US20140200360A1 (en) * 2011-07-06 2014-07-17 Basf Pharma (Callanish) Limited SMB Process
JP2014518312A (ja) * 2011-07-06 2014-07-28 ビーエイエスエフ ファーマ（コーラニッシュ）リミテッド Ｓｍｂ方法
WO2013005052A1 (fr) * 2011-07-06 2013-01-10 Basf Pharma (Callanish) Limited Procédé smb amélioré
JP2014523944A (ja) * 2011-07-06 2014-09-18 ビーエイエスエフ ファーマ（コーラニッシュ）リミテッド 新規なｓｍｂ方法
JP2014525951A (ja) * 2011-07-06 2014-10-02 ビーエイエスエフ ファーマ（コーラニッシュ）リミテッド 改善されたｓｍｂ方法
AU2012279994B2 (en) * 2011-07-06 2015-07-16 Basf Pharma (Callanish) Limited Improved SMB process
AU2012279993B2 (en) * 2011-07-06 2015-07-23 Basf Pharma (Callanish) Limited New SMB process
CN103764241B (zh) * 2011-07-06 2015-08-19 巴斯夫制药（卡兰尼什）公司 改进的smb方法
WO2013005048A1 (fr) * 2011-07-06 2013-01-10 Basf Pharma (Callanish) Limited Procédé smb pour produire epa très pur à partir d'huile de poisson
US9234157B2 (en) * 2011-07-06 2016-01-12 Basf Pharma Callanish Limited SMB process
KR101597923B1 (ko) 2011-07-06 2016-03-07 바스프 파마 (칼라니쉬) 리미티드 개선된 smb 공정
WO2013005051A1 (fr) * 2011-07-06 2013-01-10 Basf Pharma (Callanish) Limited Nouveau procédé smb
US9695382B2 (en) 2011-07-06 2017-07-04 Basf Pharma (Callanish) Limited SMB process for producing highly pure EPA from fish oil
US9694302B2 (en) 2013-01-09 2017-07-04 Basf Pharma (Callanish) Limited Multi-step separation process
US10179759B2 (en) 2013-01-09 2019-01-15 Basf Pharma (Callanish) Limited Multi-step separation process
US10214475B2 (en) 2013-01-09 2019-02-26 Basf Pharma (Callanish) Limited Multi-step separation process
US10723973B2 (en) 2013-01-09 2020-07-28 Basf Pharma (Callanish) Limited Multi-step separation process
US9428711B2 (en) 2013-05-07 2016-08-30 Groupe Novasep Chromatographic process for the production of highly purified polyunsaturated fatty acids
US9150816B2 (en) 2013-12-11 2015-10-06 Novasep Process Sas Chromatographic method for the production of polyunsaturated fatty acids
US10975031B2 (en) 2014-01-07 2021-04-13 Novasep Process Method for purifying aromatic amino acids
WO2021226072A1 (fr) * 2020-05-05 2021-11-11 Amalgamated Research Llc Systèmes comprenant des séparateurs à lit mobile simulé pour la production de fructose de haute pureté et procédés associés

Also Published As

Publication number Publication date
AU2001263192A1 (en) 2001-11-26
WO2001087452A3 (fr) 2003-07-17

Similar Documents

Publication Publication Date Title
EP1877769B1 (fr) Procede et dispositif de purification chromatographique
Aumann et al. A continuous multicolumn countercurrent solvent gradient purification (MCSGP) process
Hritzko et al. Standing‐wave design of tandem SMB for linear multicomponent systems
CA2498725C (fr) Procede de purification de polypeptides par chromatographie sur lit mobile simule
Aniceto et al. Simulated moving bed strategies and designs: from established systems to the latest developments
WO2001087452A2 (fr) Conception d'ondes stationnaires de lits mobiles simules uniques ou tandem destinee a la resolution de melanges a composants multiples
US20110168632A1 (en) Method and device for separating fractions of a mixture
Imamoglu Simulated moving bed chromatography (SMB) for application in bioseparation
EP1349866B1 (fr) Purification de l'insuline selon une technologie a lit mobile simule
Nowak et al. Theoretical study of using simulated moving bed chromatography to separate intermediately eluting target compounds
US20030216543A1 (en) Insulin purification using simulated moving bed technology
Krättli et al. Multifraction separation in countercurrent chromatography (MCSGP)
JP2014169997A (ja) 濃縮及び単離のためのクロマトグラフィープロセス
CN102271775A (zh) 多肽的逆流提纯
US11180526B2 (en) Purification method and uses thereof
Moreno-González et al. Continuous adsorption in food industry: The recovery of sinapic acid from rapeseed meal extract
WO2001087451A2 (fr) Conception d'onde stationnaire d'un lit mobile simule (smb) a neuf zones destine a la recuperation d'un solute avec une affinite intermediaire dans un melange ternaire
Aniceto et al. Preparative chromatography: batch and continuous
Huang et al. Enzyme purification and concentration by simulated moving bed chromatography: an experimental study
JP2023511283A (ja) クロマトグラフィを用いるコーンシロップからのアルロースの３成分分離
Xie et al. Preparative chromatographic separation: simulated moving bed and modified chromatography methods
Ottens et al. Multi-component fractionation in SMB chromatography for the purification of active fractions from protein hydrolysates
Fioretti et al. Continuous countercurrent chromatography for the downstream processing of bioproducts: A focus on flow-through technologies
KR20240065316A (ko) 정제 방법 및 그의 사용
Fioretti et al. Design of a twin-column countercurrent purification (MCSGP) unit for the polishing of an oligonucleotide sequence

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A2

Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EC EE ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NO NZ PL PT RO RU SD SE SG SI SK SL TJ TM TR TT TZ UA UG US UZ VN YU ZA ZW

AL Designated countries for regional patents

Kind code of ref document: A2

Designated state(s): GH GM KE LS MW MZ SD SL SZ TZ UG ZW AM AZ BY KG KZ MD RU TJ TM AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE TR BF BJ CF CG CI CM GA GN GW ML MR NE SN TD TG

121 Ep: the epo has been informed by wipo that ep was designated in this application
DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
122 Ep: pct application non-entry in european phase
NENP Non-entry into the national phase in:

Ref country code: JP