ISSN 1070-3632, Russian Journal of General Chemistry, 2009, Vol. 79, No. 9, pp. 1890–1894. © Pleiades Publishing, Ltd., 2009. Original Russian Text © Yu.E. Zevatskii, D.O. Ruzanov, D.V. Samoilov, 2009, published in Zhurnal Obshchei Khimii, 2009, Vol. 79, No. 9, pp. 1533–1537.
Experimental Testing of the Calculation Results of the Dissociation Constants of Organic Compounds in Nonaqueous Media Yu. E. Zevatskii, D. O. Ruzanov, and D. V. Samoilov St. Petersburg State Institute of Technology, Moskovskii pr. 26, St. Petersburg, 190013 Russia e-mail:
[email protected] Received April 30, 2009
Abstract―27 Dissociation constants of carboxylic and NH-acids in methanol, ethanol and formamide were measured by potentiometric method. The selection of compounds for the analysis is determined by the preliminary computation of the constants by elemental linear empirical method, and also by the absence of the consistent published data on the values for the chosen compounds. The comparison of results showed that the obtained values of dissociation constants as a rule are consistent with those predicted in the range with the relative error of 10%.
DOI: 10.1134/S107036320909014X The new elemental linear empirical method (ELEM) we previously proposed was successfully used for calculating dissociation constants (pKa) in nonaqueous media [1,2] on the basis of the array of experimental data and experimental parameters of solvents. With the use of empirical coefficients 363 values of the dissociation constants were calculated of 33 derivatives of benzoic and acetic acids in 11 solvents [1] and 192 values of pKa for 24 NHacids in 8 solvents [2]. This set included 134 values of the pKa, whose experimental values were absent in the literature [3-15] that we chose as the source of consistent and reliable data for these substances. For evaluating the accuracy of prediction of the pKa values by this method in the nonaqueous media we decided to experimentally determine the pKa values of some compounds mentioned above, for which earlier [1, 2] the calculated values were obtained in three organic solvents. From this group of substances as the objects for studies those were selected, whose experimental data were absent in the literature or were not included into examination at the development of the method [1, 2]. Furthermore, the selection of the objects to a considerable extent depends on their accessibility. The solvents for the study were selected on the basis of the considerations (see below) of possible
increase in reproducibility of measurement procedure: lowest alcohols as the typical representatives of polar proton solvents and formamide as a polar aprotic solvent with the dielectric constant larger than that of water (εs = 111). Inasmuch as the majority of values in the array of experimental data we used in [1, 2] were obtained for carboxylic acids by potentiometric titration with a solution of triethylamine in an appropriate solvent [3,4], we used the same procedure for ensuring the validity of the comparison of the experimental values obtained by us with those calculated earlier in [1]. Although to the data array for the NH-acids used in [2] were included the values of the pKa obtained by different methods, for providing the uniformity of the measurement results for all examined substances we also applied the potentiometric titration (with a solution of dry hydrogen chloride in lowest alcohols). In the titration the first 3–5 points of titration curve were used that corresponded to 15–40% of equivalent titration. The dose of titrant for different objects was within the limits of 1–2 ml. The concentrations of titrates were about 0.01 mol l–1, and of titrants, about 0.05 mol l–1. After adding the next portion of titrant and 5-min stirring, the measuring system was left for 1 h to reach equilibrium, and afterwards the emf (electromotive force) of the metering circuit was
1890
EXPERIMENTAL TESTING OF THE CALCULATION RESULTS
