Recovery of Nicotinic Acid from Aqueous Solution using Reactive

Extraction with Tri-n-Octyl Phosphine Oxide (TOPO) in Kerosene

Sushil Kumar1, Karan Gupta2, and B V Babu*
Birla Institute of Technology and Science (BITS), PILANI 333 031 (Rajasthan) India
1

Lecturer, Department of Chemical Engineering

E-mail: [email protected];
Homepage: http://discovery.bits-pilani.ac.in/Homepage/disciplines/chemical/skumar.asp
2

Student (ME), Department of Chemical Engineering

E-mail: [email protected]

*Dean-Educational Hardware Division & Professor of Chemical Engineering

E-mail: [email protected]; Phone:+91-1596-245073 Ext. 259; Fax: +91-1596-244183
Homepage: http://discovery.bits-pilani.ac.in/~bvbabu

Abstract
Nicotinic acid (also known as pyridine-3-carboxylic acid) is widely used in food,
pharmaceutical and biochemical industries. The extraction of nicotinic acid by tri-noctylphosphine oxide (TOPO) with kerosene (diluent) is studied. The equilibrium
experiments are carried out to investigate the effects of diluent, extractant (TOPO)
composition and initial acid concentration on extraction efficiency. The extraction
efficiency is found to increase with increasing TOPO composition concentration (0.10 to
0.60 mol.L-1) and found to decrease with increasing initial acid concentration (0.02 to
0.12 mol.L-1). The number of TOPO molecules in the acid:TOPO complex is estimated
through proposed mathematical model using graphical methods as well as though an
optimization route using the differential evolution algorithm and found to be 1:1
complexes of TOPO and acid with loading ratio of less than 0.5. The extraction
equilibrium constants are also determined. Equilibrium extraction constant (KE1) is found
to be 3.27 L.mol-1 in the complete range of acid concentrations and extractant (TOPO)
concentrations.

Introduction
Niacin, also known as nicotinic acid or vitamin B3, is a water-soluble vitamin whose
derivatives such as NADH (reduced form of NAD) play essential roles in energy
metabolism in the living cell. Nicotinic acid (3-pyridine carboxylic acid) widely used in
food, pharmaceutical and biochemical industries is an important chemical, mainly
obtained by chemical synthesis, using 3-picoline or 2-methyl-5-ethyl-pyridine as startingmaterials, at high temperature and pressure. Besides the technical aspects, other
parameters such as desired quality, physical and chemical properties of the final product,
and the ecological problems complicate the chemical synthesis methods. Due to these
reasons, the chemical synthesis route for nicotinic acid production will become
unattractive in the future. In recent years, the application of enzymes to organic chemical
processing has attracted the attention of researchers. Nitrilases enzymes are gaining
popularity as biocatalysts for the mild and selective hydrolysis of nitriles. The production
of nicotinic acid and nicotinamide can be intensified by enzymatic conversion of 31

cyanopyridine or biosynthesis (Kumar and Babu, 2009). Very recently amidase-catalyzed

