Description
Title of Invention: AN APPARATUS FOR CONTINUOUS
SEPARATION OF VALINE AND A METHOD FOR
CONTINUOUS SEPARATION OF VALINE USING THE SAME
Technical Field
[1] The present invention relates to an apparatus for continuous separation of valine and
a method for continuous separation of valine using the same, and more specifically to
an apparatus for continuous separation of valine from a mixture comprising amino
acids such as leucine, isoleucine, etc. and to a method for continuous separation of
valine using the same.
Background Art
[2] Valine is an L-amino acid having a formula of H0 2CCH(NH2)CH(CH3)2 and used as
one of the main raw materials in medicine and in cosmetics. It is also one of the
important ingredients utilized for animal feed. For this reason, there is a daily increase
of interest within commercial markets focusing on the applications of valine.
[3] The production of valine is generally achieved by a procedure of fermentation of
Corynebacterium. In this regard, a matter of note is that impurities are also obtained
along with valine during the fermentation procedure. These impurities include salt,
alanine, leucine, isoleucine, etc., and all of these impurities are required to be separated
from valine.
[4] Separation methods for valine used in the prior art include an ion exclusionchromatography
method (Japanese Laid-open Patent Publication No. 1987-255453), a
crystallization method using a precipitant (Japanese Laid-open Patent Publication No.
1996-333312, Japanese Laid-open Patent Publication No. 1998-237030, and U.S.
Patent No. 6,072,083), a chemical reaction method (U.S. Patent No. 4,263,450), and
the like. However, among the separation methods, the ion exclusion-chromatography
method has disadvantages in that the separation of amino acids such as leucine and
isoleucine is difficult, a lot of waste water is generated, and a post-treatment process
such as crystallization, etc. is additionally needed. Also, since the crystallization
method using the precipitant needs a process for removing the precipitant, it has the
disadvantage that the process becomes complex and that a post-treatment process for
purification is additionally needed. Further, since the chemical reaction method needs a
concentration and hydrolysis process, it has the disadvantage that the process is
complex and utilizes lot of solvents, and thus a number of labs are needed in the posttreatment
process for recovering it, and the costs thereof in the purification process are
accordingly increased.
[5] Meanwhile, the chromatographic separation process is a separation process based on
an adsorbent and is widely used in processes for the separation and purification of a
plurality of bioproducts, and according to the application is broadly divided into a
batch chromatography process and a continuous Simulated Moving Bed (SMB) chro
matography process. The SMB process was initially developed for the separation of
petrochemicals by UOP, USA in 1961, and it has been reported that its performance
and separation efficiency is far superior to that of the batch process. Due to the su
periority of the SMB process, it has been expanded to the separation and purification
of high- value added products such as sugar materials, chiral compounds, bioproducts,
medicines, etc. as well as the separation and purification of petrochemicals.
[6] Most conventional SMB processes (four-zone SMB) are composed of a structure
having four zones (columns) and four ports, as shown in Fig. 1, and each of the
mixtures to be separated and the solvent (Desorbent) flow into a port (Feed) for the
mixture to be separated and a desorption port (Desorbent) among the four ports. In
addition, materials having weak adsorptive power and strong adsorptive power are
separated and then obtained via Raffinate port (Raffinate) and Extract port (Extract). In
order for the processes corresponding to the input of the mixture and the solvent to be
separated and the recovery of the two components above to be continuously
performed, the four ports are moved along the direction in which the solvent flows at a
constant rate of switching.
[7] However, the four-zone SMB process needs at least four columns. Generally, since
both the valves of each column and the adsorbent are expensive, it is preferable to
minimize the number of the columns if possible.
[8] For this reason, a SMB process having three chromatography zones (three-zone
SMB) as shown in Fig. 2 has been introduced. Since it uses three zones, the minimum
number of columns can be reduced from four to three. In addition, the positions of the
four ports are disposed in a manner identical in terms of structure relative to that of the
conventional SMB structure having four chromatography zones.
[9] However, the low-affinity component having weak adsorptive power and thus
moving fast, such as valine in the above three-zone SMB process, is recovered via the
Raffinate port as shown in Fig. 2 but, since there is no enrichment zone for the
raffinate as in Fig. 2, excessive dilution is unavoidable. That is, since the concentration
of valine recovered is lower than the valine concentration in the mixture to be
separated, a side effect occurs in that operation cost for the post-treatment process
following the SMB separation process is increased. Therefore, the application of the
three-zone SMB process found in the new system is needed, in that it is able to
overcome the side effect.
[10] Accordingly, during research in consideration of the above-mentioned matters, the
present inventors have ascertained that valine can be continuously and efficiently
separated from a mixture comprising amino acids such as leucine, isoleucine, etc.,
through an apparatus using a simulated moving bed chromatography process, thereby
completing the present invention.
Disclosure of Invention
Technical Problem
An object of the present invention is to provide the apparatus for continuous
separation of valine from the mixture comprising amino acids such as leucine,
isoleucine, etc. and the method for continuous separation of valine using the same.
Solution to Problem
In order to resolve the above problem, the apparatus used for the separation of valine
in the present invention includes a Desorbent port (D), a Feed port (F), a Raffinate port
(R), an Extract port (E), three rotary valves (10, 20, 30), and three chromatography
zones (40, 50, 60) connected to each of three rotary valves (10, 20, 30), as shown in
Figs. 4a - 4c.
Three rotary valves (10, 20, 30) are equipped with three connection ports (10a, 10b,
10c)(20a, 20b, 20c)(30a, 30b, 30c), respectively, and only any one connection port of
each rotary valve (10, 20, 30) is opened along with the rotation of the rotary valves
(10, 20, 30), and is in fluid communication with the Desorbent port (D), the Feed port
(F), the Raffinate port (R) and the Extract port (E).
That is, the flow passage connected to the Desorbent port (D), the Feed port (F), the
Raffinate port (R) and the Extract port (E) have three branches, respectively, and thus,
these are all connected to three rotary valves (10, 20, 30), and afterward are connected
to a certain rotary valve concurrent with the opening of any one of the connection
ports.
