Isotope Effects and Hydrogen Exchange Mechanism in Compounds Having Labile Aliphatic C-H Bonds in Basic Liquid Ammonia Solutions

Russian Journal of General Chemistry, Vol. 73, No. 3, 2003, pp. 358 370. Translated from Zhurnal Obshchei Khimii, Vol. 73, No. 3, 2003,

pp. 385 399.

Original Russian Text Copyright

2003 by Tupitsyn, Zatsepina.

Isotope Effects and Hydrogen Exchange Mechanism

in Compounds Having Labile Aliphatic C H Bonds in Basic

Liquid Ammonia Solutions

I. F. Tupitsyn and N. N. Zatsepina

Prikladnaya khimiya Russian Research Center, St. Petersburg, Russia

Received April 24, 2001

Abstract Specific features of the stepwise hydrogen exchange mechanism and transition state structure in

the systems acetophenone liquid ammonia in the absence of foreign compounds and in the presence of bases

and toluene liquid ammonia in the presence of potassium amide were studied in terms of an approach based

primary and secondary kinetic isotope effects of the substrate and the solvent. The mechanisms of reactions

involving acetophenone and toluene were compared. In the first case, an elementary act of CH-acid ionization

is contributed to a small extent by diffusion-controlled separation of the carbanion and ammonia molecule.

hydrogen exchange in toluene is characterized by complete absence of the internal ion pair return effect.

NH₃

The ratio k^N_D^H/k_Tfor hydrogen exchange in acetophenone tends to decrease on addition of bases (with

simultaneous increase in its rate), which may be explained by formation of an adduct via interaction between

the unshared electron pair on the heteroatom in the base molecule and the carbonyl carbon atom. The ano-

3

NH₃

malous temperature dependence of k^N_D^H/k_Tfor hydrogen exchange in toluene is interpreted as a result of

contribution of side metalation of the C H bond by potassium amide. The change in the solvent protophilicity

due to replacement of the light solvent by deuterated one differently affects the kinetics and mechanism of

hydrogen exchange in acetophenone and toluene. Measurements of the -deuterium effect gave information

on the mode of angular deformation of C D bonds in the methyl group of toluene in the hydrogen exchange

transition state.

3

We previously examined the effects of structural

and solvation factors on the equilibrium and kinetic

parameters of proton transfer in compounds charac-

terized by different labilities of aliphatic C H bonds

[1]. We made an attempt to study the mechanism of

hydrogen isotope exchange in some compounds in

basic liquid ammonia solutions. Apart from the results

obtained in [1], there are many published data on the

effect of structural factors, solvent properties, and

catalyst on the rate of H D exchange in compounds

belonging to different classes both in liquid ammonia

in the absence of other bases and especially in the

catalytic system KNH₂liquid NH₃[2 4]. Taken

together, these results provide a good basis for

drawing conclusions related to protolytic reactivity

of CH acids, such as those concerning the order of

variation of kinetic CH acidity in series of structurally

similar substrates, dependence of the rate constant

on the base protophilicity, degree of similarity of

transition states to reagents and products, etc. How-

ever, little attention was given so far to fine details of

the mechanism of hydrogen exchange between liquid

ammonia and compounds having labile aliphatic C H

bonds. Therefore, some specific features of hydrogen

exchange mechanism have not been interpreted unam-

biguously. In addition, we should note insufficient

substantiation of the existing concepts of interrelation

of the rate-determining elementary steps, poor data on

specific linear and angular structure of the transition

states, and scanty information on possible complica-

tions caused by solvent factors. The latter issue

originates from both incomplete utilization of the

available theoretical basis and shortage of experi-

mental data on kinetic isotope effects in hydrogen

exchange in liquid ammonia solution. In rare cases,

where the mechanism of such reactions was studied,

the specificity of intermediate steps was derived as

a rule from only primary kinetic isotope effect which

was measured at a single temperature (see, e.g., [5]).

The results thus obtained could not provide a com-

plete phenomenological pattern; they should be

regarded as preliminary and requiring further explora-

tion. In terms of the above stated, it should be noted

that the situation with mechanisms of protolytic reac-

tions in aqueous or alcoholic medium is quite dif-

ferent. Here, vast experimental data on primary and

secondary isotope effects of substrate and solvent

were collected and systematized, and criteria for

1070-3632/03/7303-0358$25.00 2003 MAIK Nauka/Interperiodica

ISOTOPE EFFECTS AND HYDROGEN EXCHANGE MECHANISM

359

determination of the mechanism were found on the

basis of isotope effect measurements. Joint study of

the above effects ensured quantitative description of

all specific features intrinsic to such processes.