recorded. The work up of the curve was performed by standard procedure, taking into account the correction for ion activity according to the second DebyeHueckel approximation [16]. The ionic force during the titration did not exceed 0.01 mol l–1. Although for determining pKa in alcohols we used titration for all investigated objects, at the measurements in the formamide we used potentiometry by the method of filling in a special cell. This is due to the fact that in aprotic solvents the titration curve is distorted as a result of the phenomena of ion-molecular association and possible incompleteness of the dissociation of buffer salts [17]. Therefore for calculation of the pKa value from the titration curve in this case should be used only to a point of halfneutralization. Since in the nonaqueous media the equilibrium potential at the use of a glass electrode is established extremely slowly (several hours for each point) [17], the process of registering the complete titration curve for finding the point of halfneutralization becomes unreasonably long. Owing to these two reasons we suggested that the application of a method of filling is more expedient. For the measurements in formamide the same pair of electrodes as in alcohols was used, but the reference electrode was filled with the standard saturated solution of KCl in water. The cell of filling consisted of two U-tubes with the electrodes connected by the electrolytic bridge filled with the solution of tetrabutylammonium perchlorate in formamide with the concentration 0.1 mol l–1. The electrolyte in the bridge was renewed after 3 fillings. In the measuring part the investigated solution was placed, while the comparative part contained the saturated solution of KCl in water. Equilibrium emf of this chain was recorded on the average after 2 h after filling using a ionometer I-500 fabricated by “Aquilon” scientific and production company. For determining the pKa in formamide solutions were prepared of the mixtures of the studied carboxylic acids with the respective tetraethylammonum salts or amines with the corresponding picrates of composition 1:1 with the concentration 0.01 mol l–1 containing additionally 0.1 mol l–1 tetrabytylammonium chloride as supporting electrolyte, and the emf values of the filling cell with the indicated solutions in the measuring part were measured. The values of paH* were found according to the calibration curve, and by Eq. (1) the value of pKa was calculated: рKа = раН*(1/2) – log f1,
(1)
1891
where paH*(1/2) is the value of paH*, obtained for the buffer mixture of composition 1:1, f1 is the activity coefficient of single-charged ion computed with the use of Debye–Hueckel equation. We failed to isolate tetraethylammonium salts of acetic, isobutyric and propionic acids in solid form and therefore their pKa values were obtained along Eq. (1) from the results of titration of solutions of these acids with a solution of triethylamine in formamide with the addition of 0.01 M of tetrabutylammonium chloride. The results of the experimental determination of pKa in this work (pKaexp), computed values from [1,2] for the studied objects (pKacalc), and also the published data [18–26] (pKapubl) are listed in the table. The published values were not previously included into the array processed in [1, 2] for three reasons: (1) The data were obtained using another method; (2) The large compilation of experimental data (values in solvents) appeared in the literature [24, 25] after publishing [1]; this array drew our attention, first of all, because in these works was achieved the best in the last decade accuracy of the calculation of pKa in the nonaqueous media with the use of a QSPR method; (3) The data were consciously excluded from the array for checking the accuracy of the prediction of the pKa by empirical method without performing an additional experimental work. The gaps in the table mean that in the works [1, 2] were already given both experimental and calculated values of the corresponding pKa and any new given (obtained by us or published by other authors) in this work were not added. The relative error (ε) is calculated for the experimental values obtained in this work. As can be seen from the table, the majority of the obtained experimental values fit the previously found [1, 2] permissible range of relative error 10%, and therewith in the most cases also well agree with the added in this work published data [18–26]. A significant disagreement of the calculated pKa value of propionic acid in ethanol with that we obtained experimentally (the latter is in a good agreement with the published value) is caused by strong deviation of the empirical parameters obtained by us in [1] for this acid from a series of the data for the other aliphatic acids. Nevertheless, with these parameters the computed values fit well the experimental data for this acid used in [1].
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 79 No. 9 2009
1892
ZEVATSKII et al.
