The acid hydrolysis of alkyl β-D -galactopyranosides

CARBOHYDIL4TE

THE ACID

59

RESEARCH

HYDROLYSIS

OF ALKYL

P-D-GALACTOPYRANOSIDES

C. K. DE BRUYXE AND G. VAN DER GROEN

H.I.K. W., Lab. Anorg. Scheik. A., Ledeganckstraat35, B-9000 Gent (Belgium) (Received March 2&t,

1972; accepted for publication,

March 29th, 1972)

ABSTRACT

Twenty-one alkyl B-D-galactopyranosides were synthesised and hydrolysed in hydrochloric acid. Application of the Hammett-Zucker criterion indicates a unimolecular (A-l) mechanism. The activation parameters indicate an isoenthalpic reaction series. Galactosides having secondary-alkyd aglycon groups belong to the same series, but the rate of hydrolysis is slightly increased by the entropy factor. The influence of the aglycon group can be correlated with the Robinson-Matheson @ parameter. INTRODUCTION

In previous paperslv’, we described the acid-catalysed hydrolysis of series of alkyl and aryl P-D-glycosides and the influence3 of the acid concentration on the rate coefficients. It was the purpose of this work to investigate, in a simiIar manner, the hydrolysis of a series of alkyl /?-D-galactopyranosides, and to compare the results with those of our own previous series, and with analogous series investigated by other authors4*5. For further details on the acid-catalysed hydrolysis of glycosides, the recent review of BeMiller may be consulted. RESULTS AND DISCUSSION

Twenty-one alkyl /3-D-galactopyranosides were hydrolysed in M aqueous hydrochloric acid at five different temperatures. The pseudo-first-order reaction coefficients at three temperatures and the activation parameters are presented in Table I. Ir@ence

of the acid concentration

For five of the a&y1 /?-D-gaIactopyranosides, the pseudo-first-order rate coefficient was determined at constant temperature and various concentrations of hydrochloric acid (Table II). If the reaction proceeds via a slow, rate-limiting, unimolecular heterolysis of the glycoside conjugate acid (as is generally accepted6), log k, should show a linear dependence (slope- 1) on the Hammett acidity function Carbohyd. Res., 2.5 (1972)59-65

3

Y

e

s

Q

t;:

:

$

$.r;

2.

6

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Glycosiile

Hcptyl Octyl 2-Methyl-I-propyl 4-Methyl-I-pcntyl 2-Methyl-1-pcntyl 2-Butyl 3.Pentyl 2-Pentyl a-Hcxyl 2-Heptyl 2.Octyl Benzyl 3-Phcnylpropyl Cyclopcntyl cyclollcxyl

HCXYI

Methyl Ethyl Propyl Butyl Pen tyl

Alkyl grmrp

M HYDROCHLORIC ACID

1.28 1.43 1.56 1.67 1.57 l 1.64 1.59 1.65 I,94 1.58 I.84 3.16 4.00 3.29 3.32. 3.27 3-06 1.35 1.28 2.19 2.85

60”

5.38 5.97 6.57 6.99 6.53 6.80 6.60 6.81 8.11 6.62 7.67 13.3 16.6 13,R 14,o 13.7 1288 5.59 5.38 9.02 11.7

70”

10Sk, (set- ‘)

-I-14.9

f14.5 f15.1 + 14.2 -I-14.3 -t14.3 -1-14.4 f 14.2 -I-14.4 -I-15.2 -t 15.0 -i-16.5 -I-1602 $16.6 -I-16.6 -t 16.3 -i-16,3 + 14.0 -I-14.9 -l-14.3 -+ 14.8

20.9

23.0 25.5 27.0 25.0 26.0 25.3 26.0 31.3 25.7 29.5 51.3 63.6 53.5 54.1 52.7 49.4 21.4 20.9 34.2 44.3

80”

AS+ (GO”) (cddeg. - ’ rrrolc- ‘)

27.0 27.0 26.9 27.1 26.9 26.9 26.9 26.9 27.0 26.9 26.8 26.4 26.3 26.4 26.4 26.4 26.5 27.0 21.0 26.7 26.5

