close

Вход

Забыли?

вход по аккаунту

?

The Effect of Protein and Broth Impurities on the Partition of Citric Acid into n-Butyanol from Aqueous Solutions.

код для вставкиСкачать
Dev.Chem. Eng. Mineral Process., I 1 (3/4), pp. 223-246, 2003.
The Effect of Protein and Broth Impurities
on the Partition of Citric Acid into
n-Butanol from Aqueous Solutions
N. Carolan and L.R.Weatherley"
Dept of Chemical & Process Engineering, University of Canterbury,
Private Bag 4800, Christchurch, New Zealand
The use of liquid-liquid extraction for the recovery of biological products @om
fermentation media by direct extraction may be significantly influenced by the
presence of impurities. This can offset the process advantages of using direct whole
broth extraction which, iffeasible, can reduce the number of stages needed for the
required separation. The impurities may slow down mass transfer rate and may also
affect the thermodynamic distribution of the required solute into the extracting
solvent. In some cases the addition of reagents can be beneficial to the extraction.
Here an experimental study is described in which the partitioning of citric acid into nbutanol in the presence of a range of impurities has been measured. A range of
impurities and additives were studied in an attempt to quantia and prioritise those
compoundr which affect the partition in practice. These comprised a range of
proteins, electrolytes, a surfactant and other organic acidr. The additives considered
included: tvpsin, bovine serum albumin, sodium lauryl sulphate, glucose, sucrose,
lactic acid, magnesium sulphate and, sodium chloride. m e modification of citric acid
partitioning into n-butanol @om whole broth, filtered broth and suspensions of
reconstituted biomass were also compared. The presence of the proteins and sodium
had a detrimental effect on partition. The effect of magnesium sulphate was highly
concentration dependent.
*Authorfor correspondence.
223
N. Carolan and L.R. Weatherley
Introduction
The production of high-value products through a range of fermentations and novel
biotransformations is becoming well established especially in the pharmaceutical
industry and in the food and speciality chemicals sector. It is highly probable in the
longer term that the use of microbial processes will extend to the manufacture of low
molecular weight bulk chemical products based upon renewable feedstocks. Solvent
extraction is one of several important unit operations used for separation and
purification of such products synthesised conventionally and biologically. It can be
applied as part of a single-line continuous process as in a conventional recovery
operation, or as part of a recycle process to facilitate simultaneous product
purification and the separation of unreacted substrates. The application of such an
approach is of great of importance for viable development of extractive fermentations
and biotransformations which facilitate enhanced product yield, and the production of
unstable intermediate products. Process yield and productivity may be enhanced if
direct extraction of the untreated fermentation liquor or reactor solutions can be
achieved in line.
The negative effects of biomass components and impurities upon rates of
extraction and upon rate of phase separation are well documented [l-41. In the wider
context of industrial solvent extraction processes, the effects of surface active agents
upon kinetics of liquid-liquid extraction are also well documented [5, 61. There are
components in fermentation broths which affect extraction. These include biomass
solids, nucleic acids [7],polysaccharides which are produced during the fermentation
[8], and proteins [9]. Other components include, weacted substrate, nutrient
compounds, lipids, fatty acids, buffering electrolytes and side products of the
fermentation.
Some of the components present in fermentation broth affect the physical
processes which influence overall rates of extraction. These include rheological
properties and interfacial stability, both of which strongly influence mass transfer.
Here we are concerned with the effects which these components have upon the
partition behaviour of the main extracting solute. In this paper we are focussing on
citric acid extraction.
224
The Partition of Citric Acid into n-Butanol @om Aqueous Solutions
Submerged fermentation processes are most widely used for industrial production
of citric acid, with approximately 80% of the citric acid produced by this method [ 101.
Aspergillus niger is the most commonly used microorganism. Typical yields of citric
acid are 0.70 to 0.90 g citrate/g glucose; the maximum yield from anhydrous glucose
is 1.17 glg [ll].
Currently citric acid separation is achieved by precipitation of calcium citrate
from the fermentation broth [12]. This method results in large quantities of calcium
sulphate, and the cost of disposal can be considerable. A number of alternative
recovery procedures are being investigated and solvent extraction is one of the most
promising. It has already been applied on an industrial scale, although the citric acid
produced is only suitable for industrial rather than food industry applications.
Published studies on citric acid extraction have focussed upon toxicity effects in
extractive fermentation [13]. Extraction into 1-dodecanol and into tertiary amine from
an Aspergillus niger fermentation broth showed that fungal growth was inhibited by
the two solvents and citric acid production was severely impeded.
Studies of the extraction of citric acid from fermentation broth [14-161 into
1-butanol and 2-butanol show qualitatively that broth impurities can have a
considerable negative influence upon the extraction. The distribution coefficient
decreases as the purity of the aqueous phase diminishes. Extraction of citric acid into
2-butanol from a fermentation broth system was less than in the pure system, but was
higher than that observed into 1-butanol. There have been few attempts to identify
and analyse the impurities which exert the most influence upon the partition.
Wennersten [ 171 investigated, the extraction of citric acid (50 g/L) fiom a filtered
fermentation broth into 70% (v/v) TBP in Shellshol H, however problems with
emulsification prevented detailed comparison with the pure system.