Undoubtedly, the way of studying the reaction

mechanism tested on hydroxyl-containing media can

be applied to reactions occurring in liquid ammonia,

the more so the degrees of similarity and specificity

of the mechanisms of deuterium tritium exchange in

hydroxyl-containing and noaqueous basic polar media

(ammonia being a typical representative of the latter)

constitute an open problem.

ammonia which contained no compounds possessing

basic properties. In the second series, the initial con-

ditions were changed via variation of protophilic

properties of the medium.

Primary substrate isotope effect in hydrogen

isotope exchange between acetophenone and pure

liquid ammonia. It seemed important to determine

whether the rate of isotope exchange in the methyl

group of acetophenone is limited by the stage of C D

(C T) bond ionization (k₁) or it is an apparent

quantity depending on the preequilibrium intermediate

stage.

In keeping with the above, in the present work we

performed a detailed study of various aspects of

hydrogen isotope exchange of typical compounds

having aliphatic C H bonds with liquid ammonia

solutions. As subjects of the most detailed investiga-

tion we selected H D exchange in the methyl group

of acetophenone with pure liquid ammonia and its

solutions containing basic substances and H D

exchange in the methyl roup of toluene with a solu-

tion of potassium amide in liquid ammonia. These

media are polar and strongly basis; therefore, as a rule

they do not give rise to ion association. The selected

substrates are readily soluble therein even at low

temperatures. However, the above media are capable

of specifically interacting with various CH acids, so

that interpretation of the mechanism of hydrogen

exchange is not expected to be trivial. In some cases,

a simple stepwise mechanism may be inappropriate,

at least for some processes accompanying isotope

exchange could lead to formation of by-products.

There are data indicating side formation of an adduct

via reaction of ammonia (through unshared electron

pair on the nitrogen atom) at the electron-deficient

center of acetophenone [4, 6] and coordination of

potassium cation to CH acid (metalation of C H

bond in alkylaromatic hydrocarbons) in solutions of

potassium amide in liquid ammonia [7].

k ₁^L

k ^L

k₂

+

C L + NH₃

C , NH₃L

C H + NH₂L. (1)

1

Here, L = D or T; k^L₁and k^Lare the rate constants

of formation and decompositio¹n of intermediate con-

tact complex; and k₂is the rate constant of diffusion-

controlled separation of a solvent molecule from the

intermediate complex.

In the initial step of the study, we measured the

rate constants of protodedeuteration k^N_D^Hand protode-

3

tritiation k^N_T^Hin the methyl group of -D(T)-aceto-

3

phenone with liquid ammonia and determined the

ratio of these constants k^N_D^H/k_T. The latter varies

from 2.2 to 2.3 throughout the examined temperature

range (Table 1). Generally speaking, using the

NH₃

3

NH₃

k^N_D^H/k_T

ratios, it is impossible to quantitatively

3

estimate the contribution of each of the above stages

to the observed primary kinetic isotope effect of the

substrate for the whole reaction. The maximal in-

formation which can be derived from the measured

k^N_D^H/k_Tvalues is that is some limiting version of

the mechanism operative or not in the reaction under

study. In fact, according to scheme (1), the overall

NH₃

3

protodedeuteration rate constant k^N_D^Hor protodetritia-

3

tion rate constant k^N_T^His given by formula (2):

3

Our study of the reaction mechanism included

a number of techniques which ensured the maximum

possible utilization of different kinetic isotope effects,

analysis of the dependences of primary isotope effect

of the substrate (k_D/k_T) on the temperature and proton-

k^N_L^H= k₁^Lk₂/(k^L+ k₂).

(2)

3

1

Equation (2) is simplified in the two opposite cases

where the maximal or minimal primary isotope effect

of the substrate is directly related to the ratios of the

rate constants for elementary steps, k^L₁, k^L, and k₂. If

acceptor properties of added bases, and analysis of

NH₃

the behavior of the k^N_D^H/k_Tratio on variation of

3

k₂/k^L>> 1, according to formula (2)¹we obtain

CH acidity of the substrate in the presence of the

same base.

