Stereo-electronic Effects in the Alkaline Hydrolysis of Benzamides and

Quantitative structure-activity relationships and the unnamed science. Corwin Hansch. Accounts of Chemical Research 1993 26 (4), 147-153. Abstract | P...
0 downloads 0 Views 418KB Size
May 5, 1962

1625

HYDROLYSIS OF BENZAMIDES AND BENZONITRILES

[CONTRIBUTION FROM THE NATIONAL INSTITUTE O F ARTHRITIS AND METABOLIC DISEASES, NATIONAL INSTITUTES OF HEALTH,

BETHESDA14, MD.]

Stereo-electronic Effects in the Alkaline Hydrolysis of Benzamides and Benzonitriles' BY LOUISA. COHENAND WILLIAMM.

JONES

RECEIVEDOCTOBER27, 1961 (Pseudo)-first-order rate constants for the alkaline hydrolysis of p-substituted benzonitriles and benzamides are reported. ki kz Values for kl in the consecutive series RCN +RCONH2 -+- RCOOH NHs were obtained by automatic approximation and curve-fitting on a n electronic calculator. Introduction of t-butyl groups ortho t o the hydroxyl of p-hydroxybenzonitrile and p-hydroxybenzamide decreases the hydrolysis rate 30-fold. The phenomenon is attributed t o a highly increased density of negative charge at the site of nucleophilic attack; the latter is, in turn, the result of steric hindrance t o solvation of the charge a t the phenolate anion.

+

The negative charge of a phenolate anion is generally considered to reside partly on oxygen and partly on the ortho and para carbon atoms of the aromatic ring. When the ring carries an unsaturated functional group in the o- or p-position, further resonance structures may be written for the anion, in which the charge is dispersed by an additional heteroatom. In an earlier paper of this series,2i t was suggested on the basis of physical data that the introduction of bulky substituents ortho to the phenolate anion could alter the distribution of negative charge so that less hindered regions of the resonance hybrid would command a larger share of the electron density. The phenomenon was considered to be due t o steric hindrance to charge dispersion by the solvent a t the phenolate anion. The existence of high electron density in the p-position is also suggested by chemical evidence. Thus, from the benzoylation of I there was obtained an enol-benzoate 113 and from 111, both the benzophenone IV and its enol-benzoate IVa. However, simple steric hindrance t o acylation a t the phenolic site would

account for such data equally as well. A test of the concept was sought, therefore, by studying the rates of reactions occurring a t the p-position, where the effect of charge distribution could be observed in the absence of direct steric interference by the t-butyl groups. Indeed, we have already shown that, in an electrophilic reaction such as the acid-catalyzed decarboxylation of p-hydroxycinnamic acid, the introduction of t-butyl groups ortho to the phenolic hydroxyl effects a moderate increase in the rate of decarboxylation. Because of electrostatic repulsion, however, the same structural changes should lead to a decrease in the rate of a reaction involving nucleophilic attack at or near a center of negative charge. OH

OH

C=NH.HCI I

COOH

XI1

OCZHS V

6

1,NaOH

HCI fCzH50H

1

2,HCl

H202_

I

CN VI

I1

CONHZ VIII

C H ~ I ~ K ~ C O ~

OH CsH.COCI, I

I

CS

CONHz

VI1

IX 1. NaOH12. HCI

J. OCHj

1 . HzO, A 2, NaOH 3, HCI

I \'a ( 1 ) Paper I11 of a series on phenol-dienone tautomerism; for paper 11, cf. L. A. Cohen a n d W. hl. Jones, J . A m . Chcm. Soc.. 82, 1907 (1960). (2) L. A . Cohen, J . Oyg, Chcm., 22, 1333 (1957). (3) K. Ley, Angew. Chcm., 70, 74 (19%). (4) T. H. Coffield, A. H . Filbey, G. G. E c k e and A. J. Kolka, J . A m . Chcm. Soc., 79, 5019 (1957).

C= NFI.HC1 OC,Hs

COOH XI

X

To test the hypothesis, we have examined the rates of alkaline hydrolysis of a series of p-substi-

1626

LOUISA. COHENAND WILLIAMM.