The extraction of citric acid fiom a filtered broth [18] into trialkyl phosphine
oxide (TWO) in a diluent has also been reported. The ‘impurities’ in the broth phase
were reported to influence the magnitude of the distribution coefficient and the shape
of the extraction isotherm. There is significant influence by a number of broth
constituents which may be co-extracted, and the presence of calcium, magnesium and
ferrous ions appears to exert an important influence upon the distribution coefficient.
225
N. Carolan and L.R. Weatherley
The use of emulsion liquid membrane systems for the extraction of citric acid
from fermentation broth has been reported [19, 201. A significant increase in the
internal phase volume and some mycelium entrainment after coalescence were
reported. The successhl use of a supported liquid membrane for extraction of citric
acid from fermentation broth has been reported by Friesen [21] and Rockman [22].
Against this background the main purpose of the work described here was to study
the effects of added proteins, unreacted substrate, and nutrients upon the partition
behaviour of citric acid between aqueous solutions and n-butanol. A comparison of
the partition behaviour of citric acid from a citric acid broth is also presented, and the
effects of insoluble biomass determined independently of the other soluble
fermentation side products.
Materials Used and Experimental Methods
Partition Determination
The partition coefficients were experimentally determined using a simple batch
equilibration technique. Equal volumes (1 5 mL) of the organic phase and aqueous
phase were mixed in a sealed container in a mechanical shaker at constant
temperature of 20.0"C. The maximum shaker-bath speed used was 100 r.p.m. At
equilibrium, the phases were separated and analysed. The equilibration time for a
number of systems was determined by analysing the solute concentration in the
aqueous and organic phases at succeeding time intervals until no fiuther change in the
phase concentrations was observed. Upon equilibration, the samples were allowed to
settle for a period, and then centrifuged at 750g for five minutes to ensure complete
phase separation. Samples from the centre of each liquid layer were removed for
analysis using an automatic air displacement pipette, taking care to avoid any
entrainment of the other phase.
The aqueous-phase feed solutions were made up by appropriate volumetric
dilution of concentrated stock solutions. Automatic air displacement pipettes (error
+/- 1% ) were used to add precise volumes of the solutions for equilibration. In order
to simulate a whole broth extraction system, the additives were introduced into the
aqueous phase prior to the start of the equilibration. Concentrations of the additives
226
The Partition of Citric Acid into n-Butanolfrom Aqueous Solutions
investigated are significantly higher than those which would be present in a broth
system. This was done to exaggerate any effects which the additive may have on the
partition of the solute. The additives studied are listed in Table 1
Table 1. Additives studied.
Additive component
Bovine albumin
Bovine albumin
Trypsin
Sodium lauryl sulphate (SLS)
Sodium lauryl sulphate (SLS)
Glucose
Sucrose
Lactic acid
Magnesium sulphate
Sodium chloride
Range of concentration
(YL)
0-5
0-5
0-5
0-3
0- 1
0- 100
0-50
0-20
0- 100
0-50
Citric acid concentration *
WL)
10
100
10
10
100
10
10
10
10
10
L
(* starting concentration)
Analytical Procedure
The samples of the aqueous phase were diluted five times with water, and the organic
phase samples were diluted five times with 1-butanol (HPLC grade) prior to HPLC
analysis. The dilutions were made using a precision balance (f 0.0001 gram) to
ensure high accuracy. Quantitative analysis for citric acid was achieved using a
Shimadzu liquid chromatograph equipped with an ultraviolet-visible photodiode-array
detector (Model SPD MlOA), an automatic sample injector (SIL-IOA), and a
communications bus module (Model CBM-IOA), and further connected to a DELL
4661 computer installed with Class LClO software (Mason Technology). The system
employed a C-18 reverse phase HPLC 25 cm x 4.6 mm column (Spherisorb
SSODS2). A mobile phase of 0.17M H3P04 : methanol (90:10), was used at a
flowrate of 1.O ml/min. When resolution of citric acid fiom other fermentation broth
components was required, a mobile phase containing only 0.17M H3P04was used at a
flow rate of 1.O ml/min. All citric acid concentrations were expressed in terms of the
monohydrate form of citric acid. The detector monitored the organic acid samples at a
wavelength of 2 12 nm.
22 7
N. Carolan and L.R. Weatherley
Phase Volume Changes
Volume studies were carried out to give an indication of volume changes that may
occur in the various extraction systems examined. From the mutual solubility data it is
evident that when equal volumes of l-butanol and water are mixed and mutually
saturated, the resulting organic phase volume will be larger than the aqueous phase
volume. The effects of additives on phase volumes in l-butanol and water systems
were determined; 50 ml of aqueous phase was contacted with 50 ml l-butanol in a
graduated cylinder and allowed to settle. Results for the citric acid-water-n-butanol
system in the absence of any additives are shown in Table 2. The effect of a range of
additives upon phase volumes after equilibration in the water-citric acid-n-butanol
system was also determined and the results are shown in Table 3.