AG* (kcnholc‘)

PSEUDO-FIRST-ORDER RATE COEFFIClllNTS AND ACTIVATION I’ARAMOTCRSPOR THE HYDROLYSIS OF ALKYL B-D~CALACTOPYRANOSlDIiS IN

TABLE I

32.0 f0.2 31.8 f0.1 31,9 f0.3 31.9 ztzo.1 31.7 fO.l 31,G rtro.3 31.7 j10.3 31-6 f0.3 31.8 f0.3 3280 f0.1 31.8 f0.3 31,9 f0.2 31.7 10.2 31.9 &O.l 32.0 dzO.2 31.8 &O.l 31.9 dzo.3 31.7 -ho.1 31.9 fO.l 31.5 zlzo.1 31.4 f0.3

AH+ (/ccn/.no/e- ‘)

0

cn

HYDROLYSIS

61

OF GALACTOPYRANOSIDES

H, (Zucker-Hammett

E x p erimentally,

criterion’-‘).

cases, log k, indeed showed a linear dependence

it was found

that, in all of the

on Ha, but that the slopes

deviated from unity. A least-squares fit of the data yielded the values given in Table III, where b represents the slope, a the intercept, s, the estimated standard error on b, s,.~~the standard error on the estimate, and r the correlation the cases, the deviation from unity is statistically deviations

do not invalidate

the Zucker-Hammett

coefficient. Although, in most of significant (t-test), these small criterion,

and thus the hydrolysis

should proceed by the unimolecular A-l mechanism, without participation of water in the rate-limiting step. We did not use the Bunnett” and Bunnett-Olsen” criteria to obtain further evidence for this mechanism, because we have shown3 that, in cases of glycoside

hydrolysis

these criteria TABLE

where the Hammett

are of little mechanistic

slope b remains constant

(linear function),

value.

II

INFLUENCE

Agiycon

OF THE

ACID

CONCEhJRATION

NC1 HOa (at)

gfonp

Methyl Hexyl 3-Pentyl 2-Butyl . Benzyl

ON

-0.20 I

THE

RATE

-0.69 2

1.26 1.59 4.66 3.22 1.41

3.98 5.25 15.2 10.7 4.34

COEFFICIENT

105k,

(SEC-‘)

AT

60”

-32.5 0.87

- 31.05

- 41.40

-:.76

6.10 8-55 24.0 18.1 6.84

10.4 13.6 37.5 27.9 10.3

22.4 30.9 86.1 65.8 22.3

48.2 68.4 209 152 47.2

“From Ref. 22. TABLE SLOPES

III OF THE

Agiycon

group

Methyl Hexyl 3-Pentyl 2-Butyl Benzyl

ZUCKER-HAXfS:ETT

a - 0.085 0.010 0.458 0.303 -0.034

PLOTS

-b 1.016 1.044 1.057 1.074 0.98 1

0.019 0.015 0.004 0.012 0.014

-r

S,;r

0.9989 0.9995 0.9999 0.9997 0.9996

0.027 0.021 0.006 0.012 0.017

Isokiizetic relationship According to the theory of Exner’z*‘3, a relation between the activation enthalpy and entropy can be proved by plotting two values of log k, obtained at two different temperatures, against each other. Using the k, values of Table I, a plot of log 105k, at 80” (TI) MKSLIS log 105k, at 60” (TJ shows a linear relationship. Regression analysis, using all derivatives, yields the equation: log 10Sk,(SO”) = 1.26 f0.9977 s,=O.O09, with ssIx --0.0066,

log 105k1(60”), 1-=0.9992, and iz=21. Crrrbohpi.

Res., 25 (1972) 59-65

C. K. DE BRUYNE,

62 A t-test

on the value

different

from unity.

standard

error on b, this error causes a large uncertainty

isokinetic

of the slope Moreover,

temperature

b=0.9977

notwithstanding

j3. The only possible

indicates

that

G. VAN DER GROEN

it is not signilicantly

the high correlation

conclusion

and the small

in the calculation12V’3

of the

seems to be that b approaches

unity, and thus that j? approaches zero. In the Exner classification, this represents case 2, with dN* constant and a rate controlled by changes of AS*. From Table I, it can be seen that AH* is indeed constant throughout the whole series, which thus represents

an isoenthalpic

series.