Hereinafter, the specific mode of operation is explained.
Fig. 3 is a schematic diagram for a constitution of the simulated moving bed process
(three-zone SMB) having three chromatography zones used in the present invention.
As can be seen from Fig. 3, since the displacement order of the ports is the Desorbent
port, the Feed port, the Raffinate port and the Extract port, respectively, it is possible to
operate the enrichment zone in such a way as to prevent the dilution of the Raffinate
concentration. As a result, it has an advantage in that the degradation of the Raffinate
concentration, which is a problem of the conventional three-zone SMB, can be
prevented. In addition, since it has the structure wherein the Desorbent port and the
Feed port are connected through one column, there is an advantage in that the amount
of the solvent used can be reduced.
Further, the present invention provides the apparatus in which the process as
disclosed in Fig. 4 is continuously in operation. Fig. 4a depicts the apparatus in the first
position, Fig. 4b depicts the apparatus in the second position, and Fig. 4c depicts the
apparatus in the third position. The first, second and third positions are rotated between
continuously. That is, the apparatus according to the present invention has a mode of
operation wherein the chronological consists of the first position, the second position,
the third position, and then a return to the first position again.
[18] A setting to a certain position is made by the rotation of rotary valves (10, 20, 30).
That is, the first connection ports (10a, 20a, 30a) of rotary valves (10, 20, 30) are
opened and are thus set to the first position, the second connection ports (10b, 20, 30b)
are opened by rotating the rotary valves (10, 20, 30) and thus are set to the second
position, and the third connection ports (10c, 20c, 30c) are opened by rotating the
rotary valves (10, 20, 30) again and are set to the third position. If the rotary valves
(10, 20, 30) are rotated again, they are set to the first position again.
[19] Meanwhile, hereinafter, some connection ports (10a, 20b, 30c) are separately shown
with regard to inlet ports (lOa-1, 20b- 1, 30c- 1) and outlet ports (10a-2, 20b-2, 30c-2)
and explained for purposes of comprehension.
[20] In the first position, as shown in Fig. 4a, only the first connection ports (10a, 20a,
30a) of the rotary valves (10, 20, 30) are opened, and the second connection ports (10b,
20b, 30b) and the third connection ports (10c, 20c, 30c) are closed.
[21] In the first position, the desorbent port (D) is connected to the first rotary valve (10),
the Feed port (F) is connected to the second rotary valve (20), the Raffinate port (R) is
connected to the third rotary valve (30), and the Extract port (E) is connected to the
first rotary valve (10).
[22] Accordingly, the desorbent flowed from the Desorbent port (D) passes through the
first rotary valve (10) and the first chromatography zone (40) and then flows in to the
second rotary valve (20).
[23] The mixture to be separated that is flowed from the Feed port (F) is flowed to the
second rotary valve (20) together with the desorbent passed through the first chro
matography zone (40), and then is passed through the second chromatography zone
(50).
[24] The mixture to be separated in the present invention comprises valine, and addi
tionally comprises salt, alanine, leucine, isoleucine, etc. as impurities therein. The
separation of the constituents of the mixture to be separated is achieved by a difference
in flow rates after passing through the second chromatography zone (50). Valine is a
low- affinity component that has a weaker adsorptive power than that of other im
purities, and thereby moves faster. Therefore, the mixture to be separated is then
separated into valine and other materials, and each of them flows into the third rotary
valve (30) with the time difference.
[25] Valine intended to be separated flows out via the Raffinate port (R) but, other
materials are flowed out to the Extract port (E) by passing through the third chro
matography zone (60) and then the first rotary valve (10).
[26] After the set time (i.e., the rotational time interval of the rotary valve) has elapsed,
the rotary valves (10, 20, 30) are rotated and then changed to the second position as
shown in Fig 4b. A criterion of the set time is described below.
[27] In the second position, shown in Fig. 4b, only the second connection ports (10b, 20b,
30b) of rotary valves (10, 20, 30) are opened, and the first connection ports (10a, 20a,
30a) and the third connection ports (10c, 20c, 30c) are closed.
[28] Upon comparison with the first position, the rotary valves (10, 20, 30) connected to
ports (D, F, R, E) are moved sequentially one by one in the second position.
[29] That is, in the second position, the desorbent port (D) is connected to the second
rotary valve (20), the Feed port (F) is connected to the third rotary valve (30), the
Raffinate port (R) is connected to the first rotary valve (40), and the Extract port (E) is
connected to the second rotary valve (20).
[30] Accordingly, the desorbent flowed in from the desorbent port (D) is flowed in to the
third rotary valve (30) after passing through the second rotary valve (20) and the
second chromatography zone (50).
[31] The mixture to be separated that is flowed in from the Feed port (F) is flowed in to
the third rotary valve (30, together with the desorbent passed through the second chro
matography zone (50), and is passed through the third chromatography zone (60).
[32] However, in this case, the remaining materials of the mixture to be separated that
have flowed in from the first position as shown in Fig. 4a have not yet been flowed
into the third rotary valve (30), due to the difference in flow rates. In order to ac
complish this, the rotation time interval mentioned above is controlled.
[33] Therefore, the mixture to be separated that is flowed into the third rotary valve (30)
at the second position passes through the third chromatography zone (60) and then the
separation of the mixture to be separated is made due to the difference of the flow rates
and is then separated into valine and other materials, wherein the valine intended to be
separated is flowed out via the first rotary valve (10) and the Raffinate port (R) while
remaining materials are combined with other materials that were separated in the first
position but were not yet flowed into the third rotary valve (30), to then pass through
the second rotary valve (20) more effectively and are then flowed into the Extract port
(E).
[34] After a lapse of the predetermined time period, the rotary valves (10, 20, 30) are
rotated and changed to the third position, as shown in Fig. 4c.
[35] In the third position, shown in Fig. 4c, only the third connection ports (10c, 20c, 30c)
of the rotary valves (10, 20, 30) are opened, and the first connection ports (10a, 20a,
30a) and the second connection ports (10b, 20b, 30b) are closed.