JONES

Vol. 84

5.85 g. When a n aqueous solution of the imino ester hydrochloride was neutralized with sodium bicarbonate arid heated, the oily imino ester gradually solidified, the product being identified as the original methoxy-nitrile. Apparently, even under conditions as mild as pH 8, the imino ester lost the elements of ethanol readily to re-form Experimental' the nitrile.1° Only by heating a solution of the imino ester hydrochloride in water alone could hydrolysis be effected, Materials.-Where available, commercial materials were the oily ester separating within a few minutes. The mixused and recrystallized when necessary. p-Hydroxybenzoture was then made alkaline with 10-12 equivalents of 2 N nitrile (Aldrich) was recrystallized from hot water; m.p. 113-114'. p-Methoxybenzamide was prepared from anisoyl sodium hydroxide, enough ethanol added to effect complete solution and the mixture refluxed for several hours. After chloride and concentrated ammonium hydroxide and reremoval of ethanol under reduced pressure and acidification crystallized from water; m.p. 164-165'. p-Aminobenzof the clear, aqueous solution, a crystalline precipitate of amide was prepared from p-aminobenzonitrile by oxidation the methoxy acid separated. It was filtered, dried and rewith aqueous hydrogen peroxides and recrystallized from crystallized from ligroin (90-100°), affording a 90% yield hot water; m.p. 183-1S4.5". p-Hydroxybenzamide was of colorless granules, m.p. 192.5-194.5'. prepared from p-hydroxybenzonitrile and hydrogen peroxide Anal. Calcd. for ClaHz403: C, 72.69; H, 9.15. Found: by a similar procedure and recrystallized from water. C, 72.45; H, 9.06. After the solvated product had heen dried a t 100' in uacuo, it melted at 161-162". p-Carboxybenzamide (p-phthalamic Kinetic Runs.-Reaction mixtures were made up by addacid) was prepared from p-cyanobenzoic acid and hydrogen ing 0.01 mole of compound t o 100 ml. of 2 M sodium hyperoxide .$ droxide in ethanol. The solutions were thus 0.1 3,5-Di-t-butyl-4-methoxybenzonitrile(VII).-To a solu- molar with respect to compound and contained a twentytion of 23.1 g. (0.1 mole) of 3,5-di-t-butyl-4-hydroxpbenzofold quantity of alkali. Corning alkali-resistant glass, nitrile (VI)l in 800 ml. of dioxane was added 60 ml. of No. 7280, was used for the reaction vessel. The solutions methyl iodide and 40 g. of powdered, anhydrous potassium were heated a t reflux (82') in a n oil-bath kept at 125" while carbonate. The mixture was heated a t reflux for 48 hours, dry nitrogen was passed through the system at the rate of filtered and the residue was washed with 100 1111. of methylapproximately 20 ml. per minute. Control experiments ene chloride. The combined filtrate and washings were conshowed no significant loss of solvent from the reaction vessel centrated to a yellow sirup which was dissolved in 50 ml. of after flushing with nitrogen for 48 hours a t such a rate. methanol. The methanol solution was diluted with 250 rnl. The effluent gas was bubbled through 25 rnl. of 2%' boric of N sodium hydroxide and the precipitate of methoxyacid containing Tasliiro indicator." At various time internitrile was collected by filtration. The crude product was vals the boric acid solution was titrated back to the endrecrystallized from methanol-water after treatment n-ith point with 0.5 M hydrochloric acid. Blank experiments charcoal to give 19.7 g. (8070) of material melting a t 75-76' were performed by adding weighed quantities of ammonium (1it.l 73-75'). I n a number of similar preparations, yields chloride to the alkaline solution in the reaction flask. In varied from 75-8870. each case, at least 907, of the theoretical ammonia titer was 3,5-Di-t-butyl-4-hydroxybenzamide (VIII).