An analysis

according

to the

original

LelBer14

procedure naturally gives the same fi = 0 value. It is possible that the isokinetic relationship should be separately5 determined for glycosides of primary and secondary alcohols, respectively. Thus, we calculated the following regression lines for primary and secondary (without cyclohexyl and cyclopentyl)

alkyl groups:

primary,

log 105k,(80”)= with s,,,,=O.O05,

1.212+0.969 log 10%,(60”), s,=O.O3 and r=0.995;

secondary,

log 105k1 (80°) = 1.247+0.927 log 105k,(60”), with sYIX = 0.003, s, = 0.03, and r = 0.997.

It is clear that the slopes do not differ significantly from each other, and thus there is no proof that the two series have different isokinetic temperatures. Hence, we believe that all derivatives hydrolyse isoenthalpic reaction series.

uia the same mechanism,

and

belong

to the

same

Linear free energy relationship It is evident from the values in Table

I that there is no simple relation

between

the log k, values and the electronic (Taft polar c*) or steric (E,) effects of the aglycon group. However, Robinson, and Matheson’ 5 introduced a new parameter @, based on the alkaline hydrolysis of alkyl-substituted 3,5dinitrobenzoates as the standard reaction,

Since the substituent

variation

is in the alcohol

portion

of the ester, there is

some resemblance to the variation of the aglycon part in alkyl glycosides, and thus we tested this parameter_ Using the log 10’ kl at 60” of the derivatives for which a @ value was available (and thus derivatives 7, 10, 11, 15,and 19 are not included), we calculated: log 105k1(600) with reaction

=0.054-0.211@, constant

p’ = -0.2llf0.016,

sr,= = 0.04,

Y = 0.966,

and 12= 16.

Although a linear relationship probably exists, the correlation coefficient r is not as high as those listed by Robinson and Matheson * 5. In order to prove that the relation is more than a coincidence, we also analysed, in the same way, the reaction series of BeMiller For alkyl D-ghCOand Doyle5, and our own series of alkyl P-D-xylopyranosides’. Carbohyd. Res., 25 (1972)

59-65

HYDROLYSIS

63

OF GALACTOPYRANOSIDES

pyranosides in 0.5~ sulphuric acid at 80” (see ref. 5, Table II, and ref. 4, Table VI), the following calculations were made:

a-series, log 106k1(800)

= 0.903-O-217@,

with p’ = - 0.217 +0.05,

s,./~ = 0.09, r = -0.82,

and 12= 10;

B-series, log 106k,(80”) = 1.401-0.160@, with p’= -0.160+0.04, s,=O.O6, r= -0.84, and n = 8.

For alkyl /3-D-xylopyranosides in 0.5~ hydrochloric acid at 60” (see ref. 1, Table I), calculation

yields:

log 106k1(600) =0.793-0.297@, with p’ = -0.297 kO.029, s,,,~ = 0.06, r = -0-958, It is noteworthy

that

tert-butyl

#?-D-xylopyranoside

and IZ= 11. is the only derivative

of the

xyloside series which does not fit the relationship (log 106k, = 2.887; @ = -3.34). However 7 it is knownI that glycosides of tertiary alcohols hydroIyse with heterolysis of the exocyclic oxygen-aglycon bond, and thus the mechanism is different_

Although the correlation coefficients of the glucoside series are not very significant, we believe that the @ parameter does have some mechanistic value for these glycoside series. In particuIar, the fact that the reaction constants p’ are of the same magnitude, and thus that the influence of the aglycon group is more or less the same in each of these very similar series, seems to corroborate the value of @. The interpretation of @, which contains a steric as well as an electrical factor, is more difficult than the interpretation of the more-classical Taft or Hammett substituent constants. However, because of the absence of steric effects ’ *% ‘, in accordance with the unimolecular

nature

of the rate-limiting

step, it can be assumed

that

Qi

reflects predominantly the polar effects of the aglycon group. All derivatives of our galactoside series have a&con groups with negative @ vaIue (electron sources). The fact that the reaction constant p’ has a negative value then means that the reaction

constant is increased by electron-donating protonation Actirration

(and concentration

groups, and thus that the influence on the

of the conjugate

acid) isdominant1*4.

