Stoichiometric and Spectroscopic Study of Reactive Extraction of Phenylacetic Acid with Trin-Butyl Phosphate

Phenylacetic acid, also known as benzeneacetic acid (IUPAC Name: 2-phenylacetic acid), is an organic compound containing a phenyl ring and an acidic functional group. It has a white scale-like crystal appearance, and a honey-like odour at low concentration. Phenylacetic acid is used as a precursor in the production of penicillin G (as shown in Fig. 1),1,2 β-lactam, and amphetamine. Due to multifunctional biological and medicinal activities of phenylacetic acid, it is necessary to recover it from the effluent streams. Phenylacetic acid can also be produced by strains of Bacteroides asaccharolyticus and Bacteroides melaninogenicus subspecies isolated from human and animal sources.3 Fermentation technology for the production of carboxylic acids in the form of aqueous solutions has been known in the last few decades. There are various severe inhibiting effects on the rate of conversion and recovery methods from fermentation broths. Several separation techniques such as adsorption4–7, distillation8, electro-dialysis9,10, ion-exchange11, liquid surfactant membrane extraction12,13, liquid-liquid extraction14, precipitation15, reverse osmosis16,17 and ultrafiltration18,19 have been reported in the literature, but all these methods have inherent drawbacks. Calcium hydroxide precipitation has a few shortcomings such as consumption of large quantities of reagents (H2SO4 and lime), a large amount of waste generation per ton of acid produced, waste disposal problems of waste, and very poor sustainability. Dialysis has good potential but its drawbacks involve frequent cleaning requirement, membrane fouling, and a requirement of a larger dialysis unit as compared to a fermenter. Higher power consumption is the main problem with electrodialysis, although it allows simultaneous separation and concentration of the acid. Ion-exchange requires a large amount of chemicals, and generates a large amount of waste. The distillation method is a well-established technology, but its drawbacks involve formation of high-boiling internal esters, dimers, and greater power consumption.20–25 Reactive extraction with the proper selection of diluents and extractants can provide high selectivity and extraction but suffers from toxicity problems of solvents toward microbial strains. Selection of an extractant and diluent for reactive extraction should be on the basis of minimal toxicity and maximum capacity. Stoichiometric and Spectroscopic Study of Reactive Extraction of Phenylacetic Acid with Tri-n-Butyl Phosphate


Introduction
Phenylacetic acid, also known as benzeneacetic acid (IUPAC Name: 2-phenylacetic acid), is an organic compound containing a phenyl ring and an acidic functional group.It has a white scale-like crystal appearance, and a honey-like odour at low concentration.Phenylacetic acid is used as a precursor in the production of penicillin G (as shown in Fig. 1), 1,2 β-lactam, and amphetamine.Due to multifunctional biological and medicinal activities of phenylacetic acid, it is necessary to recover it from the effluent streams.
Phenylacetic acid can also be produced by strains of Bacteroides asaccharolyticus and Bacteroides melaninogenicus subspecies isolated from human and animal sources. 3Fermentation technology for the production of carboxylic acids in the form of aqueous solutions has been known in the last few decades.There are various severe inhibiting effects on the rate of conversion and recovery methods from fermentation broths.Several separation techniques such as adsorption [4][5][6][7] , distillation 8 , electro-dialysis 9,10 , ion-exchange 11 , liquid surfactant membrane extraction 12,13 , liquid-liquid extraction 14 , precipitation 15 , reverse osmosis 16,17 and ultrafiltration 18,19 have been reported in the literature, but all these methods have inherent drawbacks.Calcium hydroxide precipitation has a few shortcomings such as consumption of large quantities of reagents (H 2 SO 4 and lime), a large amount of waste generation per ton of acid produced, waste disposal problems of waste, and very poor sustainability.Dialysis has good potential but its drawbacks involve frequent cleaning requirement, membrane fouling, and a requirement of a larger dialysis unit as compared to a fermenter.Higher power consumption is the main problem with electrodialysis, although it allows simultaneous separation and concentration of the acid.Ion-exchange requires a large amount of chemicals, and generates a large amount of waste.1][22][23][24][25] Reactive extraction with the proper selection of diluents and extractants can provide high selectivity and extraction but suffers from toxicity problems of solvents toward microbial strains.Selection of an extractant and diluent for reactive extraction should be on the basis of minimal toxicity and maximum capacity.
Among the various available alternative processes for simultaneous removal of the product, liquid-liquid extraction is often the most suitable.][28] Extraction of carboxylic acid from dilute aqueous solution by solvent extraction has in the last few decades been drawing much attention in the scientific community.Organic solvents used for extraction can be classified into three major categories: (a) Conventional oxygen-containing and organic solvents such as alcohols, glycol, ethers, esters, ethyl acetate, ketones etc. 29 (b) Phosphorous-bonded oxygen containing extractants like trin-butyl phosphate (TBP) 30 , trialkylphosphine oxide (TOPO) etc. 31 ; (c) Higher molecular weight aliphatic amines such as tri-n-octyl amine, alamine 336, aliquat 336 etc. 29,30,32 .For extraction of carboxylic acids, the first types of solvents provide low distribution coefficient because of high affinity of acid molecule towards water molecules, whereas the second and the third types of solvents give higher distribution coefficients in organic phase.
][38] In the present work, the data on stoichiometric and spectroscopic analysis of reactive extraction of phenylacetic acid with tri-n-butyl phosphate at interface of aqueous and organic phase the data are presented in terms of extraction efficiency, distribution coefficient, loading factor, and overall equilibrium constant.Tri-n-butyl phosphate was used as extractant while methyl isobutyl ketone and petroleum ether as diluents.

