Spectrophotometric Determination of Proto- Nation Constant of N
Indian Journal of Chemistry Vol. 42A. May 2003. pp. IORI-108S
Spectrophotometric determination of proto nation constant, or acidity constant refers to when a nation constant of N-phenylbenzohydroxamic compound is potentially amphoteric. As the name im plies, on the one hand hydroxamic acids behave like a acid in mineral acids t weak acids (PKa = 8-9) and on the other, like most carbonyl compounds, they are also weak bases Kallal K Ghash* & Pankaj Tamrakar (pKBH' "" -I to -2). The suppression of acid charac- School of Studies in Chemistry. Pt. Ravishankar Shukla University. Raipur, 492 010. India ter may be attributed to the intramolecular hydrogen Received 13 February 2002, revised 6 January 2003 bonding. Thus, we may refer to pK BH' of B or pKa of BH+ (conjugate species). The experimental studies The UV spectra of N-phenylbenzohydroxamic acid generally rely on indirect methods, like changes in (C6H5CON(OH)C6H5) in sulphuric, perchloric and hydrochloric acid solutions have been analyzed by the characteristic vector NMR chemical shifts, UV or IR spectral changes and method in order to separate protonation from the medium effect. solvent extraction, etc. Protonation studies of car Using the first vector coefficient values. proton at ion constant bonyl compounds are handicapped by uncertainties, ( pK BH' ) and solvation parameters have been estimated. In the which arise from the solvent sensitivity (medium ef high acidity range, results obtained according to the Cox-Yates fect) of the UV spectra. Various methods of compen excess acidity function. Marziano Mc function. Bunnett-Olsen sating for the medium effect have been advanced, but and Hammett-acidity function methods are in good agreement. the most powerful method is the characteristic vector l3 15 The protonation equilibria of polyfunctional acids and analysis - (CV A) or principal component analysis. bases in concentrated solutions of mineral acids, have The encouraging results we obtained using vector analysis procedure for the correction of medium ef been the subject of a number of studies 1-3. On the other hand, only scanty data are available for hydrox fects in UV spectral data of some hydroxamic acids prompted us to adopt this method for estimating the amic acids both in solution4"'{' and in the gas phase7.8. N-Phenylbenzohydroxamic acid (PBHA) has been pKBH+ of N-phenylbenzohydroxamic acid widely employed in analytical chemistry9.1O. One of C6H5CON(OH)C6H5 (PBHA) in sulphuric, perchloric its characteristic properties is that it forms complex and hydrochloric acids. The Hammett acidity function compounds in strongly mineral acid media with sev method (HAFM), the Cox-Yates excess acidity eral highly charged, easily hydrolysed elements, an method (EAM), Marziano Mc method and the Bun effect that sharply increases the specificity and selec nett-Olsen (BOM) method have been compared in tivity of analytical reactions. For the last few years we order to rationalize the differences observed between have been investigating the hydrolysis of a large pKBH+ values estimated by each classical methods I. number of C- and N-substituted hydroxamic acids 11,12. Before any interpretation of these data can be made, it Experimental is necessary to correct the constant of hydrolysis for PBHA was prepared by standard procedure 16. Sul the degree of protonation and solvation parameters of phuric, perchloric and hydrochloric acid solutions the substrate. Hence precise measurement of protona were prf;pared by diluting commercial (Qualigens Ex tion constant (pK BH' ) of the substrate (Eq. 1) is es- celaR) concentrated acids and titrated with standard sential. NaOH. The stock solution of PBHA was prepared in etha nol (1.5 x 10-3 mol dm-3). An aliquat of this solution (20 ml) was diluted to 50 ml using distilled water. The samples were prepared by diluting a small portion (1- There is some confusion over what exactly proto- 2 ml) of the above solution of substrate in water with the appropriate mineral acid solutions. Practically no tpresented in part at the International Conference in Chemistry and 36 Annual Convention of Chemist, Calcutta. 