paranzeters

The isokinetic

relationship

indicated

an isoenthalpic

reaction

series. Inspection

of the data in Table I indicates that this is true, and that the reason for the small increase in reaction velocity in the case of secondary alkyl groups is the entropy factor AS-“. The rate enhancement by these strong electron-donors is in accordance with the iinear free-energy reIationship. As was put forward by BeMiIIer and Doyle5, these electron-donating groups stabilise the conjugate acid because of their ability to disperse the positive charge on the oxygen atom. But, since there is less separation of Iike charges, the bond has to be stretched farther in the transition state, and thus there is a greater increase in entropy. Carbohyd.

Res., 25

(1972) 59-65

Aglycort group

2

5

$

Methyl Ethyl Propyl Butyl Pcntyl Hcxyl Heptyl Octyl 2.Methyl-l-propyl 4-Methyl-I-pcntyl 2-Methyl-I-pcntyl 2-Butyl 3-Pentyl 2-Pentyl 2-Ncxyl 2-Octyl Bcnzyl 3.Phcnylpropyl Cyclopcntyl cyclohcxyl

P-D-GALAcToI~YRANOSIDES

b

s

TABLE IV

2

.Fi

2

s 8

Ethanol Ethnnol Acetone Acetone Acetone Acetone Acetone Acetone Ethyl acctatc Acctonc Acctonc Acetone Acetone Acetone Acctonc Acctonc Ethyl acctatc Ethyl acctatc Acetone Acetone

Crystnllisofiora s01oc/rt

’

175-176 157-159 110-l 12 104-105 114-116 116-117 98-99 99-100 117-118 126-127 141-143 121-123 145-146 155-156 144-145 119-120 102-103 100-102 104-105 134-135

- 16.9 -20.5 -20.3 -20.6 - 18.4 - 17.5 - 17.1 -16.5 -20.6 -17.7 - 18.7 -21.1 -24.6 - 16.6 - 17.7 - 17.7 - 39.5 - 12.9 -27.1 -26.7 - 34.7 - 42.4 - 40.9 -41.1 - 38.0 - 37.0 - 34.2 - 33.3 -40.1 - 36.9 - 37.6 -43.9 -46.3 - 34.6 - 37.5 --36.5 -81.3 -24.9 - 55.3 - 54.6

b-P43GO (degrees)

43.2 46.0 48.7 50.8 52.7 54.5 56.1 57.5 50.8 54.7 54.6 50.8 52.8 52.9 54.4 57.2 51.7 60.0 53.0 54.9

C

Fowrd

(%)

C,0&006

608 7.4 881 803

ClZH2406 C14H2806

906

ClZH2206

CllH2006

C15H2206

C,JHlLtOd

CllHLZ06

CllH2206

clO~~2OoG

C,ZH2406

CIZH24O6

845 888 a,7 9,l

9,l 9,2

&4H2&6

8,5

C13H2606

Cd2406

CllH2206

C,O&006

C,H,& CBHI& W-II&

Forrrwln

9-o 9,4 9,5

7,o 706 8,l a,4 808

H

43.3 4602 48,7 50.9 5208 5446 56.1 57,5 50,9 5406 54.6 50.9 5208 52.8 54,G 57.5 57.8 60.4 53.2 55.0

c

Cfllc. (%)

7.2 7.7 8.1 8.5 8.8 9.1 9.4 9.6 8.5 9.1 9.1 8.5 8.8 8.8 9.1 9.6 6.7 7.4 8.1 8.4

H

HYDROLYSIS

TimelI

65

OF GALACTOPYRANOSIDES

Our results in this and previous’ work are in agreement with the results of and of BeMiller and Doyle’, and illustrate the fact that, within series of alkyl

P-D-glycopyranosides approximately

(equatorial

aglycon group),

the activation

parameters remain

constant, whereas this is not the case for alkyl or-D-glucopyranosides5

(axial aglycon group). EXPERIMENTAL

Alkyl

synthesis,

P-D-galactopyranosides

in the presence

were

of yellow mercuric

prepared

by

the

Koenig+Knorr’

oxide and mercuric

bromide.