Materials
Tri-n-butyl phosphate, methyl isobutyl ketone, petroleum ether (mixture of volatile aliphatic hydrocarbon and primarily pentane and isohexane), phosphoric acid, and acetonitrile were procured from Merck, Germany, whereas phenylacetic acid with 98.5 % purity was supplied by Acros Organics, Belgium.All the chemicals were used as supplied.Physicochemical properties of the extractant and diluents are listed in Table 1.

Method
Extraction experiments were conducted in 100 mL Erlenmeyer flasks at room temperature 298 ± 0.5 K.The initial concentration of phenylacetic acid was 0.099 mol L -1 .The reactive extraction was carried out with tri-n-butyl phosphate in methyl isobutyl ketone, and petroleum ether.Tri-n-butyl phosphate concentration in organic solution was varied in the range of 10-60 v/v % (0.366-2.199 mol L -1 ).The volumetric ratio of the aqueous and organic phase was kept as 1:1 (15 mL of each phase).Initial pH of aqueous phenylacetic acid was measured using pH meter (Lab India, Mumbai) and observed as 2.767.The equilibrium pH was found to be different from the initial pH because of the removal of acid from aqueous phase to organic phase.
The flasks containing the mixture of aqueous and organic solution was shaken for 12 hours in a temperature-controlled water bath shaker RSB-12 (REMI Laboratory Instruments, Mumbai, India) and the solution was allowed to settle for 2-3 hours at room temperature (298 ± 0.5 K) and atmospheric pressure (101324.99Pa).The upper layer (organic phase) was used for FTIR study, and the bottom layer (aqueous phase) was taken for pH measurement and acid concentration analysis.The organic phase volume increased by approximately 5 %, with a corresponding decrease in the aqueous-phase volume, which could be due to the water transfer (water co-extraction) into the organic phase to solvate the complex 39 and the results are presented in Table 5.

Liquid chromatographic analysis
Phenylacetic acid concentrations were determined using High Performance Liquid Chromatography technique (Agilent 1200, California USA).Analysis was performed with Eclipse XDB-C18 (4.6 mm ID × 250 mm, 5 μm) column.Acetonitrile (25 %) and 20 mmol L -1 of aqueous solution of phosphoric acid (75 %) was used as a mobile phase and the flow rate was set at 1 mL min -1 .The detector (DAD) wavelength was set at λ = 215 nm.The column was operated at 308 K with the injection volume of 5 μL.Before injecting, the samples were filtered through a syringe filter with pore size 0.2 μm PVD filter media supplied by WHATMAN, U.S.A. Retention time of pure phenylacetic acid and extracted samples was found to be about 6.3 minutes (see Fig. 2).Each analysis was done in triplicate under identical conditions, and the concentration was reported as the average value.The maximum experimental error was observed as below ± 5 %.