Comparison of experimental, calculated and published pKa values of the objects studied, at 25°Ca Methanol Carboxylic acid
рKаcalc
рKаexp
2-NO2C6H4-COOH
рKаpubl
Ethanol рKаcalc
рKаexp
рKаpubl
ε, %
рKаcalc
рKаexp
рKаpubl
ε, %
7.64
8.4
8.25
8.27
2.3
4.08
4.46
4.28
8.5
10.5
9.50
9.70
10.0
9.30
10.8
10.25
10.10
5.7
11.4
10.61
10.60
7.8
9.3
9.25
9.08
0.8
5.69
5.32
6.9
9.8
10.10
10.10
3.2
5.68
5.50
3.2
9.13
ε, %
4-IC6H4-COOH 3-MeOC6H4-COOH 4-MeOC6H4-COOH 3-CNC6H4COOH
Formamide
8.46
2-MeCOC6H4COOH 4-MeSO2C6H4COOH 2-HCOPhCOOH 2-NOPhCOOH CH3-COOH
9.4
9.45
10.3
10.00
8.8
9.00
10.8
ClCH2-COOH
0.8
5.34
5.57
4.2
2.7
6.09
6.07
0.4
1.7 5.80
8.5
5.86
6.45
9.1
7.49
7.45
6.9
5.01
3.82
4.74
31.2
28.5
6.12
7.10
7.18
13.8
10.79
13.1
6.84
7.02
12.4
11.24
10.0
7.00
7.42
9.1
9.17
0.5
5.51
5.05
9.1
2.6
5.91
5.55
6.4
5.2
6.81
6.33
7.6
4.4
10.50
10.2
2.7
3.00
3.4
11.7
C2H5-COOH
11.7
10.03
9.71
16.8
13.6
10.55
(CH3)2CH-COOH
11.0
10.67
9.90
2.7
12.2
(CH3)3C-COOH
11.1
10.67
7.43
4.2
PhOCH2-COOH
8.6
8.35
1.2
Ph2CH-COOH
9.6
9.62
0.2
10.4
10.17
PhCH2-COOH
9.9
10.01
9.51
1.0
10.7
10.19
12.40
11.86
11.48
4.6
12.2
10.60
10.20
2.5 5.01
5.6
Amines BuNH3+ 1-NapthNH3+ Et2NH2+
11.90
11.92
Bu3NH+
10.60
10.13
11.20
0.2
10.7
10.97
8.2
11.50
10.4
10.4
4.6
9.3
10.23
9.1
10.70
8.9
20.8
4-Me-C6H4NH3+
4.30
4.8
9.6
3-Me-C6H4NH3+
4.00
4.3
7.6
4-Br-C6H4NH3+
2.70
3.5
22.4
3-NO2-C6H4NH3+
3.40
3.78
3.46
10.1
3.1
3.32
6.6
0.95
1.7
42.9
4-NO2-C6H4NH3+
2.30
2.71
1.55
15.1
2.2
2.84
22.5
1.20
2.0
39.1
PhNH2Me
4.10
3.8
7.9
PhNH2Et+
4.60
4.2
10.3
PhNHMe2+
4.90
3.8
27.8
PhNHEt2+
6.50
5.0
30.1
2-Me-C5H5N+
6.10
5.2
17.9
5.10
4.0
+
Quinolinium a
11.69
5.43
pKexp a
pKacalc
are experimental pKa values determined in this work, calculated values [1,2] for the studied objects, data [18–26], ε is relative deviation of pKaexp from pKacalc, expressed in percents.
The high values of a relative error in the calculation for 4-nitroaniline at the small absolute values of pKa and the absence of published data leads to a conclusion that for increasing the accuracy of prediction it is desirable to include the minimum values of the pKa into the initial array at the determination of the empirical parameters.