‘*18

Benzene

(or chloroform) (200 ml), Sikkon (calcium sulfate Fluka) (40 g), yellow mercuric oxide (21.6 g; 0.1 mole), mercuric bromide (1.5 g), the alcohol (0.1 to 1 mole), and tetra-O-acetyl-a-D-galactopyranosyi bromideI were stirred in a closed vessel for several hours at room temperature (monitoring by t.1.c.). After filtration and thorough washing with water, the solution was evaporated in LWCUO to a syrup. Since none of the galactoside tetra-acetates crystallised, the crude mixture was deacetylated with sodium methoxide”, and the product was recrystallised from the appropriate solvent. Melting points, optical rotations (in methanol, c I), and purities of products were determined as described previously ’ 1_ Data for the alkyl fi-D-gaiactopyranosides are

presented in Table IV. The polarimetric carried out with a Perkin-Elmer Model

measurements of the reaction 141 photoelectric polarimeter,

velocity, and the

calculations of the first-order rate coefficients and of the activation parameters, were performed as described in previous ‘-’ publications. REFERENCES 1 2 3 4 5

C. K. DE BRUYNE AND F. VAX WIJXENDAELE, Carboltyd. Res., 6 (1968) 367. F. VAN WIJSESDAELE AND C. K. DE BRUYNE, Carbohyd. Res., 9 (1969) 277. C. K. DE BRUY~~IE AND J. WOUTERS-LEYSEN, Carbohyd. Res., 17 (1971) 45. T. E. TIPIIELL,Can. J. C/rem., 42 (1964) 1456. J. N. BEMILLER AND E. R. DOYLE, Carbohyd. Res., 20 (1971) 23. 6 J. N. BEMILLER, Advan. Carboityd. Chent., 25 (1970) 544. 7 I_. ZUCKER AND L. P. HALIMETT, J. Amer. CheJ?Z. Sot., 61 (1939) 2791. 8 F. A. LONG, J. G. PRITCHARD, AND F. E. STAFFORD, J. Amer. Citent. Sot., 79 (1957) 2362. 9 L. L. SCHALEGER AXD F. A. LONG, Advan. Phys. Org. Chem., 1 (1963) 1. 10 J. F. BUXNEIT, J. Anter. Chew. Sot., 83 (1961) 4956, 4968, 4973, 4978. 11 J. F. BUNNETT AND F. P. OLSEN, Can. J. Chent., 44 (1966) IS99, 1917. 12 0. EXXER, Collect. C=ecit. C/tern. Cottznturt., 3 1 (1966) 65. 13 0. EXNER, AT&me (Lcndott), 227 (1970) 366. 14 J. E. LEFFLER,J. Org. C/tent., 20 (1955) 1202. 15 J. R. ROBINSON AND L. E. MATHESOS, J. Org. C/tent., 34 (1969) 3630. 16 C. AR~IO~R, C. A. BIJ~TON, S. PATAI, L. H. SEL~?AN,AND C. A. VERNON, J. Cfiem. SOL, (1961) 412. 17 W. KOE~IGS AND E. KSORR, Ber., 34 (1901) 957. 18 L. R. SCHROEDERANJ J. W. GREEK, J. Cfzem.Soc., C,(I966) 530. 19 M. BARCZAI-IMARTOS AND F. K~~RBsY,;Xrafrcre (London), 165 (1950) 369. 20 A. THOMPSON AND M. L. WOLFROM, Methods CarboItyd. Cltent., 2 (1963) 215. 21 C.K. DE BRUYNEASD J.\OUTERS-LEYSEN, Carbohyd. Res.,4 (1967) 102. 22 M. A. PAUL AND F. -4. LONG, Cheat. Rev., 57 (1957) 1. Carbohyd. Res., 25

(1972) 59-65