[36] Upon comparison with the second position, the rotary valves (10, 20, 30) connected
to ports (D, F, R, E) are sequentially moved one by one in the third position.
[37] Identical to the explanation of the first and second positions, the mixture to be
separated are flowed in from the Feed port (F) and then are flowed into the first rotary
valve (10) together with the desorbent passed through the third chromatography zone
(60), and then passed through the first chromatography zone (40), wherein, since in
this case the remaining materials of the mixture to be separated are flowed in at the
second position as shown in Fig 4b and are not yet flowed into the second rotary valve
(20) due to the difference of flow rates, they are also combined together and flowed
into the Extract port (E) after passing through the third rotary valve (30).
[38] Valine intended to be separated in the mixture to be separated is separated to the
Raffinate port (R) after passing through the second rotary valve (20).
[39] After the predetermined time period, the rotary valves (10, 20, 30) are rotated and
changed to the first position as shown in Fig. 4a, and the procedure is continuously
repeated.
[40] As the adsorbent used in the first, second and third chromatography zones in the
present invention, a porous polymer resin can be used, and preferably, the resin
consisting of an insoluble polystyrene divinylbenzene polymer material can be used.
Since the above resin has a broad surface area, a unique pore size and volume dis
tribution, it can be preferably applied for the purification of various materials, and in
particular pharmaceutical compounds. Amberchrom CG161 (Rohm & haas) or
Chromalite PCG Series (Purolite), etc. can be used as a specific example. The above
commercially available resin has a diameter of a particle of 10 - 300 , a surface area
of 700 - 900 g , a pore size of 100 - 200 A, a pore size and volume distribution of
0.7 - 1.5 ml/g, and a uniformity coefficient of less than 2.
[41] The mixture to be separated is a mixture comprising valine, isoleucine or leucine
being included in the branched amino acids together with valine, and preferably may
be valine fermentation liquor obtained through the fermentation of a microorganism.
The specific examples of the present invention utilize the mixture in which the purified
valine, isoleucine and leucine are artificially mixed, and the valine fermentation liquor
obtained by the microorganism fermentation. This is because there is an infrequent
case in which only valine is produced in the fermentation of the microorganism. In
most cases, isoleucine and leucine are additionally produced together.
[42] By the continuous separation apparatus according to the present invention, since
valine intended to be separated in the mixture to be separated is flowed out via the
Raffinate port (R), and the remaining materials are more effectively separated by the
rotation of rotary valves (10, 20, 30) as well as the difference of the flow rate and are
flowed out via the Extract port (E), thereby having valine be continuously separated.
[43] Additionally, through control of the continuous separation apparatus of valine of the
present invention, it can be controlled by regulating the flow rates of the desorbent port
(D), the Feed port (F) and the Raffinate port (R), and also by regulating the rotation
velocity (switching time) of the rotary valve. The flow rate of each port can be
regulated in consideration of the intrinsic parameters of the column (adsorption co
efficient, mass transfer coefficient, porosity, material environment, etc).
[44] In addition, the present invention provides the continuous separation method of
valine characterized in that it comprises the steps of flowing the desorbent using the
desorbent port, flowing the mixture comprising valine into the Feed port, and r e
covering valine from the Raffinate port, by way of the method for the continuous
separation of valine by using the continuous separation apparatus of the valine. The
desorbent is preferably water. In addition, it is characterized in that the purity of valine
recovered by the method is 90% to 99%.
[45] According to one example of the present invention, as a result of the separation of the
mixture to be separated comprising valine, leucine and isoleucine, by using the
continuous separation apparatus of valine according to the present invention, the yield
of the recovered valine was about 98% or more, the purity was about 98% or more, and
thus, it was verified that valine could be effectively separated from the mixture.
Advantageous Effects of Invention
[46] The present invention can continuously separate valine from the mixture comprising
amino acids such as leucine, isoleucine, etc. while having high purity and yield, by
way of the apparatus using the simulated moving bed chromatography process.
Brief Description of Drawings
[47] Fig. 1 depicts the schematic diagram of the Simulated Moving Bed (four-zone SMB)
chromatographic process having the conventional four chromatography zones.
[48] Fig. 2 is the schematic diagram of the Simulated Moving Bed (three-zone SMB)
chromatographic process having the conventional three chromatography zones.
[49] Fig. 3 is the schematic diagram of the Simulated Moving Bed chromatographic
process having three chromatography zones (three-zone SMB) used in the present
invention and the process is developed based on the port disposition mode which is dif
ferentiated from the conventional three-zone SMB.
[50] Fig. 4 shows a mimetic diagram of the apparatus for the three-zone SMB process
used in the present invention. Figs. 4a, 4b and 4c show the connections of each zone
according to the rotation of rotary valves, respectively.
[51] Fig. 5 shows the pulse test results for selecting the adsorbent. Wherein, (a) represents
the result for Amberchrom CG161C, (b) represents the result for Amberlite CG71C,
(c) represents the result for DIAION SK1B, and (d) represents the result for Amberlite
XAD-7HP.
[52] Fig. 6 shows the result of a stepwise frontal test for Amberchrom CG161C adsorbent.
Wherein, (a) represents the result for isoleucine, (b) represents the result for leucine,
and (c) represents the result for valine.
[53] Fig. 7 shows the adsorptive equilibrium data (q vs. C) on Amberchrom CG161C
adsorbent. Wherein, (a) represents the data for valine, (b) represents the data for
leucine, and (c) represents the data for isoleucine.
[54] Fig. 8 shows a comparison between the frontal test result of the mixture and the
simulation result. Wherein, (a) represents the comparison of the result for leucine, (b)
represents the comparison of the result for isoleucine, and (c) represents the
comparison of the result for valine.
[55] Fig. 9 shows the result of HPLC concentration analysis for outflow solution
discharged through the SMB desorbent port, as SMB test result. Wherein, (a)
represents the result for Raffinate concentration analysis, and (b) represents the result
for the extract concentration analysis.
[56] Fig. 10 shows a HPLC raw data chromatogram for the sample obtained from SMB
test. Wherein, (a) represents the chromatogram of the mixture to be separated, (b)
represents that of raffinate, and (c) represents that of the extract.