26.9 pKpubl a
are published
Thus, the comparison of the values of the pKa, calculated by the elemental linear empirical method [1, 2], with the experimental values we obtained in this work and that were published in the literature, confirmed a good accuracy of the prediction of pKa values in nonaqueous media by this method.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 79 No. 9 2009
EXPERIMENTAL TESTING OF THE CALCULATION RESULTS
The good agreement in most cases of the pKa values obtained by us with newly published data proved the correct selection of the measurement procedure. A significant deviation of the calculated pKa values from the experimental for the amines in formamide, in contrast to carboxylic acids, is due to fewer available experimental values used in [2] for the calculation of the empirical parameters for the amines (3–4 values for each) compared to carboxylic acids (13–14 values). EXPERIMENTAL All measurements were carried out in the air thermostat TSO-80/U at 25°C. The compounds analyzed in the work were purchased from “Acros Organics” containing a mass fraction of the basic component 98–99%, or the reagents of domestic manufacture of “analytically pure” and “chemically pure” grade were used. Their quality was verified by measuring the melting point for solids and the refractive index and the density for liquids. When necessary, the substances were purified by recrystallization or distillation. The lithium salts of buffer acids used for calibrating the metering circuit for measuring in alcohols were obtained by the neutralization of the aqueous solutions of the corresponding acids with the solution of lithium hydroxide (of the concentration ~1 mol l–1) followed by evaporation to dryness in a vacuum, recrystallization from ethanol, and drying in a vacuum at 110°C to the constant weight. The NH-acids picrates for the measurements in the formamide were obtained employing the following general procedure: a weighed sample of picric acid was dissolved in a minimum quantity of ethanol or benzene at boiling, the solution was cooled to ~60°C, and then a weighed quantity of amine corresponding to 95–98% of the theoretical amount was added to it. The picrate precipitate formed was filtered off on a vacuum filter, washed with benzene, dried in air or in a vacuum at 70–80°C. The samples were examined by determination of their melting points. If necessary, they were recrystallized from a mixture ethanol–benzene (4:1, 1:40), or from pure benzene or 2-propanol. The carboxylic acids tetraethylammonum salts except for above indicated and those of phenols for the buffer mixtures were obtained employing the Kolthoff procedures [27–29]. The reagents used for the preparation of buffer
1893
mixtures, were also preliminarily dried in a vacuum to a constant weight. The preparation of anhydrous ethanol was performed by a standard procedure [30]. Formamide from “Fluka” of “puriss” quality was purified by fractional crystallization to a constant melting point 2.55°C (measured with two Beckmann thermometers) and was used within two weeks after purification. Methanol of “chemically pure” grade was used without additional purification and drying. The solvents quality was checked by the absorption spectra (the transmission board), by the density and by the refraction index. The water content in the solvents was checked by coulometric titration according to Fischer procedure in correspondence with the GOST 24614-82 using a titrator Expert-007 produced by “Ekoniks-expert”. The water content in ethanol was 0.005–0.015 wt %, in methanol 0.02 wt % ,and in formamide 0.015–0.02 wt %. The drying of hydrogen chloride for the preparation of solutions in alcohol was carried out by its passing through two Drechsel flasks packed with balls and filled with sulfuric acid. The titration was carried out on an automatic titrator ATP-02 produced by “Aquilon.” The metering circuit consisted of glass electrode ES-10603/7 and silver chloride reference electrode ESr-10103 produced by Scientific-Producing Company “Izmeritel’naya Tekhnika” (measuring instrumens). Instead of filling with standard solution, the reference electrode was filled with the saturated solution of anhydrous LiCl in an appropriate