Citric Acid Fermentation
The Aspergillus niger strain IMI 017454 used for citric acid production was supplied
by The National Collections of Industrial and Marine Bacteria Ltd., Aberdeen,
Scotland. Sporulation of the fungus was achieved in Petri dishes at 25°C using
Czapek Dox Agar (Sigma). This medium resulted in a high yield of viable spores,
which were harvested approximately 10 days after inoculation. Supplementation of
the medium with 1.0% corn steep liquor resulted in faster growth and spore
formation, with spore harvesting possible after 3 to 4 days. Harvested spores were
suspended in sterile water containing sodium lauryl sulphate at a dilution of 1:100000,
and stored at 4°C.
Pre-seed growth of the fungus was performed using the medium shown in Table 4.
All media were sterilised for 25 minutes at 121°C prior to use. The pre-seed medium
(50 ml) was inoculated with approximately lo7 spores (0.75 ml of spore suspension)
in a 250 ml conical flask and incubated at 28°C (200 r.p.m.). Pellets, suitable for
inoculation of the submerged fermentation media, were formed usually after 24 hours.
Table 2. Volume data for water / citric acid/ n-butanol.
Aqueous phase
Aqueous volume %
Organic volume %
Water
46.6
53.3
Citric acid 100 g/L
46
54
228
The Partition of Citric Acid into n-Butanolfrom Aqueous Solutions
Table 3. Volume studies on 1-butanol systems.
Concentration
Citric acid 10 g/L
(km
% aq, vol. change xlQ'
20
+0.7
50
+18.4
100
+19.7
20
-9.9
50
-23.9
100
-26.1
Sucrose
50
-10.9
Glucose
100
+2 .o
Additive
NaCl
MgS04
Table 4. Pre-seed medium composition.
Component
Quantity (g)
Sucrose
6
NH4NO3
0.25
w2po4
0.1
MgS04.7H20
0.025
Agar
Distilled water
0.2
100 ml
Submerged fermentation was performed using a BioFlo 3000 Bench-Top
Fermentor (New Brunswick Scientific Company, Inc., Edison, N.J.) with a 5-litre
fermentation vessel, using the medium shown in Table 5 with an initial pH of 4-5.
The method was an adaptation of that used by Friedrich [23]. The sucrose used in the
fermentation was de-ionised using a cationic ion-exchange resin (Amberlite IRC 718).
After inoculation with the pre-seed pellets, fermentation was allowed to proceed
without aeration using an agitation speed of 200 r.p.m. and at a temperature of 30°C.
229
N. Carolan and L.R. Weatherley
After a period of 18 hours, aeration was started. Aeration rate and speed of
agitation were gradually increased to maximum values of 10 L/min and 500 r.p.m.
respectively, in order to maintain dissolved oxygen levels as the broth viscosity
increased due to biomass production. Fermentation was halted after 10-12 days
(maximum citric acid production).
Table 5. Submergedfermentation medium.
Component
Quantity (gll00 ml)
Sucrose de-ionised
14.0
NH4N03
0.25
m2po4
MgS04.7H20
0.1
cu2+
Zn2+
0.025
0.1
10”
0.001 x
The pH and dissolved oxygen levels of the fermentation were continuously
recorded. Samples of the broth were taken periodically from the fermentation, and
the citric acid concentration was determined by HPLC. It was not necessary to control
the pH during the fermentation because Aspergillus niger is able to produce citric acid
even at low pH [24].
In order to determine the biomass concentration, a sample of a specified volume
was centrifuged at 2000g for 20 minutes until the biomass settled to the bottom of the
centrifuge tube. Most of the liquid was decanted off and a similar volume of distilled
water was added. The solution was re-mixed prior to further centrifugation. This
washing procedure was repeated. Finally the biomass and some of the water were
transferred to a pre-weighed beaker which was placed in an oven at 95°C. Upon
attainment of constant weight, the biomass weight was determined. In order to obtain
reproducible results it was essential that the broth was homogeneously mixed when
the sample was taken. Care was taken to avoid any loss of biomass during the
washing procedure.
230
The Partition of Citric Acid into n-Butanolporn Aqueous Solutions
The concentration of soluble components in filtered broth was determined by
evaporation to dryness of a known volume of sample.
Protein in the fermentation broth was determined by the method of Bradford [25].
This method involves the binding of Coomassie Brilliant Blue G-250 to protein. The
binding of the dye to protein causes a shift in the absorption maximum of the dye
from 465 to 595 nm. The increase in absorption at 595 nm was monitored using a W
spectrophotometer (Perkin Elmer Lambda 12). The assay was found to be very
reproducible and rapid with the dye binding process virtually complete in
approximately two minutes with good colour stability for 1 hour [25]. The broth
properties as measured are shown in Table 6.
The broth samples were dosed to a concentration of 10 g/L citric acid
(monohydrate) prior to measurement of the distribution coefficients against 1-butanol
under the same regime as described above. The distribution coefficients for whole
broth, filtered broth, and the washed biomass suspension were compared against those
obtained using a control solution of citric acid at the same starting concentration.
Table 6. Citric acid broth characteristics.
Broth property
PH
Final value
1.9
Citric acid concentration
0.17 gL
Biomass concentration
18.3 g/L
Soluble components concentration
32.1 g/L
Protein concentration
0.1 gL
Surface tension
29 mN/m
Results and Discussion
Proteins
The addition of two proteins was investigated for any effects on citric acid partition.