Stoichiometry of complex formation of phenylacetic acid and tri-n-butyl phosphate
Using mass action law 40 , the extraction equilibrium for phenylacetic acid can be written as a reaction of one molecule of phenylacetic acid and γ molecule of tri-n-butyl phosphate participating in the formation of the complex.The mechanism of reactive extraction of phenylacetic acid in terms of apparent equilibrium constant, E can be written as: where, HA and T is denoted by phenylacetic acid and tri-n-butyl phosphate molecules respectively and 'org' is represented as a species in organic phase.
The distribution coefficient, D of phenylacetic acid with solvents (tri-n-butyl phosphate + diluents) can be expressed as: where, HA T org ⋅ [ ] is the complex (phenylacetic ac- id-tri-n-butyl phosphate) concentration in the organic phase and is the total phenylacetic acid concentration in the aqueous phase, i.e. the summation of associated and dissociated form of phenylacetic acid in the aqueous phase.
Dissociation of carboxylic acid in aqueous phase at equilibrium is expressed as: The dissociation constant, K a can be given by (Dippy et al., 1959) By simplifying Eqs. ( 4) and ( 6), the concentration of undissociated and unextracted phenylacetic acid molecules in aqueous phase can be written as: By rearranging the Eqs.( 3), ( 5) and ( 7), the following expression for the overall distribution coefficient is obtained as: Eq. ( 8) can be expressed in the Logarithmic form which represents a straight line The initial concentration of tri-n-butyl phosphate, 0.366 mol L -1 was higher than initial concentration of phenylacetic acid; therefore [T] org was assumed to be sufficient concentration of tri-n-butyl phosphate in the solvent phase. 42The slope of the straight line of Eq. ( 9) was used to determine the numbers of tri-n-butyl phosphate molecule participating in the interfacial reaction with the acid molecule and the extraction equilibrium constant, E is evaluated by intercept of the line.
Influence of the various concentrations of tri-nbutyl phosphate on the extraction efficiency in methyl isobutyl ketone and petroleum ether was studied to investigate the stoichiometric coefficient of tri-n-butyl phosphate molecule reacted with phenylacetic acid, and also the influence of diluent polarity on the structure of the complex at interface.The slopes of the straight lines with coefficient of determination, R 2 > 0.99 (Fig. 3) and other calculated parameters are mentioned in Tables 2-3.The results indicate the modification of the complex between extractant and acid at interface as a function of the polarity of the organic phase.In less polar solvents (petroleum ether) tri-n-butyl phosphate has low solubility, dispersibility, and low interaction with phenylacetic acid, hence only 1:1 acid-extractant complex can be formed.In polar solvent, (MIBK) tri-n-butyl phosphate has high solubility, dispersibility and interactions with phenylacetic acid, so, 2:1 complex between phenylacetic acidtri-n-butyl phosphate can be formed.
. 46.665 L 2 mol -2 0.999 0.0225 a Tri-n-butyl phosphate molecules participating in the complex with phenylacetic acid molecule Thus, it can be observed that, the extraction efficiency in methyl isobutyl ketone is increased significantly with increasing tri-n-butyl phosphate concentration up to 2.199 mol L -1 which corresponds to 1:0.5 molar ratio of phenylacetic acid and tri-n-butyl phosphate while, in petroleum ether it increased considerably with molar ratio of 1:1.2 of phenylacetic acid and tri-n-butyl phosphate.Using the mass action law, the value of γ can be explained by stoichiometric analysis of complex formation of phenylacetic acid and tri-n-butyl phosphate was 1:1 and 2:1.The first acid molecule interacts directly with the phosphate molecule to form an ion pair, and the OH of the carboxyl group of the second acid molecule forms a hydrogen bond with the conjugate CO of the carboxylate of the first acid molecule to form 2:1 complex, and it is not dependent on the ionisation strength of acid (pK a ). 43