[57] Fig. 11 shows a column profile (60.5 step), as a SMB test result. Wherein the black
line represents the simulated result of valine, the gray line represents the simulated
result of isoleucine, the black dotted line represents the simulated result of leucine, the
circle represents the test result of valine, the quadrangle represents the test result of
isoleucine, and the diamond represents the test result of leucine.
[58] Fig. 12 shows the HPLC concentration analysis result of outflow solutions dis
charging via the Raffinate port, as the SMB test result conducted based on Chromalite
PCG-600 adsorbent.
[59] Fig. 13 shows the HPLC concentration analysis result for outflow solutions
discharged via the Extract port, as the SMB test result conducted based on Chromalite
PCG-600 adsorbent.
[60] Fig. 14 shows a column profile (44.5 step), as SMB test result conducted based on
Chromalite PCG-600 adsorbent. Wherein, the black line represents the simulated result
of valine, the gray line represents the simulated result of leucine, the circle represents
the test result of valine, and the triangle represents the test result of leucine.
Mode for the Invention
[61] Hereinafter, it is intended to explain the constitution and effects of the present
invention in detail via Examples, but these Examples are intended only to illustrate the
description of the present invention and the scope and spirit of the present invention is
not intended to be limited by these Examples.
[62]
[63] Approach: Investigation for the chromatographic application of Valine pu¬
rification
[64] 1) Model-based design approach
[65] There are two matters that are initially considered upon developing a continuous
process for a new separation system. The first is the minimization of the cost and the
period required for development. The second is the conditions providing the best
output productivity and the best separation efficiency, by maintaining the optimum
state of the process to be developed. In order to satisfy both conditions as above, one
should precisely grasp the adsorption and material-transferring phenomenon based on
the detailed model, and also obtain various parameters related thereto. The method of
process design based on this detailed model and parameters is referred to as a "modelbased
design approach". Also, the present invention has developed a SMB process in
carrying out the continuous separation of valine according to the approach.
[66]
[67] 2) Computer Simulation
[68] One of the key stages of model-based design approach is a computer simulation. This
refers to the procedure of obtaining the solution by solving the detailed model equation
for the adsorption and material-transferring phenomenon of each component in a
column by way of a numerical analytic method. This numerical analytic method is
carried out by using a computer, since it requires a vast number of calculations.
[69] There are many kinds of column model equations to be used in a simulation, and
among them the lumped mass-transfer model (Kang, S.H. et al., Process Biochem.,
2010, 45, 1468-1476) is determined as the simulation model of the present invention in
consideration of accuracy and efficiency. The computer simulation based on lumped
mass-transfer model was utilized in the assessment of separation efficiency of the SMB
process as well as the measurement and assessment of base parameter of each amino
acid component. Furthermore, this model equation is utilized in the manufacture of the
SMB optimization computer tool.
[70]
[71] 3) SMB Optimization Tool
[72] Another key role in the model-based design approach that is utilized after the
computer simulation is played by the SMB optimized computer tool. This tool is used
in obtaining the optimal operating conditions of SMB process to be developed. The
matter first necessary for manufacturing this optimized tool is an optimization
algorithm. Conventionally, it is known that a gene algorithm based on stochastic theory
is the most efficient in the complex form of the process optimization such as SMB
(Lee, K.B. et al., AIChE J., 2008, 54, 2852-2871).
[73] Also, the present invention has established a SMB optimized computer program
based on a gene algorithm for optimization of continuous separation of valine. The
gene algorithm itself has been developed several times in the interim, and the NSGAII-
JG algorithm (Lee, K.B. et al., AIChE J., 2008, 54, 2852-2871), which can be said to
be the latest gene algorithm in the manufacturing stage of the optimization tool of the
present invention, is adopted as the basic algorithm.
[74] The method for preparing a SMB optimized tool codes the optimized algorithm by
using visual basic application (VBA) language installed in Microsoft Excel software,
and allows for the calculations of the detailed model equation and the NSGA-II-JG
algorithm to be simultaneously carried out therein.
[75]
[76] Preparation of Experiment
[77] 1) Materials
[78] Valine and leucine, among three amino acid components constituting the mixture to
be separated were purchased from Fluka, and isoleucine was purchased from Sigma.
Water used to dissolve amino acid was tertiary Distilled Deionized Water (DDW) and
obtained through the Milli-Q system (Millipore). Adsorbent used in the experiment
was Amberchrom CG161C (Rohm & haas), Amberlite CG71C (Sigma Aldrich),
DIAION SK1B (Mitsubishi Chemical), Amberlite XAD-7HP (Sigma Aldrich).
Methanol used in HPLC concentration analysis was purchased from Burdick &
Jackson Co. (Muskegon, MI).
[79] An Omnifit glass column used in the adsorbent selection experiment and the mea
surement experiment of the basic parameter of each material was purchased from Bio
chemical Fluidics Co. (Boonton, NJ), and in the former experiment the column having
11.6 cm of length and 1.5 cm of diameter was used, and in the latter experiment the
column having 21.7 cm of length and 2.5 cm of diameter was used.
[80]
[81] 2) Equipment
[82] - Single column experimental equipment
[83] Young-Lin SP930D pump and Young-Lin UV730D detector were used in the
adsorbent selection experiment. The control and data processing of these two pieces of
equipment was made by Autochro-3000 software. An injector used in injecting the
amino acid pulse into the column each filled with adsorbent was the Rheodyne 7725i
injector, and the injection volume was 100 b .
[84] In the measurement experiment of the basic parameter of each amino acid conducted
after completing the adsorbent selection, FPLC P-920 pump, Waters 486 UV detector,
and Amersham FPLC collector (Frac-900) were used. The control and data collection
of each of the detailed devices were made through Unicorn 5.1 software.