-To a solurealized within 5 minutes, suggesting t h a t only in the case tion of 4.62 g. (0.02 mole) of 3,5-di-t-butyl-4-hydroxybenzo of very rapid hydrolyses could a time lag introduce signifinitrile (VI) in 150 nil. of ethanol was added 40 ml. of water cant error. Several kinetic runs were repeated using inand 20 ml. of 30y0 hydrogen peroxide. The solution was creased rates of nitrogen flow without significant change adjusted t o p H 8.5 with 4 ili sodium hydroxide and kept a t in the results. Hydrolysis was carried to iO-8070 comple50-60' for 24 hours. After acidification to pH 1, the solution except for the two very slow cases, VI and VIII, in tion was concentrated t o one-fourth the original volume and which the reaction was stopped after 20V0 completion. the product was collected by filtration. The colorless solid The mixture resulting from hydrolysis of VI was worked was washed thoroughly with ether and recrystallized from up for product recovery. The alkaline solution was acidimethanol-water; 3.5 g. (7Oyc),m.p. 266-266.5'. T h e fied axid concentrated to dryness. The residue was exsolvated product was dried in Z~UCZLOa t 100" for 24 hours; tracted with ether and the ether solution was extracted with m.p. 266-266.5'. aqueous sodium hicarbonate. The hydroxy acid XIIe was obtained by acidification of the bicarbonate extract. The Anal. Calcd. for C I J - I ~ ~ O ~C, S : 72.25; H , 9.30; N, ether solution was taken to dryness arid the residue was 5.62. Found: C, 71.98; H , 9.48; ?;, 5.63. 3,5-Di-t-butyl-4-methoxybenzamide (IX) .-The methoxy- extracted with hot ligroin (90-100") to dissolve the hydroxy-nitrile VI, leaving the hydroxy-amide VI11 as a nitrile VI1 was converted into the methoxy-amide by oxidation with hydrogen peroxide as described above for the residue. I n a similar manner X I 1 was recovered from the hydroxy-amide. I n this case, the crude, dried product was hydrolysis of VIII. Recovery results agreed satisfactorily with data obtained from ammonia titration as well as from washed with hot benzene prior to recrystallization from spectrophotometric assay (see below). methanol-water. A yield of 4.2 g. ( 8 0 % ) of needles, m.p. Several attempts were made to analyze mixtures of 244-246', was obtained. The solvated product was dried nitrile, amide and acid by gas chromatography. Despite zn oacuo at 100" for 24 hours; m.p. 245-246'. the use of silicone columns operated as high as 250", only Anal. Calcd. for C1sH?sO2J: C , 72.96; H , 9.57; X, benzonitrile and anisonitrile could be volatilized and the 5.32. Found: C, 72.96; 13, 9.67; N, 5.37. method was abandoned for the present study. For VI and 3,5-Di-t-butyl-4-methoxybenzoicAcid (XI).-A solution VIII, differences in ultraviolet spectra were useful for folof 4.90 g. (0.02 mole) of 3,5-di-t-huty1-4-methoxybenzo- lowing hydrolysis rates. Aliquots of the reaction mixture were nitrile ( V I I ) in 50 nil. of ethanol was saturated with hy- withdrawn, diluted as required wit11 the same alkaline soldrogen chloride and the mixture was stored at room tem- vent, and the intensity measured a t 340 mp. In the alkaline perature for 48 hours. The solvent was removed under medium, VI11 had a n E value of 13,000 at 340 mp, whereas reduced pressure, leaving a colorless, crystalline residue VI and X I 1 had intensities below 50. which was triturated with ether, filtered and dried. The yield of imino-ester hydrochloride X, m.p . 247-24So, was Results