Influence of extractant concentration on equilibrium pH and extraction efficiency
Equilibrium pH in the extraction of phenylacetic acid depends on concentration of tri-nbutyl phosphate, and was found to increase up to 3.40-3.44due to transfer of acid molecule into organic phase.From Figs. 4(a)-(b), it was observed that the increase of tri-n-butyl phosphate concentration favours extraction efficiency, which subsequently leads to increase in equilibrium pH.At higher concentration of tri-n-butyl phosphate 2.199 mol L -1 , equilibrium pH was found as 3.44 and 3.40 in methyl isobutyl ketone and petroleum ether, respectively.In strong acidic domain (pH < 4), con-  centration of H + ions is very low because the carboxyl group is not dissociated, which allows transfer of undissociated acid molecules into the organic phase.
The above results, as shown in Table 3, confirm that, at variable concentration of tri-n-butyl phosphate, methyl isobutyl ketone gives higher extraction efficiency to the extent of 98.79 % age), whereas petroleum ether provides 97.29 % (average).This can be attributed to the strong hydrogen bonding between acid molecules and methyl isobutyl ketone, whereas, petroleum ether, being a mixture of long chain aliphatic hydrocarbons (non-polar diluent) and less interaction in water, gives lower extraction efficiency.
Effect of extractant concentration on distribution coefficient and loading ratio Fig. 5(a)-(b) reveals that the average distribution coefficients for phenylacetic acid with tri-n-butyl phosphate in methyl isobutyl ketone was (D avg = 90.5)higher than in petroleum ether (D avg = 60.40).
The loading ratio, i.e. total phenylacetic acid molecules loaded on the extractant molecules can be expressed as: of phenylacetic acid in all the forms (free acid molecule, dimer or complex) extracted in organic phase and initial concentration of tri-n-butyl phosphate in organic phase, respectively.Carboxylic acids are dissociated in water and extracted into extracted phase through their physical solubility and reactivity with extractant molecule and form 1:1, 2:1, 3:1 and 1:2 acid-extractant complexes, where 2:1 and 3:1 complexes result from the formation of dimer and trimer respectively. 44Formation of hydrogen bond between acid and extractant molecule and the hydroxyl and aldehyde group of carboxylic acid favours 1:2 complexes.Most of the complexes exist in 1:1 form when the carboxylic acid concentration is low (φ < 1). 45verloading (φ >1) occurs at relatively high equilibrium aqueous solute concentration.This phenomenon may be attributed to three reasons: co-extraction of water into the organic phase, acid dimerization in the organic phase, and the physical extraction of acid by the complex in organic phase.All three reasons are hydrogen bonding mechanism that depends only on the hydrophobicity of solute and not the extractant.The loading factor gradually decreases with increasing in tri-n-butyl phosphate concentration in both diluents, as shown in Figs.5(a)-(b).In the active diluents (e.g.1-octanol, methyl isobutyl ketone, petroleum ether, and chloroform) with high extractant concentration, the loading ratio decreases with increase in extractant concentrations at φ < 1 as the solvation occurs in the less favourable solvating medium.

Relative basicity model
Relative basicity model in terms of apparent equilibrium constant and relative basicity of extractant has been reported. 46The three major factors have been reported which interrupt the extraction equilibrium behaviour of carboxylic acids such as: acid hydrophobicity (logP), the dissociation equilibrium constant of the acid (pK a ), and the relative basicity of the extractant mixture to HCL (pK b ) except the nature of the solute.If the basicity of the extractant mixture is relative to the solute, this relative basicity of the extractant can represent the nature of the solute, diluent, and extractant, as well as special association like solvating power.A model equation is expressed as: 47     (11) were, δ 1 and δ 2 are the constants.
Extraction equilibrium behaviour of carboxylic acids with the extractant-diluent system was predicted by model (11) and the constant values are shown in Table 4. Parity plot (see Fig. 7) represents the conformity of model results with the experimental results.The average values of E 1:γ obtained for 10 -60 % by volume of tri-n-butyl phosphate are shown in Table 4.To maintain the suitability of fit, the root mean square error deviation (RMSD) value was calculated using the difference between experimental (E 1:γ, Exp. ) and the predication of the relative basicity model (E 1:γ, RBM ) by using the following equation: where N is the number of experimental data value.The Root mean square error deviation of the relative basicity model was determined to be ±4.4 %, which shows that all the predicated apparent equilibrium constants agree well with the experimental values, which is acceptable, considering the experimental uncertainty.

Spectroscopy analysis
The qualitative infrared spectra of organic phases (methyl isobutyl ketone, petroleum ether, and tri-n-butyl phosphate) were recorded on FTIR (Shimadzu -IR Affinity model Japan), in the range of 4000 − 400 cm -1 with 4 cm -1 resolution using 0.02 cm NaCl window.The spectra of pure phenylacetic acid were also taken with diffusive reflective spectrum (DRS) accessory.

Description of pure phenylacetic acid spectrum
The band 3500 − 2500 cm -1 represents the hydrogen bonding of carboxylic group (-COOH), 1699 cm -1 assigned to the characteristic stretching vibration of ketone group (C=O), presence of aromatic ring due to stretching vibration of C=C in 1600 − 1498 cm -1 region and about 1240 cm -1 represents the stretching vibration of C-O.