- Three-zone SMB apparatus for continuous separation of valine
The apparatus as shown in Fig. 4 was self-assembled and used. The flow rate of the
desorbent of the completed SMB experimental apparatus and the flow rate of the
mixture to be separated were controlled by using the Young-Lin SP930D pump and the
raffinate flow rate were controlled by the Ismatec MCP-CPF ISM 919 pump. In order
to generate the effect wherein four ports are periodically moved, a ST Valco rotary
valve (VICI, Houston, TX) was used. The mimetic diagrams of the ST valve are rep
resented in Figs. 4a-4c. A Valco rotary valve used in the experiment was automatically
controlled through Labview 8.0 software.
- HPLC concentration analysis device (analyzer)
In order to measure the concentration of the material collected through SMB ex
periment, a HPLC concentration analyzer was used. As the column for the analysis,
Waters Symmetry C-18 column (250 x 4.6 mm ID, particle size 5 ) was used. The
collection of all data related to HPLC concentration analysis was treated by using
Waters Millennium software. The flow rate of the mobile phase was controlled by
using the Waters 515 HPLC pump. The sample for the concentration analysis was
injected into the column for analysis by using a Rheodyne 9725i injector, and the
injection volume was 20 b . The detection of a sample was conducted by using a
Waters 996 PDA detector.
In the concentration analysis of the sample obtained from a frontal test of valine, a
Young-Lin SP930D pump and a Young-Lin UV730D detector were used. The control
and data processing of the Young-Lin pump and detector were conducted by Autochro-
3000 software. The sample obtained from the experiment was injected into the column
for the analysis by using a Rheodyne 7725i injector, and the injection volume was 20
id.
Example 1: Selection of an adsorbent suitable for separation of valine
1) Experimental Method
The experiment to select the most suitable adsorbent was conducted by using four
kinds of conventional adsorbents that have the possibility of separating leucine and
isoleucine from valine, such as Amberchrom CG161C (Rohm & haas), Amberlite
CG71C (Sigma Aldrich), DIAION SKIB (Mitsubishi Chemical), Amberlite XAD-7HP
(Sigma Aldrich), through the manner of a series of pulse tests.
[97] The conditions for the pulse test experiments were as follows: Each of the concen
trations of three amino acids was 2 g/L, and the injection volume was 100 b . The flow
rate of the mobile phase was 2 ml/min.
[98]
[99] 2) Experimental Results
[100] The results are represented in Fig. 5. As represented in Fig. 5, it could be ascertained
that in case of other adsorbents other than the Amberchrom CG161C adsorbent, the
selective separation of valine is difficult. It is ascertained from the above results that
the Amberchrom CG161C adsorbent is most suitable for establishing SMB process for
separating valine.
[101]
[102] Example 2 : Measurement of porosity of Amberchrom CG161C column
[103] 1) Experimental method
[104] The porosity of the Amberchrom-CG161C adsorbent as selected in Example 1 was
measured. At first, the experiment that fills the column with the adsorbent, and then
measures an inter-particle porosity (eb) was conducted. Although it is conducted in the
same manner as Example 1, in this instance a tracer molecule suitable for measuring
the inter-particle porosity was injected into the column instead of the material to be
separated, unlike as in Example 1. The tracer molecule used was blue dextran, and the
injection concentration was 1 g/L, the flow rate was 4 ml/min and the injection volume
[105]
[106] 2) Experimental results
[107] From the experimental results, the value of the inter-particle porosity of the column
was 0.391. The intra-particle porosity, which is one of the important information used
the value which is reported in the reference, was 0.737 (Nam, H.G. et al., Process
Biochem., 2011, 46, 2044-2053).
[108]
[109] Example 3 : Measurement of the basic parameter of valine, leucine, isoleucine on
Amberchrom CG161C
[110] - Measurement of adsorption coefficient
[111] 1) Experimental method
[112] A multiple frontal test was performed in order to measure the basic parameters of
valine, leucine, isoleucine that correspond to each component of the mixture which is a
subject of the separation. The multiple frontal test is the experiment that is carried out
to obtain the adsorption equilibrium data for each material to be separated, and to
define the adsorption model equation and determine the relevant adsorption parameters
based on the obtained data.
[113] Specifically, a multiple frontal test used two pumps, and one of the pumps was filled
with DDW, and another pump was filled with valine, leucine or isoleucine solution.
Then, valine, leucine or isoleucine solution was continuously injected into the column
until the equilibrium between the adsorbent phase and mobile phase in the column was
achieved. In the equilibrium state, since all of the concentrations between the adsorbent
particle and the particle as well as the concentration inside the adsorbent are
maintained similar to the concentration of valine, leucine, or isoleucine solution
injected, the adsorbent concentration on the adsorbent can be immediately investigated
by establishing a simple Mass Balance Equation. After the concentration inside the
column reaches an equilibrium state, a new equilibrium state is allowed to be
maintained by increasing the ratio of the mixture solution which is the subject of the
separation, compared to that of the previous stage. Again, the adsorption concentration
on the adsorbent is calculated by establishing a Mass Balance Equation for this
equilibrium state. While progressing the multiple frontal test, a step time for each
amino acid component is set as 2-3 times the residence time obtained from the results
of the pulse test.
[114] Also, the multiple frontal test for each amino acid was conducted as a total of five
steps, and the step time is set as 40 min for valine, 70 min for leucine, and 70 min for
isoleucine. In addition, the concentration of all samples was 5 g/L, and the flow rate
was constantly maintained as 4 ml/min. The experiments for leucine and isoleucine
were conducted in waves of 205 nm, and 218 nm at UV detector. On the other hand, in
the case of valine, a direct HPLC analysis manner (the manner sampling the column
effluents one by one and then analyzing the concentrations of them by using HPLC
device), which is not an on-line monitoring manner as used in the leucine and
isoleucine experiments was adopted. The HPLC analyzing conditions are that 10%
aqueous methanol solution was used as the mobile phase, the injection volume was 20
and the flow rate was 0.5 ml/min. Before the analysis, a self calibration curve was
secured by using a standard solution. The results are represented in Fig. 6, the ad
sorption equilibrium data is calculated based on these results, and the calculated results
are represented on Fig. 7.