The molecular weights of bovine albumin and trypsin are approximately 67,000 and
23,000 [26] respectively. Both are relatively hydrophilic proteins as indicated by their
231
N.Carolan and L.R. Weatherley
aqueous solubility. Proteins are complex macromolecules which comprise amino
acids connected to form long unbranched polymers. Many of the amino acids have
either hydrophobic or hydrophilic side chains, which govern how a protein molecule
will fold to give its characteristic three dimensional structure. Depending on the way
in which the protein molecule folds, it may have a majority of hydrophilic or
hydrophobic groups on its outer surface, thus giving it hydrophilic or hydrophobic
nature [27]. The distribution and degree of hydrophilic and hydrophobic regions
determines the degree to which a protein is soluble in an aqueous environment [28].
Another important factor which contributes to the characteristic solubility of a
protein is its size, with smaller proteins being generally more soluble than chemically
similar proteins of a larger molecular size. The availability of water to solvate the
protein is vital and properties which affect the interactions involved in solvation at the
protein surface will also affect the solubility of the protein. These interactions are
generally regarded as electrostatic in nature and therefore susceptible to changes in
parameters such as solution temperature, pH, dielectric constant and ionic strength
[28]. The presence of organic solvents, such as ethanol and acetone, can precipitate
proteins and can denature them.
Another important aspect of proteins is their tendency to denature in the presence
of electrolytes, e.g. salts [28]. 1-Butanol may denature the proteins used here [28].
The denaturing of a protein depends on the protein used, the concentration of the
protein and the concentration of the denaturant present. It is not known whether the
denatured protein will affect the partition to the same degree as the normal protein.
However, it is expected that much of the protein would be denatured due to the
violent mixing of the samples initially and the presence of citric acid and 1-butanol in
the aqueous phase. This was suggested due to the observation of interfacial crud
present in the protein samples:
Bovine albumin (BSA) clearly has a significant effect on citric acid partition. At
the lower citric acid concentrations, see Figure 1, the distribution coefficient
decreases linearly from 0.335 to almost 0.29 over the concentration range 0 to 5 gk
BSA. At higher citric acid concentrations, shown in Figure 2, the effect of the BSA is
significantly less.
232
The Partition of CitricAcid into n-Butanol from Aqueous Solutions
0.35
- 0.33
Y
;0.31
u
3
0.29
5(a 0.27
E
0.25
0
I
I
I
I
1
2
3
4
5
Bovine albumin concentration glL
Figure 1. The efect of bovine albumin on partition of citric acid at 10 g/L.
Figure 2. The effect of bovine albumin on partition of citric acid at 100 g/L.
233
N.Carolan and L.R. Weatherley
The presence of trypsin decreases the distribution coefficient of citric acid to a
lesser extent, from 0.34 to less than 0.33 (see Figure 3), over a similar concentration
range. It would be expected that citric acid could bind to hydrophilic sites on the
protein, thus making it unavailable for extraction. This would result in a decrease in
the distribution coefficient. The difference between the effects of the two proteins
could be explained as follows:
0
The presence of different hydrophilic groups and also different numbers of these
groups on each molecule will significantly affect the ability of the protein to bind
the solute. This is possibly the most important factor.
0
Differences in isoelectric point (PI) of each protein may also be important. The
surface charge of the protein will depend on pH. At a pH equal to the PI, value the
protein surface will have no charge. For a pH below PI, the surface of the protein
will be positively charged and so may bind with the citric acid or the citrate anion
to a greater degree. The isoelectric points of proteins are generally between 4.5
and 8.5 [29]. The isoelectric points of the two proteins differ with bovine albumin
having a PI value of 4.9 and trypsin having a value of 10.8 [30]. The higher PI
value for trypsin will mean that the protein surface will be more positively
charged at a given pH. So this would indicate that trypsin may be more effective at
binding the solute and thus decrease the distribution coefficient to a greater extent.
0
Steric hindrance may also affect the ability of the solute to bind to the protein.
Steric hindrance will depend on the structure of the protein [28].
8
-5
0.31 --
E
0
-g 0.27 --
n 0 29
5
Figure 3. The effect of ttypsin on partition of citric acid at 10 d L .
234
The Partition of Citric Acid into n-Butanolji-om Aqueous Solutions
For a given mass concentration, there will be more molecules of trypsin present
than bovine albumin due to its lower molecular weight. If the two protein
molecules were equal at binding the solute then trypsin would have a higher effect
on the distribution coefficient than bovine albumin for a given mass concentration.
Volume Changes
It is clear that citric acid addition decreases the aqueous phase volume, as shown in
Table 2. This indicates an increase in the mutual solubilities of the phases, and thus
such an increase would be expected.
The volume changes for a number of additives are listed in Table 3 for citric acid
with an initial aqueous acid concentration of 10 g/L. Sodium chloride increases the
aqueous phase volume in the citric acid system. Magnesium sulphate decreased the
aqueous phase volume in the citric acid system.
