The kinetics of the reaction of boric acid with salicylic acid ALANQUEEN Parker Chemical Laboratory, University ofManitoba, Winnipeg, Man., Canada R3T2N2

Can. J. Chem. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/25/17 For personal use only.

Received May 10, 1976l ALANQUEEN.Can. J. Chem. 55,3035 (1977). Salicylic acid forms a 1 : 1 complex with boric acid, the reaction involving both the fully protonated ligand, and the salicylate ion. The kinetics of this reaction have been studied by the stopped-flow method. The stability constant for the reaction involving salicylate ion has been calculated from measurements of the absorbances of solutions at equilibrium and is the same as that obtained from the kinetic data. The kinetic results at p H values in the range 3.45-4.63 suggest that, when salicylic acid is the ligand, the complex is formed in two steps. A similar process may also occur with salicylate ions. ALANQUEEN.Can. J. Chem. 55,3035 (1977). L'acide salicylique forme un complexe 1 :I avec I'acide borique; la reaction implique et le ligand completement protone et I'ion salicylate. On a etudie la cinetique de la reaction par la methode des flux stoppes. On a calcule la constante de stabilitt pour la rkaction impliquant I'ion salicylate; ces calculs ont ttk effectues a partir de mesures d'absorption de solutions en Cquilibre et cette constante de stabilite est la mCme que celle obtenue a partir de donnees cinetiques. Les rtsultats cinetiques, a de valeurs de pH allant de 3.45 a 4.63, suggerent que lorsque I'acide salicylique agit comme ligand, le complexe se forme en deux &tapes.Un processus similaire peut aussi se produire avec les ions salicylates. [Traduit par le journal]

Introduction Boric acid forms 1 : 1 and 1 :2 complexes with cis diols (1, 2), a-hydroxy acids (3), dicarboxylic acids (4) and nucleosides (5) as well as other compounds such as a-diketones (6). Most of the previous work has been concerned with equilibrium studies but recently Pizer and his coworkers have published the results of a series of kinetic studies of the reactions of boric acid with tartaric acid and phenylboronic acid with lactic, oxalic, and malonic acids (4, 7-10). The appearance of this work prompts the publication of results that have been obtained using boric and salicylic acids. Pizer has interpreted his data in terms of several factors that influence the rate of complex formation. These are the acidity of the ligand and interactions that stabilize a four coordinate species. The first of these effects seems to be well established but the other effects are less easily identified. I11 particular, a proposal that in some cases there may be attractive interactions between the aromatic ring of phenylboronic acid and the carboxylate group of the ionized ligand is quite vague. Salicylic acid was chosen for the present studies because complex formation could be fol'Revision received March 1, 1977.

lowed spectroscopically in the absence of indicators and because the rates could be conveniently measured by the stopped-flow method.

Experimental Salicyclic acid and boric acid were Fisher Certified Reagents and were used without further purification after drying in vacuo over phosphorus pentoxide. Buffered solutions were made up by volume from standard solutions of acetic acid, sodium acetate, and sodium chloride, the final volume of the mixture being adjusted to give an ionic strength of 0.1 M. These solutions were used to prepare solutions of salicyclic acid and boric acid of known strength. p H measurements were carried out at 25°C on mixtures of equal volumes of the salicylic acid and boric acid solutions. All kinetic experiments were carried out at 25"C, using a stopped-flow apparatus that has been described previously (ll), except that it was fitted with a quartz cell having a bore of 2 mm square section and length of 20 mm. The flexible light pipes were discarded and light was passed from the monochromator to the cell by means of mirrors. By changing the position of a mechanical shutter, either transmitted light or fluorescence could be measured. In the fluorescence mode, light measurements were made at right angles to the exciting beam, which illuminated a 12 mm length of the reaction cell, starting 8 mm from the mixing chamber. Rate measurements were carried out by following the changes in fluorescence intensity using exciting light at a wavelength of 335 nm. Son~ewhatlarger changes in fluorescence were observed when the exciting light was at 320 nm, but a better signalto-noise ratio was obtained at the longer wavelength, where the output of the 200 W quartz-iodine lamp was