production of nicotinic acid in batch and continuous stirred membrane reactors has been
studied by Cantarella et al (2008). Amidase enzyme, operated under mild conditions is
suitable for the synthesis of labile organic molecules and it is stable up to 50 C. This
fermentation process, because of various impurities and very low concentration of
product in the fermentation broth, requires an economic separation method to compete
with the synthetic process.
Many separation processes such as liquid extraction, ultra filtration, electro-dialysis,
direct distillation, liquid surfactant membrane extraction, anion exchange, precipitation
and adsorption in chemical industries have been employed to recover the organic acids
from aqueous solution. Among various available alternate processes for simultaneous
removal of the product, extraction is often the most suitable one. So a reactive extraction
method has been proposed to be an effective primary separation step for the recovery of
bio-products from a dilute fermentation process (Kumar et al, 2008).
Organophosphorus compounds and long-chain aliphatic amines are effective extractants
for the separation of carboxylic acids from dilute aqueous solution (Kertes and King,
1986). Phosphorus-bonded, oxygen-containing extractants have a phosphoryl group and a
stronger Lewis basicity than those of carbon-bonded oxygen-containing solvents.
Phosphorus-bonded, oxygen-containing extractants can only co-extract small amounts of
water, and show low solubilities in water. When organophosphorus extractants are used,
the solvation has a higher specificity. The distribution of nicotinic acid between water
and Alamine 300 (tri-n-octylamine) dissolved in polar and non-polar diluents, is studied
at 298 K using a phase ratio of 1:1 (v/v) by Senol (2002). The comparative study of the
reactive extraction of nicotinic acid with Amberlite LA-2 (lauryl-trialkyl-methylamine)
and di-(2-ethylhexyl)-phosphoric acid (D2EHPA) has been presented by Cascaval et al,
2007. Compared to D2EHPA, the use of Amberlite LA-2 allows the possibility to reach
higher extraction efficiency, the extraction degree being supplementarily increased by
increasing the solvent polarity. Kumar et al (2008) studied reactive extraction of nicotinic
acid with TBP and TOPO at a fixed initial acid concentration to intensify the recovery
from fermentation broth. The aim of the present study is to provide the extraction
equilibrium data on the reactive extraction of nicotinic acid (3-pyridine carboxylic acid)
with tri-n-octyl phosphine oxide (TOPO) as an extractant dissolved in kerosene for
intensification of nicotinic acid production via enzymatic route. An equilibrium model
based on mass action law is presented and used to determine the equilibrium extraction
constant (KE) and the number of extractant molecules per acid molecule (n) with
graphical method as well as an optimization procedure. Population based search
algorithm, differential evolution is used as optimization algorithm.

Theory
The extraction of nicotinic acid occurs by means of a successful competition of the
solvating phosphoryl group in TOPO against water molecules between the aqueous and
organic phase. The physical extraction of acid is neglected due to strong interactions
between acid and extractant (TOPO) molecules. The extraction mechanism of nicotinic

acid (HNc) using TOPO as an extractant with various diluents is described by Eq. (1),
showing equilibrium in the formation of complexes between acid and extractant:
HNc aq + nS org HNcS n (org)
(1)
The extraction equilibrium constant, KE, can be calculated using Eq. (2):
[HNcS n (org) ]
KE =
[HNc aq ][S org ] n

(2)

The experimentally accessible distribution coefficient, KD, is calculated with Eq. (3).
C HNc
[HNcS n (org) ]
KD =
=
(3)
C HNc [HNc aq ] + [ Nc aq ]
where [HNcSn (org) ] refers to the concentration of acid in organic phase; [HNc aq ] and
[ Nc aq ] symbolize non-dissociated and dissociated concentrations of acid in aqueous

phase respectively at equilibrium.

Using the values of [HNcS n (org) ] from Eq. (2) and [ Nc aq ] from dissociation constant
(pKa) in Eq. (3) results Eq. (4)
KD =

K E [S org ] n

(4)

K
1 + +a
[H ]

or

K
(5)
log K D + log1 + H+ = log K E + n log[S org ]
[H ]
where, [S org ] is the free TOPO concentration in the organic phase, represented as:

[S org ] = [S org ]in n[HNcSn (org) ]

(6)

Putting the value of [S org ] from Eq. (6) in the Eq. (5) results Eq. (7)

K
log K D + log1 + H+ = log K E + n log [S org ]in n[HNcS n (org) ]
[H ]

(7)

Due to apparition of n under logarithm, an optimization route for estimation of n and KE

may also be applied.
The extent to which the organic phase (extractant and diluents) may be loaded with
nicotinic acid is expressed by the loading ratio, Z (ratio of total acid concentration in the
organic phase to the total TOPO concentration) as given by Eq. (8)
Z=

C HNc
[S org ]in

(8)
3

The stoichiometry of the overall extraction equilibrium is dependent on the loading ratio
in the organic phase, Z. If the organic phase is not highly concentrated by acid, i.e., at
very low loading ratios (Z < 0.5), 1:1 complex of acid and extractant (TOPO) is formed.
A plot of Z/(1-Z) versus [HNc] yields a straight line with a slope of complexation
constant (KE1) as given by Eq. (9):
Z
= K E1 [HNc]
(9)
1 Z