1999, Abs. No. : corrections were necessary for the chemical reactions PHY (A)-I, P B-1 of PBHA in acids during the measurement. The sam- 1082 INDIAN J CHEM. SEC. A. MA Y 2003 pIes were cooled in an ice bath before and during the (S042-, Cn would be more effective catalysts, i.e., addition to prevent the substrate from any hydrolysis protonating power is greater. It seems probable that an effect. The resulting cold solution was then thermo acid having anion of low charge density (C104 -) sta stated to 25±O.1 DC, and the volumetric flask was bilizes carbonium ion like transition state. It can be eventually filled up to the mark with water. The solu seen from the results that H2S04 is the suitable me tions were transferred into I cm quartz cells and the dium for studying protonation behaviour. The batho spectra recorded against an acid solution of the same chromic shifts indicate that more than one process is concentration using Unicam UV-2 300 spectropho occurring; it is apparent that each of the two spectral tometer. The final concentration of PBHA in the me processes contributes approximately equally to the dium was 6.0 x 10-5 mol dm-'. total spectral variation. No clear isosbestic point has been observed. At low sulphuric acid concentration Characteristic vector analysis of UV data (water-5.0 mol dm-3), no significant differences in the UV data were collected into a matrix, each row cor spectra have been observed. At any given acid con responding to a complete spectrum at a given acidity centration, ionization ratios (I = [BH+/[B]) were cal and each column to a particular wavelength. The pa culated from Eq. 2, rameters on which pKRH' was assumed to be de- pendent are protonation and medium (solvation/H ... (2) bonding). This data matrix was then subjected to CV A, according to the procedure originally described where All is the absorbance of free base (molecular by Simonds l7 . The output was a matrix resulting from species), A is the absorbance value at intermediate the coefficients of the first and second characteristic acidity and ARH' is the absorbance of the protonated vectors only, a column of which was processed as base. This study allows us to select the most suitable usual. wavelength (280 nm maximum variation for H2S04
and HCI04 and 260 nm for HCl) and All and AIlH' as Results and discussion the limits. The UV spectra of PBHA at different acidities (H2S04 = 0.0-16.2, HCl04 = 0.0-9.0, HCI = 0.0-9.3 In strong acid media the convention to write proto mol dm-') are very similar for all the three mineral nation equilibria (Eq. I) for the determination of acids. However, in the spectrum of PBHA in hydro pKBH , is as follows (Eq. 3) chloric acid solutions, no significant shifts in Amax is observed with respect to acid concentration. . .. (3) In perchloric acid at low acidities (water-5 mol dm -'), a peak is present at about 256 nm, which shifts where CIl+ is the hydrogen ion concentration andfBfHI to 268 nm in strongly acidic solutions. In sulphuric hH + represents molar activity coefficient ratio. Gener acid a clear cut bathochromic shift (250 nm to 276 nm) is observed. Protonation and medium effects ally the pKBH' values are calculated by Hammett contribute equally to this spectral variation. Solutions acidity function (HAFM) (Eq. 4) and Bunnett-Olsen of PBHA in mineral acids are colourless. They un method (BOM) (Eq. 5). dergo no visible changes on standing. In all the three acids the molar extinction coefficients increased suc log 1= m (-Ho) + pKBH , ... (4) cessively with increasing acid concentrations: [H2S04 (Amax 276) = 5841 to 15094, HCI04 (Amax 268) = 5807 to 15859, HCI (Amax 262) = 8487 to 12730] morl cm-I log 1+ Ho = pKRH' + <1>c (Ho + log CH+) ... (5) at their respective Amax.
We have already studied kinetics of hydrolysis of where Ho is the Hammett acidity function. Bunnett ll PBHA in the mineral acids • For the hydrolysis of Olsen equation may also be represented as Eq. 6. PBHA the catalytic order of strong mineral acids is
HCI > H2S04 > HCl04 . It may be concluded that strong acids having anions of high charge density log I-log CH+ = (<1>-1) (Ho + log CH+) + pKBH+ ... (6) NOTES 1083
The pKRH+ values were obtained by plotting log I Each method has its equivalent for kinetic studies. In against Ho (HAFM) and (log I - log CH+) against (Ho all these plots the standard deviations in slope and + log CH+) (BOM). Thus, only one acidity function intercept were in the range 0.01 < s < 0.11. The
(Ho) is needed for the purpose of estimating pK BHO • pKBH , values of PBHA in HCI and HCI04 solution were low. Amongst these acids, PBHA seems to be logh* JH / h*ft = Mc or X ... (8) stronger base in sulfuric acid solutions .