3036

CAN. J. CHEM. VOL. 55, 1977

TABLE1. Rates of reaction of boric acid and salicylic acid at 25°C. The effects of wavelength changes

[Salicyclic acid] (M x lo3)

Can. J. Chem. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/25/17 For personal use only.

pH

[H3B03] (MI

h (nm)

Absorbance method

Fluorescence method

TABLE2. Stability constants for the complexation of boric acid and salicylate ions, 1.043 x M , at p H 5.34 and 25°C CBoa (MI

Absorbance at 322 nm

Absorbance at 332 nm

KHA-b" (M-l)

KHA-bsd (M-l)

0C.O is the initial concentration of boric acid.

b K H A -= 10.61 -

EI =

snectivelv.

+

0.29 M - I . 1610, where and

E>

are the extinction coefficients of the complex 1 and salicylate ion, re-

much greater. The apparatus used a Bausch and Lamb high intensity monochromator set at a bandwidth of 15 nm. It was shown that identical rate constants were obtained using either fluorescence or absorbance measurements. The fluorescence method was preferred because considerable loss of transmitted light - occurs in our apparatus below 400 nm, resulting in a poor signal-tonoise ratio. The rate constants were calculated by an on-line method (12) and the results are sun~marizedin Table 1. Each value is the mean of at least ten independent experiments. Good first-order rates were obtained over at least four half lives.

Results Stability constants, KHz, and KHA-,for the formation of the complex 1 from salicylic acid and salicylate ion, respectively, are defined according to [I] and [2]. Values of KHA-have been obtained at 322 and 332 nm from the measure-

ments of the absorbances of ten different solutions of boric acid and salicylic acid. l-he p~ the was 5.34 and measurements were made at 25°C using a Gilford s ~ e c t r o ~ h o t o m eter. model 2400-2. For these conditions less than 0.5% of the ligand is in the protonated form. Calculations were based on the method of Rose and Drago (13) and the results are summarized in Table 2. The average value of KHA-= 10.6 can be compared to the approximate value of 17 3, previously obtained at an ionic strength of 3 M and an unknown temperature using infrared data (14). Rose and Drago's method assumes that only two absorbing species are present in solution. Support for this assumption is gained from the convergence of the KHA--' values in the Rose-Drago plots and from the

,

3037

QUEEN

Can. J. Chem. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/25/17 For personal use only.

fact that excellent isosbestic points were observed at 258 and 299 nm when the p H of the solutions was 5.34. [complex] [H '1 [I] = [H2A][boric acid] c21

KHA-=

[complex] [HA-] [boric acid]

be readily shown that the observed rate constants are given by [4].

Within the range of p H used, boric acid is virtually undissociated and only the first ionisation of salicyclic acid needs to be considered. The boric acid was always in large excess over salicyclic acid but at concentrations where the formation of dimers and larger aggregates could be neglected (15). For these conditions, the simplest reaction scheme is given by [3]. It can

instead to check the validity of using [4] to obtain k, and k,. It can be seen that the values of these rate constants are in good agreement with those of A and B, respectively, obtained at low hydrogen ion concentration (pH = 5.29). Using these values of the rate constants, the ratio k,/k, = 10.7 0.8 M - I is in good agreement with the value of KHA-given in Table 2. If k,/k, is calculated from the data obtained at p H 6.29, it equals 11.1 -t 0.3 M - l . The best value of k,' leading to calculated values of B in agreement with those observed is 135 s- M - l.

+

In agreement with this scheme, plots of k,,, against boric acid concentration at constant p H were linear. The slopes and intercepts of these plots were obtained by a least-squares method and the values are given in Table 3. The values of the intercepts, A, varied linearly with hydrogen ion concentration, as required by [4]. The corresponding slopes, B, also increased linearly with hydrogen ion concentration, within the limits of the experimental errors, which is not inconsistent with [4], depending on the values of the constants and the range of hydrogen ion concentrations used. K,, the acidity constant of salicylic acid, has been accurately measured in these laboratories by Dunn and Kung (16) at 25"C.and an ionic strength of 0.1 M, the value being 1.02 x Values of k, and k,' were obtained from the variation of A with hydrogen ion concentration by a least squares method. k, was similarly calculated from the hydrogen ion dependence of B. Values of k,, k,', and k, obtained in these ways are shown at the bottom of Table 3. The results obtained at p H 5.29 were not included in these calculations, being used