The change in phase volume indicates a change in the mutual solubility of the
phases and is likely to affect the partition of a solute. When salt is added it is likely
that 1-butanol is "salted out" from the aqueous phase. Water may also be "salted in
water" fiom the organic phase into the aqueous phase. The different effects of
sodium chloride and magnesium sulphate upon volume change suggests that they may
"salt out" and "salt in" to different degrees, with "salting out" dominating with sodium
chloride and "salting in" dominating with magnesium sulphate, but the reasons for
this difference are unclear.
The effects of sucrose and glucose upon phase volumes were also investigated.
Sucrose decreases the aqueous phase volume in both systems while glucose increases
the aqueous phase volume. Both these observations proved significant in helping to
explain changes to citric acid partitioning in the presence of each of these compounds.
Other Additives
Sodium Lauryl Sulphate
Sodium lauryl sulphate clearly has a strong influence on the distribution coefficient of
citric acid, decreasing it approximately linearly fiom 0.33 to 0.25 over a concentration
0 to 3 g/L(see Figure 4). A smaller decrease (0.374 to 0.365) is observed at higher
citric acid concentrations as shown in Figure 5 . The significant decrease in the
235
N. Carolan and L.R. Weatherley
distribution coefficient indicates a strong interaction between the solute and the
surfactant molecule. The SLS molecule has a considerable aqueous solubility due to
the strong hydrophilic portion of the SLS molecule and so it is expected to reside
primarily in the aqueous phase. The decrease in citric acid partition is assumed to be
caused by the interaction of SLS with the citric acid which is thus less available for
extraction into the organic phase.
Although the SLS molecule is an anionic
surfactant, thus capable of attracting charged ions, only cations would be expected to
be attracted to SLS. It is unlikely that the citrate anion will bind with SLS. It would
however be possible for the citric acid in molecular form to bind with SLS.
-C
o
IE
0.32
0.30
u
E
2 0.28
5
n
'cm, 0.26
6
0.24
0
1
05
15
2
2.5
3
Sodium lauryl sulphate concontration g l l
Figure 4. The efect of sodium lauryl sulphate on partition of citric acid at 10 g/L.
"."""
I
!I
3 0.380
E
0"V
0.375
C
0.370
n
0.365
I
B
0.360 4
0.0
I
1
0.2
0.4
0.6
Sodium lauryl sulphate concentration gL
0.8
1 .o
Figure 5. The efect of sodium lauryl sulphate on partition of citric acid at 100 g/L.
236
The Partition of Citric Acid into n-Butanolfiom Aqueous Solutions
Glucose and Sucrose
Glucose increases the distribution coefficient only slightly, as shown in Figure 6.
Modification of the citric acid distribution by glucose could occur in a number of
ways. It is very likely that glucose will alter the activity of the water in the aqueous
phase which would result in the salting out of some 1-butanol from the aqueous
phase. This would make the aqueous phase less polar and thus would tend to
decrease the distribution coefficient. However, by altering the activity of water in the
aqueous phase, some citric acid would be salted out thus increasing the distribution
coefficient. Hence the overall effect of glucose would depend on the resultant of these
opposing effects.
The sucrose has virtually no effect on the distribution coefficient as shown in
Figure 7. Figure 8 shows explicitly the variation of aqueous and organic phase citric
acid concentrations at equilibrium for a range of sucrose concentrations. Examination
of the data in this form shows that the concentration of citric acid in both the aqueous
and organic phase decreases in an almost identical manner. Another relevant factor
here is the changes in solution volume which occur in the presence of the different
additives (see Table 3), where the effects of a range of additives upon the specific
volume of the aqueous citric acid solutions are indicated, and are expressed as a
percentage volume change. In the case of sucrose addition, a small decrease in the
aqueous phase volume was observed as the sucrose concentration is increased up to
50 g/L (see Table 3).
z0.31
t
0
30.29
---
n
g0.27 --
E
0.25
!
0
20
40
60
80
100
Glucose concentration @I.
Figure 6. The effect of glucose on partition of citric acid at I0 g/L.
23 7
N. Carolan and L.R. Weatherley
0.33
0
Figure 7. The effect of sucrose on partition of citric acid at 10 g/L,
=
0
B
1.85
1.80
1.75
0
10
20
30
Sucrose concentratlong/L
40
50
Figure 8. The effect of sucrose on phase concentrations of citric acid at 10 dL.
This indicates a change in the mutual solubility,of the phases. If citric acid is
transferred to the organic phase during this volume change then, although the aqueous
and organic phase concentrations will both decrease (if a sufficient quantity of citric
acid is transferred), the distribution coefficient may remain constant if the
concentration changes are such that the ratio of the concentrations (the distribution
coefficient) remains the same. This is indicated in Figure 8 where the slopes of the
two lines are approximately equal, and so the percentage change in each phase is the
same. The distribution coefficient for sucrose into 1-butanol is very small, 0.0157
238
The Partition of Citric Acid into n-Butanolfiom Aqueous Solutions
[3 13, and so sucrose will remain in the aqueous phase. Sucrose may alter the activity
of water in the aqueous phase, and so cause citric acid to transfer to the organic phase.
1-Butanol may also be salted out which would explain the volume reduction in the
aqueous phase.