Description of presence of phenylacetic acid in solvents
Figs. 6(a-b) confirm that stretching vibration of ketone group (C=O) in phenylacetic acid at 1699 cm -1 was overlapping with methyl isobutyl ketone peaks in 1714 -1680 cm -1 region and the aromatic ring peak of phenylacetic acid at 1498 cm -1 merged with the band of bending vibration of C-H at 1467 cm -1 .Thus, parent peaks of phenylacetic acid are not clearly visible in methyl isobutyl ketone, whereas distinct peaks about 1240 cm -1 are evidence of the existence of phenylacetic acid with tri-n-butyl phosphate in methyl ketone.On the other hand, strong peaks of 1728 cm -1 and 1256 cm -1 assigned to the stretching vibrations of C=O and C-O, respectively, indicate clearly the presence of phenylacetic acid with tri-n-butyl phosphate in petroleum ether.

Evidence of the presence of tri-n-butyl phosphate in diluents
Characteristic parent peak of tri-n-butyl phosphate i.e. 3545 and 3491 cm -1 appears in the spectra with partial shift as 3479, 3414 cm -1 and 3475, 3444 cm -1 in methyl isobutyl ketone and petroleum ether, respectively.1028 cm -1 appeared, which may be attributed to symmetric stretching vibrations of phosphate (P-O) with both diluents in finger print region.
Ion-pair complexation of phenylacetic acid and tri-n-butyl phosphate (1:1) The 3500 -2500 cm -1 band represents the hydrogen bonding of carboxylic group (COOH) and its disappearance after extraction into the organic phase (TBP+MIBK and TBP+PE), indicates that phenylacetic acid and tri-n-butyl phosphate form the ionpair complex.In other words, the reactive extraction reaction is a proton transfer from acid to extractant.

Conclusion
The reactive extraction of phenylacetic acid with tri-n-butyl phosphate in methyl isobutyl ketone and petroleum ether was investigated.The distribution coefficient, loading ratio, and extraction efficiency were obtained for the existing system.Average extraction efficiency, η avg = 98.79 % and 97.29 % were found for TBP+MIBK and TBP+PE with average distribution coefficient, D avg = 90.05 and 60.40, respectively.The higher values of distribution coefficient are attributed to ion pair formation between phenylacetic acid and tri-n-butyl phosphate.Stoichiometric analysis of reactive extraction mechanism confirms 1:1 and 2:1 complex formation between phenylacetic acid and tri-n-butyl phosphate molecule.Ion pair complex was confirmed by IR spectroscopy.To represent the experimental data, relative basicity model was used and found within ± 5 %.

F i g . 3 -
Determination of the number of moles of TBP per moles of phenylacetic acid in phenylacetic acid-trin-butyl phosphate complex as per equation (9)Ta b l e 2 -Number of extractant molecules and overall equilibrium constantDiluent

F
i g .4 (a) -Influence of tri-n-butyl concentration on equilibrium pH * and extraction efficiency for methyl isobutyl ketone F i g .4 (b) -Influence of tri-n-butyl concentration on equilibrium pH* and extraction efficiency for petroleum ether T o org [ ] , are the concentration

F
i g .5(a) -Effect of tri-n-butyl concentration on distribution coefficient and loading ratio for methyl isobutyl ketone F i g .5 (b) -Effect of tri-n-butyl concentration on distribution coefficient and loading ratio for petroleum ether

N
o m e n c l a t u r e D -Distribution coefficient E -Overall equilibrium complexation constant, L mol -1 HA -Undissociated acid in aqueous phase [HA] -Concentration of phenylacetic acid, mol L -1 pK a -Dissociation constant pK b -Relative basicity of tri-n-butyl phosphate

F i g . 7 -
Parity plot for relative basicity model predicted E 1:γ for reactive extraction of phenylacetic acid with tri-n-butyl phosphate in methyl isobutyl ketone and petroleum ether N -Total number of experimental data T -Free tri-n-butyl phosphate molecules in organic phase [T] org -Tri-n-butyl phosphate concentration in organic phase, mol L -1 A b b r e v i a t i o n s DRS -Diffusive reflectance spectrum MIBK -Methyl isobutyl ketone PE Petroleum ether RBM -Relative basicity model RMSD -Root mean square error deviation RSME -Root square mean error TBP -Tri-n-butyl phosphate G r e e k w o r d s γ -Stoichiometric coefficient of extractant η -Extraction efficiency φ -Loading ratio δ 1 and δ 2 -Constants for relative basicity model S u p e r s c r i p t s a n d S u b s c r i p t s Ta b l e 1 -Physical properties of the diluents and extractant F i g . 2 -HPLC chromatogram of phenylacetic acid standard HA HA A aq aq Ta b l e 5 -Water co-extraction results for phenylacetic acid without TBP (physical extraction) system at 298 K for various concentrations of TBP at fixed concentration of phenylacetic acid