1

k^N_D^H/k_T; k₁^D/k^T₁; i.e., we have the isotope difference

between the rates of ionization, which reflects the

true isotope effect. By contrast, if the intermediate

contact complex is so strongly reactive that it abstracts

proton from the solvent molecule (which participates

in the protolytic process) at a higher rate than the rate

NH₃

3

MECHANISM OF HYDROGEN ISOTOPE

EXCHANGE IN THE METHYL GROUP

OF ACETOPHENONE

We performed two series of experiments. In the

first series, the reaction medium was pure liquid

RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 73 No. 3 2003

360

TUPITSYN, ZATSEPINA

Table 2. Average rate constants for deuterodeprotonation

Table 1. Average rate constants for protodedeuteration

(k^N_D^H) and protodetritiation (k_T) and k_D/k_Tratios

for hydrogen exchange in -D(T)-acetophenone with

liquid ammonia^a

k^N_H^Dand tritiodeprotonation k^N_T^Dand k_H^ND/k_T

ratios

NH₃

NH₃NH₃

ND₃

3

for hydrogen exchange in -H-acetophenone and -T-

acetophenone with liquid trideuteroammonia^a

k^N_D^H10⁵,^b

k^N_T^H10⁵,

Tempera-

ture,

k^N_H^D10⁵,

k^N_T^D10⁵,

3

Tempera-

ture,

NH₃

ND₃

k^N_D^H/k_T

k^N_H^D/k_T

3

1

C

s

C

s

0

15

25

1.0 0.1

2.3 0.2

5.8 0.3

0.45 0.05

1.0 0.1

2.6 0.2

2.3

2.2

0

15

25

8.0 0.7

17 1

40 2

1.4 0.1

12.1

a

Mean values from the results of 8 12 experiments.

Deuterium concentration in the initial ND 98% (of the

3

b

Tentative values of the Arrhenius parameters:

NH₃

E

=

theoretical value). Here, as well as in Tables 3, 4, and 6,

average values from 5 6 experiments (at each temperature)

are given.

9.2 kcal/mol, logA = 2.3. A rough estimate of k

at 40 C

D

7

1

is 2.5 10

s .

of transfer of the latter into bulk solution (k^L/k₂>> 1),

formula (2) is reduced to Eq. (3).

ammonia requires the following to be made. Initially,

it is necessary to identify on a qualitative level the

possibility for contribution of internal ion pair return

to the stage of C H bond ionization or to make

certain that such effect is absent. For this purpose,

the empirical exponent in the Swain Schaad formula

1

k_L= k^L₁/k^Lk₂= K_e^L_qk₂.

(3)

1

Taking into account that the equilibrium isotope

effect K^D_eq/K_e^T_qis negligible [8] and that the stage

described by the rate constant k₂is insensitive to

isotope substitution, kinetic isotope effect is not

was compared to the theoretical maximum y_max

=

3.34. The latter corresponds to the case when the stage

k^L₁in scheme (2) is fully rate-determining.

NH₃

observed at all: k^N_D^H/k_T= 1.

3

ND₃

NH₃

log (k^N_H^D/k_T) = y_maxlog (k^N_D^H/k_T).

(4)

3

Our experiments with a solution of -D(T)-aceto-

NH₃

phenone in liquid ammonia gave k^N_D^H/k_Tvalues

3

Substitution of the rate constants from Tables 1 and

2 (15 C) shows that the exponent y for the reaction

under study differs from the maximal value (y = 3.0).

Further on, using the experimental isotope effects

which correspond to neither theoretical maximum

NH₃

(k^N_D^H/k_T= 2.6 at 25 C) nor total absence of isotope

3

effect. Then, the primary substrate isotope effect is

a complex function of the parameter a_L= k^L/k₂which

1

NH₃

k^N_D^H/k_Tand k_H^NH/k_Tin combination with Eq. (4)

and assuming the absence od isotope effect at the

stage k₂, we can calculate the contribution of internal

ion pair return in the tritium and protium exchange

(a_Tand a_H) and the true isotope effect at the ioniza-

tion stage k^H₁/k₁^Taccording to Eqs. (5) (7).

3

characterizes the effect of internal ion pair return [3].

In other words, both elementary stages k and k₂

partialy limit the reaction rate, leading to the ¹observed

decrease of isotope effect relative to the maximal

value. Nevertheless, it is possible to estimate the

partial rate constants k^L₁and k^Land their isotope

1

ratios if the experimental k^N_D^Hand k_T^NHvalues are

3

y_max

supplemented by the corresponding values of deutero-

1

a_T(B^1/y

1)

= A + a_T(A

B);

(5)

deprotonation (k^N_H^D) and tritiodeprotonation rate con-

3

k^H₁/k^T₁= (k^N_H^D/k_T

)

ND₃

3

stants (k^N_T^D), determined independently. Such experi-

3

T

H

ND₃

1

ments were performed for solutions of -T-aceto-

phenone and light -H-acetophenone in liquid ND₃

(Table 2). The rate constants thus obtained were

treated by a univesal procedure which utilizes several

empirical formulas. The procedure makes it possible

to deduce the true kinetic isotope effect which reflects

isotope difference in the rates of dissociation of C D

(C T) bond at the stage k₁(k^D₁/k₁^T) and to estimate

on a quantitative level the effect of internal ion pair

{[1

a_T((K_eq/K_eq)(k^N_H^D/k_T

)

1)]} ;

(6)

(7)

3

a_H= a (K^T_eq/K_e^H_q)(k^H₁/k^T₁).

T

^NH₃y

3

Here,

A

)

=

(k^N_D^H/k_T

)

^max/(k^N_H^H³/k^N_T^H³);

B

=

^NH₃y

(k^N_D^H/k_T

^max/(K^T_eq/K^H_eq); K_e^T_q/K_e^H_qis the equilibrium

3

constant.