Experimental
Materials
Tri-n-octyl phosphine oxide (TOPO), a phosphorus-bonded oxygen donor supplied by
Sigma-Aldrich Co. USA, with purity of 99% is a white powder with the molar mass of
386.65 g.mol-1. Nicotinic acid, a white powdered material as analytical grade reagents
(99.5%) is procured from HIMEDIA, India. Kerosene (diluent) of commercial grade is
used. Organic solutions are prepared using TOPO as extractant dissolved in the diluent.
The initial TOPO concentrations are varied in the range of 0.10 to 0.60 mol.L-1. Deionized water is used to prepare the aqueous solutions of various concentrations of
nicotinic acid. Since acid concentration in the fermentation broth is found very low, the
aqueous solutions of acid are prepared in the range of (0.02 to 0.12) mol.L-1. All
chemicals are used without any pretreatment.
Procedure
The extraction equilibrium experiments are carried out at constant temperature (298 K)
with equal volumes (16 cm3) of the aqueous and organic solutions shaken at 100 rpm for
8 hours in conical flasks of 100 mL on a temperature controlled reciprocal shaker bath.
This mixing time could be considered as the appropriate time for attaining equilibrium
based on our preliminary studies. After attaining equilibrium, the phases are brought into
contact with each other using the separating funnel of 125 mL for separation. The
concentration of acid in the aqueous phase is determined using an UV spectrophotometer
(Systronics, 119 model, 262 nm). The acid concentration in the organic phase is
calculated by mass balance. The initial and equilibrium pH values of aqueous solutions
are measured using a digital pH-meter of Arm-Field Instruments (PCT 40, Basic Process
Module) which varied in the range of (2.45 to 2.92) and (3.41 to 3.94) respectively. The
reproducibility is checked by carrying out the experiments twice in some selected cases.
The results are found to be reproducible within 5 %.

The extraction process is analyzed by means of the degree of extraction and distribution
coefficient. The distribution coefficient, KD, is calculated using Eq. (3). The degree of
extraction is defined as the ratio of acid concentration in the extracted phase to the initial
acid concentration in aqueous solution by assuming no change in volume at equilibrium
as given by Eq. (10).

E=

KD
100
1+ KD

(10)

Results and Discussion

The physical extraction of nicotinic acid in pure kerosene is remarkably small with a
value of distribution coefficient of about 0.017 as given by Kumar et al, (2008). It is
extracted by phosphorous-bonded oxygen-bearing extractants with a significantly higher
distribution ratio than carbon-bonded oxygen-bearing extractants under comparable
conditions. Tri-n-octyl phosphine oxide (TOPO) is used as extractant to study the
extraction equilibria of nicotinic acid because of its excellent chemical stability, higher
basicity and low solubility in water. In the case of TOPO, the alkoxy groups as found in
TBP are substituted by alkyl groups. So, the Lewis basicity is increased through inductive
effects. The isotherms for nicotinic acid are determined from five aqueous solution
concentrations, four concentrations of TOPO dissolved in kerosene as shown in Figure 1.
The concentration of components is expressed in molar units (mol.L-1). For a higher
range of TOPO concentration, there is a linear relationship between acid concentration in
the two phases, and slightly nonlinear relationship for lower concentrations of TOPO. It
may be noted that for low concentrations of acid with respect to higher range of TOPO
concentrations, Henrys law is valid.
0.08
-1

0.60 mol.L
-1
0.40 mol.L
-1
0.25 mol.L
-1
0.10 mol.L

[HNc]org

0.06

0.04

0.02

0.00
0.000

0.025

0.050

0.075

0.100

[HNc]aq

Figure 1. Equilibrium isotherms of nicotinic acid for different concentrations of TOPO

dissolved in kerosene
The degree of extraction (E) is found to increase with an increase in TOPO concentration
(0.10 to 0.60 mol.L-1) as shown in Fig (2) with different concentrations of acid (0.02,
0.05, 0.08 and 0.12 mol.L-1). Initial concentration of nicotinic acid also affects the
extraction efficiency as shown in Fig (3). The degree of extraction (E) decrease when the
concentration of acid is increased from (0.02 to 0.12) molm-3. Different concentrations of
extractant (TOPO) have been used to derive the effect of initial acid concentration on
extraction efficiency. From Fig. (3), the effect of acid concentration on the degree of
extraction is more significant at the low concentration of TOPO. The trends in
experimental results of this study on equilibrium concentrations of acid in aqueous phase
are in good agreement with the results reported by Kertes and King (1986).
5