[115]
[116] 2) Experimental Results
[117] As shown in Figure 7, all of valine, leucine and isoleucine represent the linear ad
sorption relationship on Amberchrom CG161C. Therefore, each of the amino acid
components is modeled by the linear adsorption equation, and the linear adsorption co
efficient was determined therefrom. The values of the determined adsorption coef
ficients are represented in Table 1, as below.
[118]
[119] - Determination of a mass-transfer coefficient
[120] 1) Experimental method
[121] A mass transfer coefficient which can be referred to as another important basic
parameter together with the adsorption coefficient was determined as per the following
method.
[122] At first, an axial dispersion coefficient and film mass-transfer coefficient were re
spectively expected by using Chung & Wen correlation (Chung, S. F. et al., AIChE J.,
1968, 14, 857-866) and Wilson & Geankopolis correlation (Wilson, E. J. et al., Ind.
Eng. Chem. Fundam., 1966, 5, 9-14), and a molecular diffusivity was calculated by
using Wilke & Chang correlation (Wilke, C.R. et al., AIChE J.. 1955, 1, 264-270).
Also, intra-particle diffusivity was determined while fitting the frontal experimental
data and the simulation results based on the lumped mass-transfer model each other.
[123]
[124] 2) Experimental results
[125] The values of the molecular diffusivity and the intra-particle diffusivity determined
above are represented in Table 1 as below:
[126] Table 1
[Table 1]
[127]
[128] In order to ascertain the suitability of the values of the adsorption and mass transfer
coefficient as determined above, the results of model simulation and frontal ex
perimental results substituting the values were compared as shown in Figure 6.
[129] As shown in Figure 6, it can be ascertained that the simulation results and the ex
perimental data coincide well. Therefore, the value of the basic parameter as de
termined above can be sufficiently utilized for the optimization of SMB process, as
conducted later.
[130]
[131] Example 4 : Investigation of the basic parameter by a frontal experiment of the
mixture
[132] 1) Experimental method
[133] A frontal experiment of the mixture was carried out in order to investigate the basic
parameter value determined in Example 3. In this experiment, the frontal experiment
was conducted by using the mixture solution comprising all of three amino acids as the
mixture to be separated, unlike as in Example 3.
[134] Also, the model simulation was conducted on grounds similar to those of the basic
parameter values determined in Example 3 and the mixture frontal experimental
conditions. Additionally, the obtained simulation results were compared to the frontal
experimental data of the mixture, and the results are represented in Figure 8.
[135] 2) Experimental results
[136] As shown in Figure 8, it can be ascertained that the experimental data coincided well
with the simulation results. This means that the basic parameter values determined in
Example 3 can well explain the behavior of each amino acid in the column in the state
of being in the mixture as well as in the state of being a separate component. Fur
thermore, it means that interaction between each of the amino acids is rarely present.
[137]
[138] Example 5 : Optimization of the process for continuous separation of valine
[139] The matter focused upon in the optimization procedure is to maximize the pro
ductivity of valine while assuring the high purity and high yield of valine that is the
target of the separation. Wherein the productivity of valine is defined as per the
following equation:
[140] - Productivity of valine = Qraf * C f,V i
[141] (wherein, Q f is the flow rate in the Raffinate port and, C e f i e means the con
centration of valine in the Raffinate port.)
[142] The productivity of valine was established as an objective function and the purity and
yield of valine were established as a constraint, to conduct the optimization procedures.
The concrete optimization procedures are summarized as follows:
[143] Max J = Productivity [Qfeed, Qraf, tsw]
[144] Subject to Valine purity = 98%
[145] Valine yield = 98%
[146] Fixed variables Qdes = 5 mL/min
[147] Cfeed for each component = 5 g/L
[148] Lc = 21.7 cm, dc = 2.5 cm
[149] (wherein, Qfeed and Qdes independently refer to the flow rate of the mixture to be
separated and the flow rate of the desorbent, respectively, and tsw refers to the
switching time.)
[150]
[151] In order to optimize the separation process of valine using the three-zone SMB
device of the present invention according to the equation, the optimized computer
program based on the NSGA-II-JG algorism was self-coded, and the computer
program was connected to an Aspen simulator to optimize the relevant process. The
results are represented in Table 2 as below.
[152] Table 2
[Table 2]
[153]
[154] Example 6 : SMB Experiment
[155] By using the date obtained above, valine was separated by the continuous separation
device of valine according to the present invention, and the yield and purity thereof
were measured as follows:
[156]
[157] 1) Experimental method
[158] As shown in Fig. 4, a SMB experiment was carried out by using the optimized
conditions derived from above Example 5. Amberchrom CG161C was used as the
column.
[159] To begin, the column was connected prior to the start of the experiment. At this time,
the flow rate of desorbent (water) was maintained as 2 ml/min, and the operation of the
other pump was halted. Each of the valve and column were connected, while at this
time care was taken such that no air is entered inside the column. After completing all
the connections of the columns, the flow rate of the desorbent was increased up to the
target value. Since the start of the SMB experiment is for the time that the mixture
solution to be separated is injected, the pump of the mixture to be separated was
operated concurrent to the operation of Labview 8.0 software. The switching and
switching time of the valve were simultaneously controlled via Labview 8.0.
[160] The SMB experiment was progressed until it sufficiently reached a cyclic steady
state. It was ascertained that the steady state was achieved after the 25th step.
Therefore, the SMB experiment progressed up to the 60th step, which is a value greatly
in excess of that of the steady state. Throughout the SMB experiment, at every step, the
concentration of the solution eluted from each outlet port was analyzed by using the
HPLC device. Also, in order to secure the column profile, the relevant sample was
taken at the last step. For the purpose of this, all driving pumps were stopped when the
SMB experiment reached the 60.5 step. Additionally, the solution exiting from the
column by opening the lower part of the column connected to the valve was collected,
and the concentration of the solution was analyzed. HPLC analysis conditions were
that: 10% aqueous methanol solution was used as the mobile phase, the injection
volume was 20 b and the flow rate was 0.5 ml/min. Before the analysis, the selfcalibration
curve was secured by using the standard solution. The results are rep
resented in Figure 9. In Figure 10, (a) is the analysis result of the raffinate con
centration, and (b) is the analysis result of the extract concentration.