C T + NH₃(liq.)

C H + NH₂T(liq.)

return (k^L/k₂) [9]. This approach was developed by

According to the data of [10], K^T_eq/K_e^H_q= 1.1. The

calculation of a_T= k^T/k₂by Eq. (5) involves some

inconvenience, for Eq. ¹(5) cannot be solved in an ana-

1

Streitwieser and co-workers [9]. Its extension to

hydrogen exchange in acetophenone with liquid

RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 73 No. 3 2003

ISOTOPE EFFECTS AND HYDROGEN EXCHANGE MECHANISM

361

lytical form. Therefore, the solution was found by

iteration, and a_Tand a_Hwere estimated at 0.022 and

0.44, respectively (15 C). Thus, the noncatalytic

hydrogen exchange in acetophenone can be classed

with reactions characterized by a weak (a_T) but never-

theless perceptible (a_H) internal ion pair return. The

latter does not hamper attainment of a high isotope

effect at the stage of proton transfer from CH acid

to base: k^H₁/k₁^T= 18. The k₁^H/k^T₁value thus obtained

is very close to the theoretical value (k^H₁/k₁^T= 19 [11]).

In keeping with the simplest version of the transition

state theory, this result can be regarded as an evidence

for symmetric structure of the transition state in the

kinetically controlled C H ionization stage, which

is postulated by the three-center linear model.

(by 4.5 log unit) thermodynamic CH acidity of indene

in DMSO (pK_a20.1) than that of acetophenone

NH₃

(pK_a24.6). Likewise, the ratio k^N_D^H/k_Tequal to 2.2

at 25 C for hydrogen exchange in acetophenone is

smaller than the corresponding value for the more

3

NH₃

acidic fluorene: k^N_D^H/k_T= 1.9 at 25 C (pK_a= 22.4,

3

k^N_D^H= 16 10

s

[5]).

5

1

3

Thus, increase in CH acidity leads to a slightly

increased asymmetry of the transition state. Unfortu-

nately, the lack of kinetic data for a weaker CH acid

than acetophenone does not allow us to speak about

attainment of the theoretical maximum of isotope

effect in the reactions under comparison.

An additional information on the ratio of partial

rate constants k₁, k , and k₂can be derived from

1

Taking into account the structure-dependent nature

analysis of the temperature dependences of the isotope

effect. We believe that the number of measurements

and their accuracy are sufficient to differentiate

NH₃

of isotope effect, we presume that k^N_D^H/k_Tshould

decrease in going from acetophenone to other ana-

logous CH acids. Indeed, according to the modern

concepts, the dependence of k_D/k_Tupon CH acidity

for structurally similar substrates consists of two

branches, ascending and descending, with a maximum

3

NH₃

k^N_D^H/k_Tvalues obtained at 0 and 25 C. The data

3

in Table 1 indicate that k^N_D^H/k_Tdoes not tend to

NH₃

3

decrease in the temperature range from 0 to 25 C:

NH₃

k^N_D^H/k_T= 2.3, 2.3, and 2.2 at 0, 15, and 25 C,

3

at pK_a= 0, where pK_a= pK_a(substrate)

pK_a

respectively. This contradicts the commonly accepted

NH₃

view, according to which k^N_D^H/k_T= exp[h(

3

(autoprotolysis constant of the solvent); in a few

cases, the maximum is displaced toward positive or

negative pK_avalues [12]. The ascending branch of

the (k_D/k_T)_maxpK_acurve corresponds to early transi-

tion states which are structurally more similar to the

initial state, and the descending branch corresponds

to late transition states. A typical example in support

of the theoretical views is hydrogen exchange in

a series of thiazolium ions in aqueous medium [13].

In the series under consideration, the k_D/k_Tratio attains

the maximal value when the difference pK_afor two

acids, one of which is a CH acid participating in the

reaction and the other is an acid conjugate to a common

organic base, is equal to zero. This means that the

linear three-center model is valid [13]. An analogous

relation between k_D/k_Tand pK_awas observed for

exchange reactions of fluorene derivatives in a solu-

tion of potassium methoxide in methanol [14]. On the

basis of available experimental data it is impossible

to trace consistently substrate influence on the isotope

effect in hydrogen exchange between acetophenone

and liquid ammonia. Indene and fluorene are the only

hydrocarbons studied, which are structurally related to

acetophenone. Their relative inertness suggests that

the mechanism of H D exchange is not complicated.

As was shown in [5], the average rate constants and

CD

_CT)/kT] on the assumption that the corresponding

preexponential factors are similar (A_D= A_T). The

decrease in the isotope effect in the temperature range

0 25 C attains 10%, if k_D/k_Tis taken equal to 2.7.