Degree of Extraction (%)

60

50

40

30
-1

0.02 mol.L
-1
0.05 mol.L
-1
0.08 mol.L
-1
0.12 mol.L

20

0.1

0.2

0.3

0.4

0.5

0.6

0.7

-1

TOPO Concentration, mol.L

Figure 2. Effect of TOPO concentration on degree of extraction with different acid

concentration

Degree of Extraction (%)

60

50

40

30
-1

0.60 mol.L
-1
0.40 mol.L
-1
0.25 mol.L
-1
0.10 mol.L

20

0.05

0.10

0.15
-1

Acid Concentration, mol.L

Figure 3. Effect of initial acid concentration on degree of extraction with different TOPO
concentration
For the estimation of equilibrium extraction constant (KE) and the number of extractant
(TOPO) molecules per acid molecule (n), the mass action law modeling approach is used
as performed in the section of Theory. A plot of equation (6) by taking,
K

log K D + log1 + +a on y-axis and log[S org ] on x-axis yields the straight line with a
[H ]
slope of n and intercept of log KE as shown in Fig. 4. For the extraction equilibrium of
6

nicotinic acid with TOPO, the slopes of the straight lines suggest the formation of
complex between one molecule each of both the reactants in all cases of acid
concentration. The results indicate a stoichiometric association between the individual
phosphoryl group and individual acid group, and display the effect of acid concentration
on the experimentally determined distribution ratio. From Figure 4 and Table 1, it can be
seen that there is not a significant difference between the slopes and intercepts for
different initial acid concentrations except very low concentration of acid (0.02 mol.L-1).
At very low concentration of acid, kerosene molecules significantly compete with water
molecules for the physical extraction.

log KD + log (1+KH/[H ])

0.2

0.0

-0.2

-1

-0.4

0.02 mol.L
-1
0.05 mol.L
-1
0.08 mol.L
-1
0.12 mol.L

-0.6
-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

[S]org

Figure 4. Determination of extraction constants (KE) and apparent number of reacting

molecules (n) using TOPO dissolved in kerosene with different initial acid concentration
The values of equilibrium extraction constant (KE) and the number of reacting extractant
molecules (n) are also estimated using an optimization routine. A Population based
search algorithm, differential evolution (DE), which is simple and robust and has proven
successful record (Babu, 2004; Babu and Munawar, 2007), is also employed to solve the
model equation (7) for estimation of extraction equilibrium constants (KE) and the
number of reacting extractant molecules (n). An objective function based on least square
K

error between experimental data and predicted value of log K D + log1 + +a has been
[H ]
minimized. The values of equilibrium extraction constant (KE) and the number of reacting
extractant molecules (n) determined by this computational procedure are also given in
Table 1.

Table 1. Values of Equilibrium Extraction Constant (KE) and Number of Reacting

Extractant (TOPO) Molecules (n) with Different Concentrations of Nicotinic Acid.
[HNc]in
/mol.L-1

0.02
0.05
0.08
0.12

Using graphical method (Figure 4)

Using DE (optimization
procedure

KE

R2

KE

2.40
2.43
2.35
2.37

0.74
0.82
0.82
0.83

0.99
0.99
0.99
0.99

2.47
2.36
2.53
2.10

0.89
0.82
0.83
0.65

The obtained results of KE and n, for TOPO extractant using computational procedure
shows the same trend as obtained by graphical method using 1:1 complex, but more
accurate. The diluents, having dielectric constant values and solvent polarity also
contribute to the extraction of organic acid and results in a less than one value of
stoichiometry coefficients. The dielectric constant may be considered as an indicator of
solventsolute local interactions, inducing the limitation of solute solvation by solvent or
extractant, due to the presence of ionizable groups in the solute chemical structure. The
values of coefficient of determination (R2) near about 0.99 are showing minimal
deviation in the results.
Since nicotinic acid is used in the low concentration range of 0.02 to 0.12 mol.L-1and
TOPO is diluted in the range of 0.10 to 0.60 mol.L-1, the loading ratio is found to be very
low (Z < 0.5). 1:1 complexes of acid and TOPO are formed and Z/(1 - Z) versus [HNc]
are plotted for complete range of TOPO concentrations to obtain the value of equilibrium
complexation constant (KE1) as shown in Figure 5. Equilibrium extraction constant (KE1)
for 1:1 complex of acid and TOPO at 298 K for the extraction of acid are given in Table
2.

