Reactive Extraction Results
The complex formation between acid and extractant molecules is the chief difference between the reactive and solvent extractions. Trioctylphosphine oxide, tributyl phosphate, and Cyanex 23 are the types of organophosphorus extractants that have been broadly used to recover carboxylic acids [5, 28, 33–36]. Organophosphorus extractants have good chemical stability, they are effective in acid recovery and their recycle is possible. Besides, they co-extract water in negligible amounts. The phosphoryl group present in the organophosphorus extractant and the carboxylic group of the carboxylic acid react with each other, and so an acid-extractant complex forms.
Reactive extraction results are presented in terms of KDand E % in Table 3. Fig. 1a-b shows the change in KDand E % values versus the change in Ph3PO concentration. The results in the presence of reactive extractant were better compared to the conventional solvents alone. The physical extraction efficiencies were 23, 21, 44, 54, and 48 % with oleyl alcohol, dimethyl adipate, isobutanol, methyl isopropyl ketone, and methyl ethyl ketone, respectively. Methyl isopropyl ketone and methyl ethyl ketone showed the best performance in terms of extraction efficiency, followed by isobutanol, dimethyl adipate, and oleyl alcohol. As seen, inadequate extraction efficiencies were obtained in the case of physical extraction. For the reactive extraction experiments, Ph3PO was varied from 12 % to 44 % by volume in the solvents to see its effect on the efficiency. Remarkably, in the presence of Ph3PO, the efficiencies increased up to 61, 76, 86, 67, and 84 %, respectively. All the reactive extraction efficiencies, even at the lowest TPPO amount, are quite higher than those obtained in the physical extractions. These great differences emphasize the importance of the reactive extractant. As seen from Fig. 1 and Table 3 that KD values followed identical sequences with E % values. It is clear from the reactive extraction results that the acid-extractant complex was poorly solved by oleyl alcohol and dimethyl adipate, and thus the use of them with Ph3PO would not be suggested. Whereas, acid-extractant complex was efficiently solved by the ketones and isobutanol, and therefore the use of isobutanol, methyl isopropyl ketone, and methyl ethyl ketone with Ph3PO would be suggested. Acid recovery is highly dependent on the extractant concentration. However, the extractants are expensive. For the sake of the economy, the extractant amount should be optimized. Irrespective of cost, a further increase in the Ph3PO amount can cause a third phase formation if solvent amount would not be enough to solve the complex.
The solvent properties like dielectric constant, boiling point, density, molecular weight, Dimroth-Reichardt ET parameter, and dipole moment (µ) have been attempted to correlate with KD [37]. ET parameter gives information about the ionization power of the solvent. However, there was no significant correlation found between KD and them.
Z =\(\frac{{[HGA]}_{\text{org}}}{[Ph_{3}\text{PO}]}\)(4)
The loading factor (Z) is the ratio of organic phase acid concentration to the organic phase extractant concentration. Basically, it is the fraction that shows how many acid molecules are loaded on an extractant molecule. The loading factors were calculated and written in Table 3. As seen, the higher extractant concentrations were used in the solvents, the less were the loading factor values, which can be expressed by the definition of Z [37]. The loading factor values were calculated to be less than 0.5, showing that there was no overload on the extractant molecule. In other words, one extractant molecule reacted with one acid molecule, showing 1:1 acid-extractant complex formation. The reaction between carboxylic group and phosphoryl group is shown in Fig. 2. The complexation constant KE gives information about the type of the formed complex and estimated from the loading factors. If loading factors are less than 0.5, complexation constant values are calculated as in the below Eq. (5) [37, 38]:
\(\frac{Z}{1-Z}\) = KE [HGA]aq (5)
where KE is the experimental complexation constant of 1:1 acid-extractant complex formation, and it is found from the slope when \(\frac{Z}{1-Z}\) is plotted versus [HGA]aq[39].
Mass Action Law Model explains the nature and type of the formed complex. First, acid and extractant molecules interact with each other at the interphase, then complexation occurs, and afterwards the formed complex is solubilized in the solvent. Dipole-dipole interactions and hydrogen bonding enable dissolution of the complex in the solvent. Although according to the Mass Action Law model, activities of species are proportional to the concentrations of species, a non-ideal behavior can be observed for the equilibrium constant. The following equation gives the relationship of distribution coefficient with Mass Action Law model’s equilibrium constant [27, 37]:
Log KD = Log KE,MAL + s Log [Ph3PO]org (6)
where KE,MAL is the equilibrium constant of the Mass Action Law model, s is the solvation number. [Ph3PO]org is the organic phase Ph3PO concentration. KE,MAL might be obtained from the intercept, when Log KD is plotted against Log [Ph3PO]org. Higher KE,MAL value means higher complexation. In this circumstance, solvent would have strong ability to solve the complex and the extractant would display high complexation potential with the acid [27]. KE,MAL values were found to be dissimilar to KE values.