'

The simple mechanism shown in [3] requires However, that K, = kf1kb/kb'k,= 1.02 x the rate constant quotient equals 2.8 x lop3, which suggests that the simple mechanism is either incorrect or incomplete. A logical expansion of this scheme allows for stepwise formation of the final complex 1 from both salicylate ion 2 and salicylic acid 4, as shown in [6]. It might be argued that k, and k-, should be associated with the quantities k,' and k,', so that K, in [4] should be replaced by a composite-K,', possibly equal to k-,K,/k3. However, it is not possible to choose a value for K,* which will simultaneously yield k,' and a fit to the observed values of B. We have not attempted a full kinetic treatment of [6]. Instead we have preferred to argue from a consideration of a treatment in-

3038

CAN. .I.CHEM. VOL. 55, 1977

TABLE 3. Rates of reaction of boric acid and salicylic acida at 25°C and a n ionic strength of 0.1 M

K,kr

+ [H+]kfJ + [H+l

kobs (s - I)

A = (kb k b f [ H + I ) (s-l)

5.290

0.03380 0.05995 0.07453 0.09387 0.1143

4.90 5.90 6.48 7.18 8.01

3.55k0.06

39.4k0.8

39.4

4.625

0.03333 0.05615 0.07314 0.09284 0.1156

4.98 6.07 6.68 7.50 8.33

3.969

0.03679 0.05662 0.07636 0.09115 0.1170

5.86 6.74 7.63 8.39 9.50

3.648

0.03623 0.05472 0.07557 0.09570 0.1231

6.71 7.79 9.02 10.2 11.9

3.450

0.03688 0.05725 0.07352 0.09261 0.1148

7.63 8.88 9.96 11.2 12.6

5.25+_0.03

64.1k0.4

63.7

PH

Can. J. Chem. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/25/17 For personal use only.

B=

+

[HBBO~I (MI

"[Salicylic acid] = 1.5 Y 10-3 M, K , 360s-I M - I , kr'= 135 s-I M - l .

=

1.02 x 10-3, k,

volving only the species 1, 2, and 3; i.e. for the situation that would hold when the p H > 5. If all three species change concentration simultaneously, then they will also contribute to the changes in fluorescence. It can then be shown that the value of k,,, is given by [7].

C ~ kobs I =

(Ye + Ze) ln { ( Y e + Z e I - ( Y + Z )

?

RZe)l( Ye + e' ) Q(Ye - Y ) + R(Z= - Z) (Ye + Ze) - (Y + 2 )

- (Q Ye

=

3.62 i 0.d8

s-I,

Ka

(s-I

kr = 38.9

Bcaic,

M-I)

+ 2.6 s - I

M - ' , k;

=

4483

+

In this equation, Ye = (C," - C,'), Y= (C,' - c,~), Ze = (Cle - ClO). The C values are concentrations at times 0, t , and equilibrium, as indicated by the superscripts 0, t , and e, respectively. The quantities P, Q, and R are the fluorescence coefficients of the species 2,3, and 1, respectively. These are likely to be different at a given wavelength and to change differently with wavelength. Hence, kobs would be expected to vary with the wavelength of the incident light in a given experiment where the boric acid concentration and pH are fixed. If the reaction is followed by monitoring changes of absorbance,

Can. J. Chem. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/25/17 For personal use only.

QUEEN

the P,Q, and R values are extinction coefficients. It is likely that there will not be a 1 : 1 relation between these quantities for fluorescence and absorbance, so that kobsmight be expected to vary depending on the method used to measure it. The data in Table 1 show that neither of these differences are observed when the p H is 5.29. Indeed, the values of k,,, are independent of wavelength at p H 4.63, where the salicylic acid/ salicylate ion ratio is 0.023. This suggests either that the complex 3 is not formed at all or that it reaches a low steady-state concentration early in the course of the reaction. The possibility that Q and R are equal is considered to be less likely. When the concentration of the complex 3 is small, Y and Ye are small causing the second term of [7] to become zero. The steady state approximation may be applied to complex 3 and kobsis then given by [8].