Unfortunately, published data on the temperature

dependence of isotope effect in hydrogen exchange

between compounds having a labile aliphatic C H

bond and liquid ammonia are very scanty. The experi-

NH₃

mental temperature dependence of k^N_D^H/k_T

was

3

obtained only for indene [5]. According to [5], the

NH₃

ratio k^N_D^H/k_Tis 2.3 at 30 C and 2.0 at 0 C. In this

case, the temperature dependence of isotope effect

is similar to that predicted by the conventional theory

3

NH₃

for processes in which k^N_D^H/k_Tdecreases due to

variation of the degree of asymmetry of the three-

center transition state at the stage of C H bond

ionization (k₁) [15]. As applied to processes involving

internal ion pair return, the temperature dependence

3

NH₃

of k^N_D^H/k_Tshould be different, for the effect of

temperature on the constant k₂[see scheme (1)] should

also be taken into account. In this case, the tempera-

ture gradient of the experimentally measured

3

NH₃

k^N_D^H/k_Tratio can be divided arbitrarily into two

parts, one of which characterizes change of isotope

effect at the k₁stage, and the other, its variation due

to temperature contribution of internal ion pair return

(k₂). The latter contribution has the opposite sign,

and it acts to increase the overall istope effect as the

3

NH₃

k^N_D^H/k_Tratios for tritium exchange in indene are

3

5

1

NH₃

3

25 10

0.1 s and k^N_D^H/k_T= 2.0 at 0 C, i.e.,

a 25-fold increase in the reaction rate relative to aceto-

phenone corresponds to decrease of the isotope effect

from 2.3 to 2.0. This is consistent with the greater

RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 73 No. 3 2003

362

TUPITSYN, ZATSEPINA

temperature rises. This increase directly follows from

the close relation between the numerical value of k₂

and diffusion motion which is known to increase with

rise in temperature. Insofar as hydrogen exchange in

acetophenone involves two concurrent stages, k₁and

k₂, either zero temperature gradient of k^N_D^H/k_Tor

even its positive value could be expected. An ana-

logous pattern was observed in [13, 16]. Thus the

results of studying temperature dependence of the

kinetic isotope effect in hydrogen exchange in aceto-

phenone are consistent with the occurrence of internal

ion pair return in this reaction.

isotope effect of the solvent for reactions in water

and liquid ammonia [similar deuterium partition coef-

ficients in both liquid media and similar deuterium

isotope effects (which are reflected in the ionic prod-

ucts of the solvents)], we used Eq. (8) to estimate the

NH₃

3

Brønsted coefficient

for the reaction under study.

The value of was thus found to be 0.47 at 15 C.

This value can serve as an internal criterion of the

validity of the previous [1] conclusion on the structure

of transition state for hydrogen exchange in liquid

ammonia, which (as noted above) was derived from

direct comparison of logk^N_D^Hand pK_a. According to

3

[1], the Brønsted coefficient

is 0.74 at 25 C and

Secondary isotope effect of the solvent in hydro-

gen exchange between acetophenone and liquid

ammonia. Further study of the mechanism of hydro-

gen exchange in acetophenone in liquid ammonia

requires fine structure of the transition state to be

known. One of the ways to elucidate this problem

0.6 at 120 C. The disagreement between the values of

determined by different methods may be explained

by the fact that two different kinds of CH acids were

included in a single reaction series. First of all, this

applies to isomeric carboranes which give rise to hard

intermediate carbanions (i.e., the latter are charac-

terized by a high degree of electron density localiza-

tion). Therefore, in keeping with the Leffler Grun-

wald postulate, inclusion of carboranes should lead

to approaching of the transition state structure to that

of the final carbanion. Outwardly, this corresponds to

some overestimation of relative to the true Brønsted

coefficient for hydrogen exchange in the methyl group

of acetophenone. According to the latest concepts,

the transition state for hydrogen exchange in aceto-

phenone should approach completely symmetric one

with respect to both the magnitude of charge and its

distribution. On the whole, the Brønsted coefficient

is consistent with the conclusion drawn in the

previous section, i.e., that the transition state in the

reaction under study has a symmetric structure.

consists of estimation of the Brønsted coefficient

.

Our previous study [1] of the Brønsted relation for

hydrogen exchange in liquid ammonia was performed

with CH acids belonging to two structural classes:

(1) compounds possessing a labile aliphatic C H

bond and giving rise to delocalized carbanions and

(2) quasiaromatic compounds (isomeric carboranes).

Despite differences between these classes, the relation

between the Gibbs activation energy of hydrogen

exchange in liquid ammonia (logk_D) and pK_awas

found to be linear. The Brønsted coefficient

of

the corresponding equation was used to evaluate the

degree of structural and energetic similarity of the

transition state to products of proton transfer.