alcohol. Measurements were performed in the specially designed glass measuring cell with a magnetic stirrer, protected from the air moisture by calcium chloride drying tube. The electrodes and hydraulic circuit of the titrator were connected with the cell by an ait-tight ground-glass joint. For calibrating the system at the measurements in alcohols the buffer mixtures were used of 1:1 composition of picric, or benzoic, or salycilic, or 5,5diethylbarbituric acid with the respective lithium salt of the concentration 0.01 mol l–1. The calibration for the measurements in the formamide was performed on analogous buffer mixtures on the basis of triethylamine, 2,4-dichloroaniline, 2,4-dinitrophenol, and benzoic acid. The pKa values of buffer acids for constructing the calibration curves were taken from [14, 15, 17] for methanol, ethanol, and formamide, respectively. Below are given the parameters of calibration curves for the studied solvents.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 79 No. 9 2009
1894
ZEVATSKII et al. Solvent
Methanol
Ethanol
Formamide
k, mV/pH unit
57.98
58.18
55.87
E0, mV
575.80
586.0
308.30
ACKNOWLEDGMENTS The authors express their gratitude to N.O. Mchedlov-Petrosyan (Karazin Khar’kov National University of Ukraine) for the valuable recommendations regarding the measurement procedure and to A.V. Selitrennikov for purification of technical ethanol and its drying. REFERENCES 1. Zevatskii, Yu.E. and Samoilov, D.V., Zh. Org. Khim., 2008, vol. 44, no. 1, p. 59. 2. Zevatskii, Yu.E. and Samoilov, D.V., Zh. Org. Khim., 2008, vol. 44, no. 12, p. 1764. 3. Bartnicka, H., Bojanowska, L., and Kalinowski, M.K., Austr. J. Chem., 1991, vol. 44, no. 8, p. 1077. 4. Bartnicka, H., Bojanowska, L., and Kalinowski, M.K., Austr. J. Chem., 1993, vol. 46, no. 1, p. 31. 5. Pytela, O. and Kulhanek, J., Coll. Czech Chem. Comm., 2002, vol. 67, no. 5, p. 596. 6. Kolthoff, I.M., Chantooni, M.K., and Bhowmik, S., J. Am. Chem. Soc., 1968, vol. 90, p. 23. 7. Korolev, B.A., Zh. Obshch. Khim., 1980, vol. 50, p. 841. 8. Korolev, B.A., Zh. Obshch. Khim., 1979, vol. 49, p. 2360. 9. Izmailov, N.A. and Mozharova, T.V., Zh. Fiz. Khim., 1960, vol. 34, p. 1709. 10. Bacarella, A.L., Grunwald, E., Marshall H.P., and Purlee, E.L., J. Org. Chem., 1955, vol. 20, p. 747. 11. Izmailov, N.A. and Mozharova, T.V., Zh. Fiz. Khim., 1960, vol. 34, p. 1543. 12. Gutbezahl, B., J. Am. Chem. Soc., 1953, vol. 75, p. 559.
13. Kaljurand, I., http://tera.chem.ut.ee/~manna/pkadata. html. 14. Courtot-Coupez, J. and Le Demezet, M., Bull. Soc. Chim. France, 1969, p. 1033. 15. Lange’s Handbook of Chemistry, John A. Dean, Ed., McGraw-Hill Inc., 15th edition, 1999. 16. Al’bert, A. and Serzhent, E., Konstanty ionizatsii kislot i osnovanii (Ionization Constants of Acids and Bases), Moscow: Khimiya, 1964. 17. Izutsu, K., Acid-Base Dissociation Constants in Dipolar Aprotic Solvents. IUPAC, Blackwell Scientific Publications, Oxford, 1990. 18. Mchedlov-Petrosyan, N.O. and Vasetskaya, L.V., Zh. Obshch. Khim., 1989, vol. 59, no. 3, p. 691. 19. Mchedlov-Petrosyan, N.O., Zh. Obshch. Khim., 2003, vol. 73, no. 2, p. 288. 20. Verhoek, F.H., J. Am. Chem. Soc., 1936, vol. 58, no. 12, p. 2577. 21. Ludwig, M., Baron, V., Kalfus, K., Pytela, O., and Večeřa, M., Coll. Czech. Chem. Comm., 1986, vol. 51, no. 10, p. 2135. 22. Mollin, J., Pavelek, Z., Navratilová, J., and Recmanová, A., Coll. Czech. Chem. Commun., 1985, vol. 50, p. 2670. 23. Exner, O. and Kalfus, K., Collect. Czech. Chem. Commun., 1976, vol. 41, p. 569. 24. Jover, J., Bosque, R., and Sales, J., QSAR Comb. Sci., 2008, vol. 27, no. 4, p. 563. 25. Jover, J., Bosque, R., and Sales, J., QSAR Comb. Sci., 2008, vol. 27, no. 10, p. 1179. 26. Rived, F., Roses, R., and Bosch, E., Anal. Chim. Acta., 1998, vol. 374, p. 309. 27. Kolthoff, I.M. and Chantooni, M.K., J. Phys. Chem., 1966, vol. 70, no. 3, p. 856. 28. Kolthoff, I.M. and Chantooni, M.K., J. Am. Chem. Soc., 1970, vol. 92, no. 24, p. 7025. 29. Kolthoff, I.M., Chantooni, M.K., and Bhowmik, S., J. Am. Chem. Soc., 1966, vol. 88, no. 23, p. 5430. 30. Khramkina, M.N., Praktikum po organicheskomu sintezu (Laboratory Course of Organic Synthesis), 5th edition, Leningrad: Khimiya, 1988.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 79 No. 9 2009