Lactic Acid
Figure 9 shows the effect of lactic acid upon the distribution of citric acid. Since
lactic acid itself distributes significantly into n-butanol, the distribution coefficient for
lactic acid as a function of concentration is also shown on the right hand axis. The
presence of other organic acids would tend to increase the distribution coefficient, and
indeed the addition of lactic acid from 0 to 20 g L increases the distribution
coefficient of citric acid from 0.335 to 0.36. The distribution coefficient of lactic acid
increases although to a lesser extent than for citric acid.
0.42
--
0.40
0-74
D = distribution coefficient
--
1
I
L
'C
0.72
e
-- 0.70
=
V
0.36 --
- - 0.68
S
J
I
-- 0.66 0
1
I
I
;
I
- 0.64
Figure 9. The effect of lactic acid on partition of citric acid at I0 g/L.
Citric and lactic acid are weak organic acids with pKHAlvalues of 3.13 and 3.86
respectively, so citric acid is a stronger acid. Since both acids would be expected to
extract preferentially into n-butanol in undissociated form, a decrease in pH would
increase the partition of weak organic acids The pH of a 10 g/L citric acid solution
may be calculated at a value of 2.52, which is below the pKwl value for both acids.
239
N. Carolan and L.R. Weatherley
On addition of lactic acid, pH will be further reduced. On a mass concentration scale
lactic acid has a similar effect on pH to citric acid.
The sensitivity of the distribution coefficient of the acid to pH is illustrated for
citric acid in Figure 10. Lactic acid is expected to behave similarly except an increase
in the distribution coefficient with pH will occur at a higher pH because of the higher
pKWl value for lactic acid. The curve in Figure 10 for citric acid would be expected
to shift to the right by 0.7 pH units (the difference between the two pKHAlvalues) for
lactic acid. For lactic acid the distribution coefficients will, of course, all be higher
but the partition curve will have the similar shape. Thus at lower pH values the
distribution coefficient of citric acid will be expected to be more sensitive to changes
in pH than for lactic acid.
This is in line with the observations in Figure 9. Another factor is that increase in
overall organic acid concentration tends to increase the mutual solubility of the phases
[12]. This would also be consistent with an increase in distribution coefficient with
increasing organic acid concentration.
.
0.38 1
D
,
3.0
(distribution coefficient)
0.34
n
0.32
0.28
Om3'
n
t
-L
2.1
'
3
1.5
0
10
20
30
40
50
Citric acid concentration g/L
60
Figurelo. The relationship between pH, citric acid concentration andpartition.
Salts
The presence of salts or other electrolytes tend to decrease the mutual solubility of 1butanol and water, and are most likely to be the main factors in the decrease in the
distribution coefficient at the lower concentrations found in broth systems. The effect
240
The Partition of Citric Acid into n-Butanolfiom Aqueous Solutions
of the electrolytes sodium chloride and magnesium sulphate on the distribution
coefficient are illustrated in Figures 11 and 12 respectively.
Figure 11. The effect of sodium chloride on partition of citric acid at I0 g/L.
0.32
0.30
0.28
0.26
0 .a4
0
20
40
60
80
100
Magneslum sulphate concentration gA
Figure 12. The effect of magnesium sulphate on partition of citric acid at I0 g/L.
The addition of sodium chloride has a decreasing effect on the distribution
coefficient of citric acid over the concentration range 0 to 100 g/L (see Figure 11).
The decrease is sharpest over the lower concentrations and any additional effect levels
off at approximately 50 g/L. The overall decrease in the distribution coefficient is
substantial, fiom 0.33 to below 0.25. Other work [32] has shown that the mutual
241
N. Carolan and L.R Weatherley
solubility of 1-butanol and water is reduced upon the addition of sodium chloride.
Water is salted in fiom the organic phase and 1-butanol is salted out from the aqueous
phase. The result is a more polar aqueous phase and a more non-polar organic phase.
This would make it more difficult for the polar citric acid to transfer to the organic
phase. However, because the water molecules become "unavailable" for solvating
citric acid in the aqueous phase on the addition of salt, there are two opposite effects,
one results in a decrease in the distribution coefficient and the other in an increase.
The "salting in" of water from the organic phase appears to be the likely dominant
factor which explains the decrease in partition of citric acid in the presence of sodium
chloride.
The data in Figure 12 show that addition of magnesium sulphate up to a salt
concentration of approximately 20 gK decreases the distribution coefficient of citric
acid from 0.33 to a value of approximately 0.275. For magnesium sulphate
concentrations in the range 20 g/L to 50 g/L, the distribution coefficient remains
almost constant. Beyond 50 g/L, the distribution coefficient increases until at 100 g/L
it reaches almost 0.3 1 as shown in Figure 12. The unusual shape of the curve can be
explained in a manner similar to the effect of sodium chloride. Again there are four
possible factors which affect the partition of citric acid:
The mutual solubility of the phases.
Phase volume changes.
The salting out of citric acid.
The formation of a complex between the metal cation and the citrate anion.
In the concentration range fiom 0 to 20 g/L of magnesium sulphate, the increase in
aqueous phase polarity and the decrease in the organic phase polarity due to a
reduction in the mutual solubility of the phases is likely to be the dominant factor
affecting changes in the citric acid partition as observed. Beyond approx. 30 g/L
magnesium sulphate, it is possible that the mutual solubility of the phases will not be
reduced as significantly and therefore will exert less influence on the distribution
coefficient. The salting out of citric acid could then become more important. A higher
concentration of magnesium sulphate is likely to reduce the amount of water available
for citric acid solvation thus promoting "salting out" into the organic phase.