tuted benzonitriles6 and benzamides,6 both with and without t-butyl substitution, and have observed impressive changes in reaction rate in the expected direction.

(5) For an earlier study of the kinetics of alkaline hydrolysis of benzonitriles, cf. Y.Ogata and M. Okana, J. Chcm. SOC.Japan, 70, 32 (1949). (6) Cf.1. Meloche and K . J. Laidler, J . A m . Fhcm. Soc., 73, 1712 (1951): E. E. Reid, A m . Chcm. J.,21, 284 (1899); 24, 397 (1900). (7) Melting points are uncorrected. Ultraviolet spectra were run on a Cary recording spectrophotometer, model 14. The authors thank Mr. H. G. McCann and his associates of this Institute for carrying out the microanalyses, (8) M. T. Bogert and L. Kohnstamm, J . A m . Chcm. Soc.. 25, 483 ( 1903). ( 0 ) p. Kattwinkel and R . Wolffenstein, Ber.. 31, 3223 (1904).

For all amides studied, second-order or (pseudo)first-order rate laws were obeyed equally as well (the ratio of alkali to amide being 1 O : l or 2O:l) up to 700/, hydrolysis. Rate constants were determined graphically and are summarized in Table I. (IO) Similar difficu1tie.r were encountered in the preparation of X I 1 from V. The procedure described below should be substituted for that reported in the earlier paper (ref. 2 ) . (11) N. C. Davis and E. L. Smith, "Methods of Biochemical Analysis," Vol. 2, Interscience Publishers, Inc., New York, N . Y., 1955, p. 232.

HYDROLYSIS OF BENZAMIDES AND BENZONITRILES

May 5 , 1062

1627

TABLE I ( PSEUDO)-FIRST-ORDER RATECONSTANTS FOR THE ALKALINE OF BENZAMIDES’ HYDROLYSIS H series

Benzamide

h

X lor, +in.-$ Di-f-butyl series

p-Carboxy 3100 .. p-Hydrogen 2460 .. p-Methoxy 963 505 p-Amino 285 .. p-Hydroxy 63 2 4 Reaction mixtures were composed of 0.01 mole of compound in 100 ml. of 2 M sodium hydroxide in 60% ethanol; hydrolysis rates were measured a t reflux (82’).

Since the hydrolysis of nitriles involves a set of consecutive reactions following, in this case, (pseudo)-first-order kinetics, the set of integrated equations derived by Esson12 for the reaction sequence A

I

I

-1

ki

kt

+B +C ( + D )

IO00

0

may be applied, where A

and A +B

5

2000

MIN.

AOC-kif

Fig. 1.-Comparison of experimental and calculated rates of hydrolysis of benzonitriles: -, calculated curve; 0. experimental values.

+ C = A o o r A o- C = A + B

I

I

I

Then p-COOH

1

By setting A . = 1 and C = per cent. of theoretical amount of ammonia released/100 k 1- C= +A (e-h; e-klt} (4) kz - kl Using the values of kz recorded in Table I and tentative values for k1, eq. 4 was solved a t various values of t and the results compared with the experimentally measured values of C or D. By successive approximations for the value of K1, a satisfactory fit to each experimental curve was realized. Computation was greatly facilitated by programming eq. 4 for an IBM 650 computer which performed successive approximations and curve-fitting automatically. l 3 A representative comparison of experimental and calculated results is shown in Fig. 1. Values for kl are collected in Table 11.

-

TABLE I1 ( PSEUDO)-FIRSP-ORDER RATECOXSTANTS FOR THE ALKALINE OF BENZONITRILES~ HYDROLYSIS Eenzonitrile

ki X 106, min.-l I1 series Di-1 butyl series

$1-Carboxy 12000 .. p-Hydrogen 7250 .. p-Methoxy 3000 2000 630 .. p-Amino p-Hydroxy 125 4 a Reaction mixtures were composed of 0.01 mole of compound in 100 ml. of 2 M sodium hydroxide in 60% ethanol; hydrolysis rates were measured at reflux (82’). (12) W. Esson, Phil. Trans. Roy. SOC.(London), 166, 220 (1866): A. A. Frost and R. G. Pearson. “Kinetics and Mechanism,” John Wiley and Sons, Inc., New York. N . Y., 1953, p . 153. (13) We are indebted t o Dr. N. 2. Shapiro and Mr. R . H . Brunelle, Computation and Data Processing Branch, National Institutes of Health, for performing the necessary computation.

-‘I

Y

1

-*r

5 1 I

-3

t

-t L

-5

0 N I T R I L E S - H Series

A NITRILES-1-Butyl Series .AMIDES-H

Series

A AMIDES-I-ButyI Series

A

A

L -1.0 -0.5

Fig. 2.-Hamrnett

O

plot for alkaline hydrolysis of benzonitriles and benzamides.