[161]
[162] 2) Experiment results
[163] As shown in Figure 9a, it could be ascertained that most of the components in the
raffinate solution were valine. Leucine and isoleucine, belonging to the impurities,
were rarely present. This means that the purity of valine being maintained is very high.
Also, as shown in Figure 9b from the analysis result of the extract concentration, most
of the components were leucine and isoleucine, and the detected amount of valine was
of a negligible quantity. This means that the loss of valine via the outlet port is
minimized.
[164] Also, the computer simulation for the relevant SMB process was carried out simul
taneously with the progress of the SMB experiment, and the result was directly
compared with the HPLC concentration analysis data of the raffinate and the extract.
As shown in Figure 9, it can be seen that the simulation result and SMB experimental
data correlated well.
[165] In order to make a quantitative measure of the final purity and yield of valine
separated by SMB experiment as mentioned above, the effluent solutions obtained
during the last six steps were mixed in the same ratio and HPLC concentration analysis
for the solution was then conducted. Based on the analyzed concentration data, the
purity and yield of valine were calculated, and the results were represented in Table 3
as below.
[166] Table 3
[Table 3]
[167]
[168] From the results of Table 3 and Fig. 9, it could be ascertained that the SMB process
using Amberchrom CG161C of the present invention is superior in assuring the
continuous separation of valine and maintaining high purity and high yield throughout.
As the experimental evidence data for this, a raw chromatogram of HPLC con
centration analysis (HPLC raw data) was represented in Figure 10.
[169] At first, in HPLC raw data (Fig. 10a), for the mixture solution to be separated all of
three amino acid components represented large peaks. In HPLC raw data (Fig. 10b) for
the Raffinate port effluent that belongs to the production port of valine, only the valine
component clearly showed a large peak while the peak of isoleucine showed a very
small peak, and further the peak of leucine was rarely found. In HPLC raw data (Fig.
10c) for the Extract port effluent which is established so as to obtain only data corre
sponding to the impurity, the peak of the valine component is negligible, while the two
remaining amino acid components showed a large peak. Through this series of HPLC
raw data, it could be ascertained again that the SMB experiment for the continuous
separation of valine carried out as in the present invention was successfully conducted.
[170] In addition to the concentration graphs of the raffinate and extract as mentioned
above, the column profile data is also important SMB experimental data. For this
reason, the samples required for securing the column profile data were taken during the
halfway mark of the final switching period, that is, at step 60.5. After analyzing the
sample concentration, the result was represented in Fig. 12. As can be seen from Fig.
11, it can be ascertained that the column profile data also correlates well with the
simulation result. This discloses that the column profile proves the SMB experiment
was conducted well. That is to say, it can be said that the facts as experimentally
verified that solute waves of each of the amino acids in the SMB column were dis
tributed so that they are fully advantageous to a high purity and yield.
[171] Therefore, from all of the SMB experimental data (the concentration graph at the
outlet port, column profile) mentioned above, it could be seen that the SMB process of
the present invention using Amberchrom CG161C is sufficiently applicable to the
continuous separation process of valine on an industrial scale.
[172] The above mentioned SMB process was optimized based on Amberchrom CG161C
adsorbent, and also verified experimentally. In addition to this adsorbent, SMB process
based on Chromalite PCG-600 wherein the effect on the valine separation was verified
was also conducted for optimization, and the experiment in this regard was also
conducted. As a result, it was ascertained that Chromalite PCG-600 adsorbent was also
sufficiently utilizable for the continuous separation of valine upon being applied to the
SMB process.
[173]
[174] Example 7 : Experiment in the actual fermentation mixture
[175] According to a method similar to that of the previous experiment, an experiment was
carried out by using the actual fermentation mixture.
[176] In order to measure the basic parameter (adsorption and mass transfer coefficient) of
valine and leucine components in the actual fermentation mixture, a mixture frontal ex
periment was conducted wherein the solution of the fermentation mixture was injected
into a single column filled with Chromalite PCG600C resin while the concentration of
column effluent was measured over the time.
[177] The basic parameters of valine and leucine were determined by using the con
centration profile data obtained from the experiment and the inverse method, and the
results are given in Table 4.
[178] Table 4
[Table 4]
[179]
[180] On the basis of the basic parameters given in the above Table 4, the PCG600C-SMB
process continuously separating valine and leucine from the actual fermentation
mixture was optimally designed. This SMB process was also based on the new port
disposition order of the present invention, and the column configuration adapted the
three-zone structure as shown in Fig. 3.
[181] The matter being focused upon in the optimization procedure of PCG600C-SMB
process was the maximization of the productivity of valine while assuring a high yield
of valine and a high removal efficiency of leucine. In order to achieve this optimization
purpose, the productivity of valine was established as an objective function, and the
yield of valine and the removal rates of leucine were established as the constraint to the
conduct of the optimization procedure. The concrete optimization procedures are
summarized below. Qdes is the value which is controlled within the scope that the
purity and yield of valine can be increased.
[182] Max J = Productivity [Qfeed, Qraf, tsw]
[183] Subject to Valine purity = 97%
[184] Leucine removal efficiency = 90%
[185] Fixed variables Q des = 6 mL/min
[186] Lc = 21.7 cm, dc = 2.5 cm
[187] Column configuration = 1 - 1 - 2
[188] (wherein, Qfeed and Qdes refer to the flow rate of the mixture to be separated and the
flow rate of the desorbent, respectively, and tsw refers to the switching time.)
[189] In order to conduct the optimization of PCG600C-SMB process according to the
above equation, the optimized computer program based on NSGA-II-JG algorithm was
self-coded, and the computer program was connected to the Aspen simulator to obtain
the optimization of the relevant process. The optimization results are given in Table 5.
[190] Table 5
[Table 5]
In order to experimentally verify the optimization results of Table 5, PCG600C-SMB
process device was self-assembled. The mimetic diagram of the assembled process
device is given in Fig. 4.
The SMB experiment for separating valine wherein the actual fermentation mixture
is the subject was conducted by using the optimization result of Table 2 and the
process device of Fig. 4. The concentrations of valine and leucine in the actual fer
mentation mixture used in this experiment were 71.8 g/L and 0.742 g/L, respectively.