Cox and Yates I excess acidi ty method is denoted Characteristic vector analysis by Eq. 7 CY (principle component analysis) analysis can be used to obtain an abstract solution where all data are l2 ... (7) expessed as linear sums of product terms . Thus any absorbance, A".p may be represented by Eq. 8; in which CI represents the principal component for According to this method pK • values were deter BH wavelength p, v I represents characteristic vector for mined as intercepts by plotting the log I-log cft the nth spectrum, subscripts I and 2 describe first and against X or Mc values. Theoretical and practical as second components (weighing factors) and finally p pects of the above methods have been treated exten and n describe certain wavelengths and acidities, re sively. Marziano et al. 18 demonstrated the differences spectively. A is mean absorbance at a given wave between activity coefficient function and acidity length. function, as well as the parameter related to proton at ing ability of the solvent. ... (8)
The values of pKBH , calculated by the different methods closely resemble each other (Table I). It is In the present case the first vector accounted for 94% difficult to comment on the validity of each method. of the total data variation and the second vector
Table 1-pKBH+ values of PBHA calculated by different methods and after applying characteristic vector analysis Mineral acid HAFM (mol dm") BaM H" HA m pKBII+ m pKBH+ Q> pKllH+
HCI 0.60 -1.37 0.992 0.86 -1.67 0.991 0.46 -1.43 0.968 (0.0-9.28) (0.55) (-1.12) (0.985) (0.91 ) (-1.25) (0.989) (0.38) (-1.20) (0.979) H2S0~ 0.38 -2.12 0.948 0.69 -2.52 0.968 0.31 -2.46 0.934 (0.0- I 6.2) (0.73) (-2.01 ) (0.989) ( 1.02) (-2.34) (0.980) (0.64) (-2.06) (0.978) HCI04 0.71 -2.07 0.970 0.41 -1.89 0.905 0.33 -1.88 0.927 (0.0-9.00) (0.36) (-1.19) (0.982) (0.53) (-1.37) (0.984) (0.21 ) (-1.33) (0.935)
Mineral acid EAM (mol dm") MCP m* pKBH+ nB pKBH+
HCI 0.44 -1.39 0.961 0.65 -1.45 0.969 (0.0-9.28) (0.48) (-I.I5) (0.97 I) (0.75) (-1.22) (0.982) H2SO4 0.31 -2.41 0.938 0.19 -2.43 0.951 (0.0- 16.2) (0.63) (-2.03) (0.969) (0.94) (-2.12) (0.975) HCI04 0.31 -1.99 0.938 0.34 -1.92 0.953 (0.0-9.00) (0.20) (-1.34) (0.940) (0.21) (-1.38) (0.970)
r = Regression coefficient HAFM = Hammett acidity function method; 80M = Bunnett-Olsen Method; EAM = Excess acidity method; MCP = Marziano-Cimino method t Values in parentheses indicate data after applying CV analysis. 1084 INDIAN J CHEM. SEC. A. MAY 2003
accounting for 6%. The first vector (principal compo curve, changing from a constant CB to a constant nent) is associated predominately with protonation CBH+, when plotted in this way. It is evident that CI and the second reflects only the medium effects. i.e., behaves like this initially but does not reach a stable hydrogen bonding and solvation. The spectra recon final value and is roughly linear, with -Ho. C2 has no stituted using vectors and coefficient obtained are in obvious physical chemical counterpart. Reconstituted good agreement with the originals. The coefficients of spectra show good isosbestic points. The mean curve the two vectors are plotted as a function of -Ho, (for for the spectrum of PBHA in H2S04 is shown in Fig. H2S04 only) for illustrative purposes, (Fig 1). Now 2A. There is a contribution at each wavelength from simple protonation would give a sigmoid titration the first characteristic vector, given in Fig. 2 B weighted by the factor CI from Table 2, and another 0.6 contribution at this wavelength from the second char 0.4 acteristic vector, given in Fig. 2e, weighted by the 0.2 c, factor C2 from Table 2. It should be emphasized that the curves of Fig. 2 are obtained by a purely mathe u 0 matical processing of the raw data by the methods of - 0.2 If vector analysis, with no assumptions about the nature V - 0.4 of the effects (other than their orthogonality). From -0.6 these reconstituted curves, the ionization ratios -2 0 2 4 6 8 10 -110 [BH+]/[B] at varying acid concentrations could be obtained. These ionization ratios are independent of Fig. I--Coefficients of the first and second characteristic vectors the wavelength chosen to calculate them, as they plotted against a function of (-Ho) for H2S04 , should be in the absence of medium effects. It must be noted that our data analysed by aU the methods gener ally gave nearly coincident results. The Bunnett-Olsen 0.9 0.8 A and excess acidity methods have been shown to be 0.7 essentially equivalent, giving the same pKBH+ and 1:: 0.6 § 0.5 .0 [m* = (1 -