0.4

Z/(1-Z)

0.3

0.2

0.1

0.0
0.00

0.02

0.04

0.06

[HNc]aq, mol.L

0.08

0.10

-1

Figure 5. Plot of Z/(1 - Z) versus [HNc] for the estimation of (1:1) nicotinic acid-TOPO
equilibrium complexation constant (KE1) in the entire range of TOPO concentration.
Table 2. Equilibrium extraction constant (KE1) for the extraction of nicotinic acid
(0.02 0.12 mol.L-1) in the entire range (0.10 - 0.60 mol L-1) of TOPO from Figure 5
Extractant system

KE1 (mol-1m3)

R2

SD

TOPO + kerosene

3.270.11

0.97

0.02

Conclusions
The studies on reactive extraction of nicotinic acid with tri-n-octyl phosphine oxide
(TOPO) dissolved in kerosene at various TOPO and acid concentrations indicated that the
reactive extraction occurs by means of the interfacial formation of solvates between acid
and extractant. The experimental data clearly show that the chemical extraction using
organophasphorus compound (TOPO) is far better than the physical extraction.
The distribution coefficients (KD) and degree of extraction (E) are found to increase with
an increase in TOPO concentration (0.10 to 0.60 mol.L-1) and found to decrease when the
concentration of acid is increased from 0.02 to 0.12 mol.L-1. Different parameters like
distribution coefficient, degree of extraction, loading ratio, and equilibrium complexation
constants are determined. Chemical modeling approach is used for the determination of
the equilibrium extraction constant (KE) and the number of extractant reacting molecules
(n), the estimated values of KE and n depend on the applied method. Optimum values of
KE and n have been found when optimization procedure (differential evolution algorithm)
is used to solve the model equations. Since the loading ratio was less than 0.5 in all the
cases, no overloading was obtained and only 1:1 complexes of acid and TOPO were
9

formed using graphical method and differential evolution algorithm. With the assumption
of n = 1, the maximum equilibrium extraction constant is found to be 3.27 L.mol-1 in the
complete range of acid concentrations and extractant (TOPO) concentrations.
References
Babu, B.V. (2004) Process plant simulation, Oxford University Press, India.
Babu, B.V.; Munawar, S.A. Differential Evolution Strategies for Optimal Design of
Shell-and-Tube Heat Exchangers. Chem. Eng. Sci. 2007, 62(14), 3720.
Cantarella, M., Cantarella, L., Gallifuico, A., Intellini, R., Kaplan, O., Spera, A.,
Martnkov, L. (2008) Enzyme Microb. Technol. 42, 222.
Cascaval, D., Galaction, A.I., Blaga, A.C., (2007) Camarut, M. Comparative Study on
Reactive Extraction of Nicotinic Acid with Amberlite LA-2 and D2EHPA. Sep. Sci.
Technol. 42, 1-13.
Kertes, A.S., King, C. (1986) Extraction Chemistry of Fermentation product Carboxylic
Acids. Biotechnol. Bioeng. 28, 269-282.
Kumar, S., Babu, B.V. (2008) Process intensification for separation of carboxylic acids
from fermentation broths using reactive extraction, J. Fut. Eng. Technol. 3, 19.
Kumar, S.; Wasewar, K.L.; Babu, B.V. (2008) Intensification of Nicotinic Acid
Separation using Organophosphorous Solvating Extractants: Reactive Extraction.
Chem. Eng. Technol. 31, 1584-1590.
Kumar, S.; Babu B.V. (2009) Process Intensification of Nicotinic Acid Production via
Enzymatic Conversion using Reactive Extraction. Chem. Biochem. Eng. Q. accepted
(In press).
Senol, A. (2002) Extraction Equilibria of Nicotinic Acid using Alamine 300/Diluent and
Conventional Solvent Systems. Turk. J. Chem. 26, 77.

10