242
The Partition of Citric Acid into n-Butanolj?om Aqueous Solutions
Broth System
The effects of broth components fiom a sucrose-based Aspergillus niger broth are
shown in Figure 13. The control sample and the washed biomass sample yielded
distribution coefficients higher than 0.325. The whole broth and centrifuged broth
samples show a lower partition than for the control sample with a distribution
coefficient slightly greater than 0.30. No change in the distribution coefficient for the
washed biomass sample and for the control sample indicate that biomass has no effect
on partition. This is confirmed by the whole broth and centrifuged broth samples also
having the same distribution coefficient value.
0.34
E0
0.33 -
I
0.32 -
8
5
0.31
-
Whde
broth
Centrifuged
broth
Washed
biomass
control
Figure 13. The effect of broth components on partition.
The effect of the soluble broth components differentiate the centrifuged broth fiom
the control sample. It is these constituents which appear to cause a decrease in the
distribution coefficient.
The soluble components include the broth medium
components remaining after the fermentation, such as sucrose and salts. Sucrose
breaks down to form fi-uctose and glucose so these sugars may also have been present
[33]. Protein, biosurfactants, and any by-products formed during the fermentation
such as other organic acids would also be present in the broth. A whole broth system
may contain 50-100% more protein than filtered broth [34], and this may influence
partition in whole broth systems.
243
N. Carolan and L.R Weatherley
The presence of biomass increased the error in the values of the distribution
coefficients obtained. Also the presence of the soluble broth components appear to
have a negative influence on the reproducibility of the partition experiments with the
centrifuged broth sample exhibiting a greater variation in the distribution coefficient
values compared to the control sample. The distribution coefficient error for the
control sample was typically 0.33 f 0.004; for the whole broth sample 0.303 f 0.007;
and for the centrifhged broth sample, 0.304 f 0.006. The washed biomass sample
also showed an increased error of 0.327 f 0.006.
Conclusions
The extraction of citric acid into 1-butanol exhibits variable behaviour in the
presence of a number of compounds and components present in fermentation
media and final whole-broth product. The organic acid molecule is a weak acid
capable of dissociation and this has a major influence on the partition of the
organic acid in the presence of additives.
A number of additives, including proteins, were capable of binding to the citric
acid molecule and hence altering the distribution coefficient.
Possible changes in mutual phase solubility are of considerable importance in
determining how citric acid distribution is affected by the presence of a particular
additive.
In this study the presence of biomass solids per se did not appear to influence
partition of citric acid into 1-butanol. The presence of other soluble components
associated with the biomass was very significant and resulted in substantial
reductions in the overall distribution coefficient.
The presence of proteins can have a significant effect upon partition of citric acid.
In an extraction system where the partition of the broth is influenced by protein,
such as the citric acid system studied where the two proteins investigated reduced
partition, it is possible that the whole broth and filtered broth samples may exhibit
different partition behaviour. However, the concentration of these compounds in
the broth may be quite low, with protein having a concentration of O.lg/L in the
244
The Partition of Citric Acid into n-Butanolfrom Aqueous Solutions
citric acid broth system (see Table 6) and so any influence on partition will be
small. However, this was not observed in this study. The presence of ionic
surfactants also exerts a significant influence on the partition equilibrium of citric
acid.
0
The extraction of citric acid in the broth systems has been analysed in terms of the
effect of individual components present in the broth. The decreased partition of
citric acid into I-butanol in the broth system may be explained by the negative
influence of electrolytes at low concentration on the partition of citric acid.
References
1. Schugerl, K. 1994. Solvent Extraction in Biotechnology, Recovery of Primary and S e c o h r y
Metabolites, Springer-Verlag.
2. Akzo Nobel. 1988. Fermentation Production, Extraction Demulsrfiers, Company Brochure.
3. Weatherley, L.R., Allan, G., and Goodwin, J.A.S. 1993. Interfacial Mass Transfer and Partition Studies
During Whole Broth Extraction, Proceedings of the International Solvent Extraction Conference,
Vol.1, 1003-1009.
4. Stuckey, D.C. 1996. Solvent Extraction in Biotechnology, Proceedings of the International Solvent
Extraction Conference, Vol. I, 25-34.
5. Liang T.B., and Slater, M.J. 1990. Liquid-liquid extraction drop formation: mass transfer and the
influence of surfactant. Chem. Eng. Sci., 45(1), 97-105.
6. Agble, D., and Mendes-Tatsis, M.A. 1996. Mass Transfer with lnterfacial Convection and Added
Surfactants, Proceedings of the International Solvent Extraction Conference, Vol. 1,267-272.
7. Crabbe, P.G., Tse, C.W., and Munro, P.A., 1986. Effect of Microorganisms on Rate of Liquid
Extraction of Ethanol from Fermentation Broths, Biotechnology and Bioengineering. 28(7), 939-943.