Although the Brønsted equation is helpful in explain-

ing the mechanism of hydrogen transfer, the results

obtained therewith are not always consistent with

those derived from the kinetic isotope effects.

Effect of bases on the kinetics and mechanism

of hydrogen exchange in acetophenone. In the next

step of our study we made an attempt to estimate the

effect of substances capable of acting as proton

acceptors on the kinetics and mechanism of hydrogen

exchange in acetophenone. The advantage of this

technique is that macroscopic parameters of the

system (primarily, its dielectric permittivity) remain

unchanged, for the concentration of the base added

is small. However, the degree of protophilicity of the

solution is expected to change in going from one base

to another, so that favorable conditions are created

for detalization of the role of bases in the mechanism

of the process. Preliminary experiments with various

bases showed that a positive catalytic effect is pro-

duced by addition of an organic base, piperidine,

poorly soluble sodium amide, and an inorganic com-

pound, calcium nitrate. Triethylamine was also among

the examined bases; its basicity is similar to the

basicity of piperidine. The rate constants were

An independent way of elucidating the structure of

transition state may be determination of on the basis

of the secondary isotope effect of the solvent, i.e., of

the ratio of rate constant for tritium exchange in the

same substrate with deuterated and and nondeuterated

solvent (k^N_T^D/k_T). A theory of secondary isotope

effects of a solvent has been developed [17], which

is valid for protolytic reactions in water or alcohol

[17]. The theory establishes relations between the

experimental isotope effect k^O_T^D/k_T^OH, on the one hand,

NH₃

3

and Brønsted coefficient

and maximum attainable

ratio (k^O_T^D/k^O_T^H

)

= 2.4, on the other.

max

k^O_T^D/k_T^OH= (k_T^OD/k_T^OH

)

= 2.4 .

(8)

max

Taking into account an almost complete identity of

factors which (in keeping with the theoretical con-

cepts) are responsible for the appearance of secondary

RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 73 No. 3 2003

ISOTOPE EFFECTS AND HYDROGEN EXCHANGE MECHANISM

363

NH₃

NH₃NH₃

Table 3. Average rate constants of deuterodeprotonation (k^N_D^H) and tritiodeprotonation (k_T) and k_D/k_Tratios for

3

hydrogen exchange in -D(T)-acetophenone with liquid ammonia in the presence of bases^a

Tempera-

ture,

k^N_D^H10⁵,

k^N_T^H10⁵,

3

NH₃

Added base

Piperidine

Concentration, M

k^N_D^H/k_T

3

1

C

s

0.11

0.22

0.58

0.67

(0.02)^a

3.9

15

15 1

22 1

24 1

6.7 0.6

2.2

15

40

0

30 2

Sodium amide

Calcium nitrate

220 15

3.6 0.2

11 2

157 12

2.1 0.1

7.9 0.5

17 2

1.4

1.7

1.4

1.1

3.9

15

25

20 2

a

Molar ratio sodium amide:acetophenone 1.4; molar ratio solvent:acetophenone 140 150.

NH₃

Table 4. Average rate constants of deuterodeprotonation (k^N_H^D) and protodedeuteration (k_D) of acetophenone, toluene,

3

and benzene with liquid ammonia containing basic substances

k^N_H^D10⁵,

k^N_D^H10⁵,

3

Tempera- Base concen-

NH₃

Catalytic system^a

k^N_H^D/k_D

3

1

ture,

C

tration, M

s

C₆H₅COCH₂L + NL₃(liq.)

15

17 2

16 2

14 1

5.5 0.5

85 10

69 5

2.3 0.1

15 1

7.4

1

S₆H₅COCH₂L + NL(liq.) + piperidine

C₆H₅COCH₂L + NL₃(liq.) + triethylamine

C₅H₅CONH₂L + NL₃+ Ca(NO₃)₂

C₆H₅CH₂L + NL₃+ KNL₂

15

35

0

0.11

0.09

0.18

3.9

0.05

11 2

5.6 0.7

1.1 0.1

7.7

12

13

C₆H₅L + L + KNL₂

14 1

a

L = H or D. Deuterium concentration in NL (L = D) 98% of the theoretical value.

3

measured at several concentrations of the base added.

The kinetic data are collected in Tables 3 and 4. The

following facts should be noted:

tends to slightly decrease as the concentration of tri-

ethylamine rises (Table 4);

(4) The effect of salts on the rate of hydrogen

exchange was studied using calcium nitrate. Nitrate

ions act as proton acceptors, though the acceleration

effect is weak: At an NO₃concentration of 3.9 M,

the rate constant of deuterium exchange increases by

a factor of 4 5 at 15 C and by a factor of 3 at 25 C.