The results of the SMB experiment for separating valine, wherein the actual fer
mentation mixture was the subject are represented in Figs. 12-14. As shown in the
result of the raffinate effluent history in Fig. 12, the high concentration recovery of
valine was made in accordance with the level being expected by the simulation, and
the effluent concentration level of leucine was low enough to the extent of being
considered a negligible quantity. As shown in the result of the extract effluent history
of Fig. 13, it could be ascertained that leucine is removed in accordance with the level
expected by the simulation. Also, it could be ascertained that the loss of valine via an
Extract port was minimized. These experimental results mean that a high yield of
valine and a high removal rate of leucine can be sufficiently assured by the continuous
separation of valine and leucine by the PCG600C-SMB process.
Further to the effluent history results of Figs. 12 and 13, the column profile result of
Fig. 14 also comprehensively shows that the continuous separation of valine and
leucine was successfully conducted. As shown in Fig. 14, it can be ascertained that the
concentration distribution of valine and leucine in each column is made to be very ad
vantageous in assuring a high yield recovery of valine and a high removal rate of
leucine. Due to these results, 99.7 % of valine in the actual fermentation mixture could
be recovered via the Raffinate port, and at the same time 98.0 % of leucine could be
removed via the Extract port.
Claims
[Claim 1] An apparatus for continuous separation of valine, which comprising:
a Desorbent port (D);
a Feed port (F);
a Raffinate port (R);
an Extract port (E);
a number of rotary valves (10, 20, 30) each selectively connected to the
ports (D, F, R, E); and
a number of chromatography zones (40, 50, 60) each equipped with
every a number of rotary valves (10, 20, 30),
wherein the number of rotary valves (10, 20, 30) are connected to each
other,
wherein the number of rotary valves (10, 20, 30) are equipped with the
number of connection ports (10a, 10b, 10c) (20a, 20b, 20c) (30a, 30b,
30c), respectively, and
wherein only any one of the number of connection ports (10a, 10b, 10c)
(20a, 20b, 20c) (30a, 30b, 30c) is opened, along with the rotation of the
number of rotary valves (10, 20, 30), and subsequently any one of the
rotary valves (10, 20, 30) selectively connected to each of the ports (D,
F, R, E) are changed.
[Claim 2] The apparatus for continuous separation of valine according to claim 1,
wherein:
the number of rotary valves (10, 20, 30) rotate after a short period of
time, and
the rotary valves (10, 20, 30) connected to the ports (D, F, R, E) are
changed, along with the rotation of the number of rotary valves (10, 20,
30).
[Claim 3] The apparatus for continuous separation of valine according to claim 2,
wherein:
the number of rotary valves (10, 20, 30) are continuously changed so as
to rotate among the first, second and third positions along with the
rotation after the short period of time,
the first connection ports (10a, 20a, 30a) are opened at the first
position,
the second connection ports (10b, 20b, 30b) are opened at the second
position, and
the third connection ports (10c, 20c, 30c) are opened at the third
position.
[Claim 4] The apparatus for continuous separation of valine according to claim 3,
wherein, at the first position,
the Desorbent port (D) is fluid-communicated with the first chro
matography zone (40) through the first connection port (10a) of the first
rotary valve (10),
the Feed port (F) is fluid-communicated with the second chro
matography zone (50) through the second connection port (20a) of the
second rotary valve (20),
the second chromatography zone (50) is fluid-communicated with the
Raffinate port (R) through the third connection port (30a) of the third
rotary valve (30), and
the Extract port (E) is fluid-communicated with the third chro
matography zone (60) through the first connection port (10a) of the first
rotary valve (10).
[Claim 5] The apparatus for continuous separation of valine according to claim 4,
wherein:
the mixture to be separated flowed in via the Feed port (F) at the first
position is separated into valine and the remaining materials, along with
passing through the second chromatography zone (50),
the valine separated from the first position flows in to the third rotary
valve (30) and then flows out via the Raffinate port (R), and
the remaining materials separated from the first position flow into the
first rotary valve (10) and then flow out via the Extract port (E).
[Claim 6] The apparatus for continuous separation of valine according to claim 5,
wherein:
the short period of time is the time that the valine separated from the
mixture to be separated is flowed into the third rotary valve (30), but
the remaining materials are not flowed into the third rotary valve (30).
[Claim 7] The apparatus for continuous separation of valine according to claim 6,
wherein:
when the rotary valves (10, 20, 30) rotate from the first position to the
second position, the mixture to be separated flowed in via the Feed port
(F) is separated to valine and the remaining materials, along with
passing through the third chromatography zone (60),
the valine separated at the second position is flowed into the first rotary
valve (10) and then flowed out via the Raffinate port (R), and
the remaining materials separated from the second position are flowed
into the second rotary valve (20), along with the remaining materials
separated during the first position and then flowed out via the Extract
port (E).
[Claim 8] The apparatus for continuous separation of valine according to claim 1,
wherein:
the number of rotary valves (10, 20, 30) are alternatively changed at the
first, second and third positions by the continuous rotation, and
the rotary valves to which the ports (D, F, R, E) are connected at the
first, second and third positions are different from each other.
[Claim 9] The apparatus for continuous separation of valine according to claim 8,
wherein:
the mixture to be separated flowed in via the Feed port (F) is separated
to the valine and the remaining materials, and
the rotation interval in which the number of rotary valves (10, 20, 30)
change their positions is the time that the valine moves any one of the
rotary valves to another rotary valve but the remaining materials do not
move.
[Claim 10] A method for continuously separating valine by using the apparatus for
continuous separation of valine as defined in any one of claims 1 to 9,
comprising:
flowing a desorbent into the Desorbent port,
flowing a mixture comprising valine into the Feed port, and
recovering valine from the Raffinate port.
[Claim 11] The method for continuously separating valine according to claim 10,
wherein the desorbent is water.
[Claim 12] The method for continuously separating valine according to claim 10,
wherein the purity of the valine recovered is 90% to 99%.