U’hen log k was plotted against for each series of unhindered compounds, an approximately linear relationship was obtained (Fig. 2). The value of p for the hydrolysis of benzamides was found to be 1.5, while that for benzonitriles was 1.75.

Discussion The alkaline hydrolysis of propionitrile has been shown to proceed a t a rate one-tenth that of pro(14) H . H. JaffC, Chem. Revs., 63, 191 (1953).

LOUISA. COHENAND WILLIAMN. JONES

1628

pionamide. l5 Aromatic nitriles, on the other hand, appear to hydrolyze from 2-4 times as fast as the corresponding amides, according to the data of Tables I and 11. The inverted sequence may be due to resonance factors. As the result of the formation of a tetrahedral intermediate in benzamide hydrolysis (XIII), resonance coupling between the amide and the aromatic ring is totally lost; however, the trigonal intermediate in nitrile hydrolysis (XIV) is still capable of resonance interaction with the ring. Consequently, benzamides might be expected to show a higher energy of activation for hydrolysis.

Vol. 84

by Vlc and VIIIc; (c) impaired ability of hydroxide ion to add to the quinonemethine structures VIc and VIIIc in a Michael fashion, as compared with the unhindered analogs. Although no

VIIIb t

VIb

f

XI11

C=N-

XIV

From a comparison of the rate data (Tables I and 11) for p-methoxybenzamides as well as for pmethoxybenzonitriles, i t is evident that t-butyl groups affect the hydrolysis rate to a relatively small extent. The observed difference may result from a small steric effect of the bulky t-butyl group mefa to the site of nucleophilic attack or, more likely, from increased contributions of forms such as XV in addition to a small inductive effect. However, the introduction of t-butyl groups into the phenolic derivatives VI and VI11 serves to reduce the rates of alkaline hydrolysis by a factor of 30. l6 Undoubtedly, the increased contributions of I-Ib and 17c and of VIIIb and lTIIIc to the cH3;+c/

x \’ respective resonance hybrids are reflected in the rates of hvdrolysis. The effect may be attributed to (a) greater electrostatic repulsion of hydroxide ion by VIb and YIIIb; (b) a similar repulsion (15) B. S. Rabinuvitch a n d

C . A Winkler, C a n . J . Revearch, 20B,

183 (lY42). (16) During t h e course of this investigation, t h e resistance of VI11 t o alkaline hydrolysis was also observed by E. Muller, A. Rieker, K. Ley, R. h‘layer and K. Schiemer, Chem. Bcr., 92, 2278 (1959).

VIIIC

VIC

rigid assessment of the relative importance of each factor can be made a t present, i t is instructive to examine the effect of t-butyl groups on the ultraviolet spectra of the compounds in question. The data of Table 111 suggest that VIc and VIIIc contribute measurably to the structures of the phenolate anions and may affect reaction rates to a greater extent than VIb or VIIIb.” Preliminary experiments on the alkaline hydrolysis of ethyl p-hydroxybenzoate a t 25’, both with and without TABLE I11 EFFECTOF ALKALIO S ULTRAVIOLET SPECTRA OF PHENOLS

0 ‘3

“H,

“H:

o$WO

---Wave length. muH series Di-t-b;tyl series AlkaAlkaNeuSeuline linea tral tral ~

Compound

~~~

y

~

~

~

l

252 281 252 303 p-Hydroxybenzonitrile 257 294 262 325 p-Hydroxybenzamide 262 308 p-Hydroxybenzoic acid 259 281 258 300 262 330 Ethyl p-hydroxybenzoate 0 n’eutral spectra were measured in 957; ethanol; alkaline spectra in 2 M sodium hydroxide (BOY0 ethanol).

t-butyl substitution ortho to the phenol, indicate that a considerable stereo-electronic hindrance to saponification exists in this series as well. (17) G. W , Wheland, “Resonance in Organic Chemistry,” John Wiley a n d Sons, Inc., New York, N . Y., 1955, p. 292.