(1) The data in Tables 1 and 3 indicate that the

rate of deuterium exchange in -D-acetophenone with

liquid ammonia at 40 C in the absence of bases is

lower by a factor of 10⁴than in the presence of

sodium amide (a suspension; c_NH= 0.02 M);

2

As in the H D exchange in pure liquid ammonia,

sufficient information on the reaction mechanism can

be obtained by simultaneously considering both

kinetic and thermodynamic aspects of the catalytic

reaction. Therefore, let us analyze variations of the

reaction rate, caused by addition of the above-listed

substances, in terms of homogeneous base catalysis.

For this purpose, the rate constants should be brought

into correlation with the equilibrium parameter of the

added bases. Although pK_avalues in water for all the

bases used are now available, we believe it un-

reasonable to treat them as basicity criteria in liquid

ammonia. Unlike water or alcohol, liquid ammonia

(2) Addition to a solution of acetophenone in

liquid ammonia of piperidine, which is a stronger base

than ammonia, induces a positive catalytic effect. The

latter increases as the concentration of piperidine rises

(general base catalysis: the rate constant increases by

a factor of 13 as the concentration of piperidine rises

from 0 to 0.67 M);

(3) A specific feature of triethylamine as base

additive to liquid ammonia is that the rate constant

k^N_D^Hfor deuterium exchange in acetophenone changes

3

insignificantly upon variation of triethylamine concen-

tration. Moreover, the rate of deuterium exchange

RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 73 No. 3 2003

364

TUPITSYN, ZATSEPINA

weakly solvates and stabilizes both the bases them-

selves (B) and their conjugate acids (BH⁺) via hydro-

gen bonding, for its ability to form such bonds is

insignificant (as compared to water). A more appro-

priate basicity parameter in liquid ammonia may be

the basicity in the gas phase (or proton affinity). The

proton affinities of the selected bases are greater than

that of ammonia (207 kcal/mol) but smaller than of

amide ion (419 kcal/mol). The data in Table 3 indicate

that increase in the reaction rate in going from pure

liquid ammonia to its solutions containing piperidine

and amide ion is consistent with the corresponding

proton affinity order in the gas phase (207, 232, and

419 kcal/mol, respectively [18]). The absence of

catalytic effect in the presence of triethylamine, whose

gas-phase proton affinity (236 kcal/mol [18]) appro-

aches that of piperidine, is not surprising. Presumably,

this is the result of steric shielding of the unshared

electron pair on the nitrogen atom by bulky ethyl

groups. An analogous pattern was observed previously

while studying the kinetics and thermodynamics of

proton transfer in the gase phase [19]. Meot-Ner and

Smith [19] found that there exist considerable dif-

ferences between unhindered and sterically hindered

basic centers: The rates of proton transfer in such

systems differ by more than 20 orders of magnitude.

supports the assumption that the overall reaction

mechanism is contributed to a greater extent by

deuteron transfer from the C D bond of unperturbed

acetophenone molecule to the base (path 1) rather than

by the process involving intermediate complex. By

contrast, initial formation of a solvate complex by

CH acid and basic reagent could slow the reaction

down (path 2). Both tendencies a priori make the

overall rate of H D exchange dependent on the con-

tributions of the two paths on addition of bases.

The above stated may be illustrated by the following

scheme.

Path 1

k₁

/2

+

/2

PhCOCH₂D + B

[PhCH₂

O

D

B

]

k

1

k₂

PhCOCH₃+ BD⁺,

(9)

Path 2

NH₃+ NO₃

NH₄NO₃+ NH₂,

(10)

/2

O

PhCH₂D,

PhCOCH₂D + NO₃

/2

O

NO₂

/2

O⁺

[PhCH₂

O

k₁

It is most difficult to interpret the disagreement

between the difference in the proton affinities of

piperidine (232 kcal/mol) and nitrate ion

(318 kcal/mol [20]) and appreciably smaller increase

in the rate of H D exchange under catalysis by the

latter. No plausible explanation may be given to the

observed anomaly on the basis of the experimental

data presented in Table 3. Some clarity can be intro-

duced by analysis of published data. There are both

direct and indirect indications that chemical reactions

of ketones, including acetophenone, with bases

involve intermediate formation of an adduct via attack

by unshared electron pair of the base on the carbonyl

carbon atom. The new bond is considered to be

covalent [6]. Alikhanov et al. [4] advanced a similar

hypothesis, according to which H D exchange of

acetophenone with liquid ammonia slows down due

to establishing of an equilibrium between the above

adduct and its decomposition products. Therefore,

we cannot rule out a mechanism where hydrogen

exchange is preceded by formation in solution of

a molecular complex with either ammonia or added

base. Obviuously, this process should affect in one

or another way the subsequent exchange reaction. In

other words, the formation of an adduct is equivalent

to complication of the reaction mechanism. The

existence of a correlation between the rate of H D

exchange and the gas-phase basicity of the reagent

+

D

NH₂]

PhCH₂D

+ NH₂

k

1

/2

O₃N

O

NO₂

k₂

(11)

PhCOCH₃+ NHD + NO₃.