8 . Weatherley, L.R., Doherty-Spiers, E.A., Goodwin, J.A.S., and Slaughter, J.C., 1994. A Study of
Biomass Morphology on Pattition in the System PropanollWatertln-Decanol in the Presence of
Aureobasidium pullulans. Separations for Biotechnology Volume 3, ed. D.L. Pyle, SCl Publishing,
539-545.
9. Lennie, S., Halling, P.J., and Bell, G. 1990. Causes of Emulsion Formation During Solvent-Extraction
of Fermentation Broths and its Reduction by Surfactants Biotechnology and Bioengineering, 35(9),
948-950.
10. Thornton, J.D. 1992. Science and Practice of Liquid-Liquid Extraction. Volume 2, Clarendon Press.
11. Blanch, H.W., and Clark, D.S. 1996.Biochemical Engineering, Marcel Dekker, New York.
12. Kertes, A.S., and King, C.J., 1986. Extraction Chemistry of Fermentation Product Carboxylic-Acids.
Biotechnology and Bioengineering.28(2), 269-282.
13. Bar, R.,and Gainer, J.L. 1987. Acid Fermentation in Water-Organic Solvent 2-Liquid Phase Systems,
Biotechnol. Prog., 3(2), 109-1 14.
14. Colin, P. 1967. The Recovery of Citric Acid Using Extraction, French Patent 555,810.
15. Sporidi, G. 1964. Process for Organic Acid Recovery, German Patent No. 1,493,196.
16. Les Usines de Melle. 1961. Process for Extracting Citric Acid from Aqueous Solutions Thereof,
French Patent No. 874,030.
17. Wennersten, R., 1980. New Method for the Purification of Citric Acid by Liquid-Liquid Extraction.
Proceedings of the International Solvent &traction Conference, 8 , 1-7.
18. Yu-Ming, J., Dao-Chen, L., and Yuan-Fu, S. 1983. Study on Extraction of Citric Acid, Proceedings of
the International Solvent Extraction Conference, Vol. 2, 5 17-5 18.
19. Boey, S.C., Delcerro, M.C.G., Pyle. D.L.(1987). Extraction of Citric-Acid by Liquid Membrane
Extraction, Chem. Eng. Res. Des., 65(3), 218-223.
245
N. Carolan and L.R. Weatherley
20. Chaudhuri, J., and Pyle, D.L. 1987. In: Verrall M.S., and Hudson, M.J. (eds.), Separations for
Biotechnology,Volume 1, Ellis Horwood Ltd., 241-259.
21. Friesen, D.T., Babcock, W.C., and Chambers, A.R. 1986.Separation of Citric-Acid from Fermentation
1Papers
of the American Chemical Sociefy,
Beer using Supported-Liquid Membranes, Abstracts 0
.
.
191(4), lfi-195:
22. Rockman, J.T., Kehat, E., and Lavie, R. 1996.Thermally Enhanced Liquid-Liquid Extraction of Citric
Acid using Supported Liquid Membranes. Proceedings of the International Solvent Extraction
Conference. 2.857-862.
23. Frihrich, J:, Cimerman, A,, and Steiner, W. 1994.Concomitant Biosynthesis of Aspergillus-niger
Pectolytic Enzymes and Citric-Acid on Sucrose, Enzyme and Microbid Technology, 16(8), 703-707.
24. Grewal, H.S., and Kalra, K.L. 1995.Fungal Production of Citric Acid, Biotechnologv Advances, 13(2),
209-234.
25. Bradford, M.M. 1976.A Rapid and Sensitive Method for the Quantitation of Microgram Quantities of
Protein Utilizing the Principle of Protein-Dye Binding, Anafytical Biochemistry, 72.248-254.
26. Fluka Chemica-Biochemika Analytika Catalogue. 1995.Fluka Chemicals, Donet, England.
27. Moody, P.C.E., and Wilkinson, A.J. 1990.Protein Engineering. Oxford University Press.
28. Weatherley, L.R. 1994.Engineering Processesfor Bioseparations, Butterworth-Heinemam
29. Verrall, M.S. 1996.Downstream Processing ofNatural Products, A Practical Handbook, John Wiley.
30. Oxender, D.L., and Fox, C.F. 1987. Protein Engineering, Liss.
31. Weiser, R.B., and Geankoplis, C.J. 1955.Lactic Acid Purification by Extraction, Ind. Eng. Chem.,
47(4), 858-863.
32. Kertes, AS., and King, C.J. 1987. Extraction Chemistry of Low-Molecular Weight AliphaticAlcohols, Chemical Reviews, 87(4), 687-710.
33. Grewal, H.S., and Kalra, K.L. 1995.Fungal Production of Citric-Acid. Eiorechnologv Advances, 13(2),
209-234.
34. Likidis, Z.,Schlichting, E., Bischoff, L., and Schugerl, K. 1989.Reactive Extraction of Penicil1in-G
from Mycel-Containing Broth in a Countercurrent Extraction Decanter, Biotechnologv and
Biwngineering, 33(1 l), 1385-1392.
246
Документ
Категория
Без категории
Просмотров
0
Размер файла
974 Кб
Теги
acid, partition, solutions, effect, citrix, butyanol, protein, aqueous, impurities, brot
1/--страниц
Пожаловаться на содержимое документа