C81

kobs =

k - 1k2 k-, k,

+

k1k2 +

k-,

+ k,

[boric acid]

Hence,

This is consistent with the observation that the kinetic and equilibrium values of KHL-are the same. The conclusion that complex 3 does not contribute significantly to changes in fluorescence suggests that the inequality of K, and the quotient kf'kb/kbfkf may require that complex 5 is not in low or steady state concentration during the course of the reaction. If this is so, the values of k,,, at low pH, when no salicylate ion is present, should be given by an equation similar to [7] and so should be wavelength dependent. No changes in light absorption or fluorescence could be detected at p H 1 and were too small to be useful at p H 2. A similar, but more complicated equation than [7] should apply to kobsfor intermediate p H values where both salicylic acid and salicylate ion react with boric acid. Table 1 shows that kobsis indeed wavelength dependent at p H 3.45. It is therefore proposed that the results for the boric acid, salicylic acid system support a stepwise formation of the complex 1 from salicylic acid and probably from salicylate ion also. For this system, complex 3 is in low concentration during the reaction but complex 5 is not.

3039

Pizer has previously assumed the formation of complexes corresponding to 3 and 5 in the reactions he has studied. However, he also assumes that these intermediates are formed at diffusion controlled rates, so that k, and k,' are considered to measure the rates of ring closure on loss of H,O and H,O+, respectively. Although the present data have not been interpreted in the same way, mechanism [6] would appear to be capable of accommodating a variety of different possibilities since the various steps would be expected to be sensitive to the effects of substituent groups in the ligands. Indeed, it may be possible on this basis to explain the fact that k, is larger than k,' for the complexation of lactic and boric acids, this reaction being the only one so far reported where the fully protonated acid reacts more slowly than its conjugate base. It is worth noting that lactic acid is the only ligand having an electron donating group close to the reaction sites. Moreover, it is not certain that the same group acts as the nucleophile towards the boron atom in all cases. It may well be that the phenolic group acts in this way in the fully protonated ligands but that the carboxylate group is the reactive centre in the corresponding ionized ligands, at least in some cases.

1. J. KNOECKand J. P. TAYLOR. Anal. Chem. 41, 1730 (1969) and literature cited therein. 2. L. B. MAGNUSSON. J. Inorg. Nucl. Chem. 33, 3602 (1970). Recl. Trav. Chim. Pays-Bas, 51, 955 3. N. VEERMAAS. (1932). 4. S. FRIEDMAN and R. PIZER.J. Am. Chem. Soc. 97, 6059 (1975). 5. U. WESER.Z . Naturforsch. 22,457 (1967). 6. B. PESETSKYand N. R. ELDRED.Tetrahedron, 25, 4137 (1969). 7. K. KUSTINand R. PIZER.J. Am. Chem. Soc. 91, 317 (1969). 8. S. FRIEDMAN, B . PACE,and R. PIZER.J. Am. Chem. SOC.96, 5381 (1974). and R. PIZER.J. Am. Chem. Soc. 97, 9. S. FRIEDMAN 6059 (1975). and R. PIZER.Inorg. Chem. 15,978 (1976). 10. G. LORBER J. E. CROOKS,and A. QUEEN.J. Phys. 11. E. F. CALDIN, E, 6,930 (1973). 12. A. QUEEN,J. L. CHARLTON, E. DAWSON,and W. BUCHANNON. Chem. Instrum. 6, 153 (1975). 13. N . J. ROSEand R. S. DRAGO.J. Am. Chem. Soc. 81, 6138 (1959). Acta Chem. Scand. 14. R. LARSSONand G. NUNZIATA. 26, 1503 (1972). 15. N. INGRI.Acta Chem. Scand. 16,439 (1962). 16. G. E. DUNNand FEI-LINKUNG.Can. J. Chem. 44, 1261 (1966).