It is readily seen that our attempt to correlate the

order of variation of the rate constants of hydrogen

exchange in acetophenone with gas-phase proton

affinities of the corresponding bases gives a reason-

able result. Path 1 is more significant for the series

of active bases such as ammonia, piperidine, and

sodium amide. On the other hand, catalysis of hydro-

gen exchange by calcium nitrate is favorable for

stronger (than in the other catalytic systems) inter-

action between the anion and the positively charged

carbonyl carbon atom. Addition of nitrate ion (through

unshared electron pair on its oxygen atom) to the

-carbon atom of acetophenone neutralizes the positive

charge on the latter, thus reducing proton mobility of

the methyl hydrogen atoms. Despite such inhibitory

effect, the reaction along path 2 provides a slightly

higher rate of H D exchange than the reaction rate

in the pure solvent, presumably as a result of genera-

tion of excess amide ions in solution. Thus, an appre-

ciable catalytic effect of nitrate ions may be explained

by a simultaneously occurring reversible reaction

between ammonia and nitrate ions. The assumption

RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 73 No. 3 2003

ISOTOPE EFFECTS AND HYDROGEN EXCHANGE MECHANISM

365

that an additional amount of NH₂ions is generated

by reaction (10) seems to be justified if we take into

account the following data. As shown in [21], addition

of potassium nitrate to a solution of some structurally

related CH acids (such as 3-methylpyridine and

2-methylnaphthalene) in liquid ammonia containing

potassium amide leads to increase in the intensity of

the long-wave carbanion band. Extra NH₂ions formed

by reaction (10) cause the equilibrium between the

molecular form of the substrate and its carbanion to

shift toward the latter.

states in the reactions under comparison are located at

different parts of the reaction path, each being charac-

terized by a high degree of asymmetry.

Thus, the assumed change of the reaction mecha-

nism, induced by addition of calcium nitrate, is

indirectly supported by analysis of the kinetic isotope

effects.

Effect of the protophilicity of solutions of aceto-

phenone in liquid ammonia on the kinetics and

mechanism of hydrogen exchange. The procedure

utilizing addition of basic reagents gives helpful

information on the mechanism of hydrogen exchange

in acetophenone. However, it remains unclear whether

a proton-abstracting species may be identified with

the added base or it is common for different catalytic

systems. For example, such a species may be amide

ion generated by reaction (10). In the further treat-

ment, we will try to substantiate an approach to

determination of the nature of protophilic species,

which implies examination of the differentiating effect

on the kinetic CH acidity of acetophenone in each

particular case in going from light liquid ammonia

to its deuterated analog having similar properties.

It is reasonable to presume that such a transition, as

well as addition of a base, will result in a minimal

change of polar properties of the medium (dielectric

constant and dipole moment), while the main con-

tribution to variation of the reaction rate will be that

arising from change in the solvent protophilicity.

The results of experiments carried out with both

acetophenone and hydrocarbon CH acids (benzene and

toluene) are given in Table 4. The data for benzene

and toluene were included in order to elucidate how

does the same base (amide ion) change the rate of

deuterium exchange in aliphatic and aromatic C H

bonds on replacement of liquid ammonia by its

deuterated analog. As follows from Table 4, the

examined bases exert different accelerating effects

on deuterium exchange in going from NH₃to ND₃,

other conditions being equal. The strongest accelara-

tion is observed for potassium amide: it gives rise to

the maximal difference (by a factor of 11 12) between

the rate constants of deuterium exchange in each of

the above listed hydrocarbons in going from the

system NH₃(liq.) + KNH₂to ND₃(liq.) + KND₂.

Nitrate ion and ammonia occupy an intermediate place

(the reaction with acetophenone is accelerated by

a factor of 7.5). Piperidine turned out to be the least

active (no acceleration was observed). These results

can be explained on the assumption that the thermo-

dynamic basicity of ND₂ion in deuterated liquid

ammonia is greater than the basicity of NH₂ion in

liquid NH₃, like OD ion in D₂O is a stronger base

(by a factor of 1.47) [22, 23] than OH ion in H₂O.

Changes of the reaction path and transition state

structure, postulated above as a result of addition of

basic compounds, could lead to variation of primary

NH₃

substrate isotope effect k^N_D^H/k_Twhich is an experi-

3

mentally measured parameter of the transition state.

The experimental results are given in Table 3. The

isotope effect in hydrogen exchange between -D(T)-

acetophenone and liquid ammonia in the presence of

piperidine remains almost the same as in the absence

of catalyst (Table 1). In experiments on reverse

exchange in -D(T)-acetophenone with addition of

NH₃

sodium amide, the ratio k^N_D^H/k_Tsharply decreases

3

(1.4 at 40 C).

A specific feature of the catalytic system containing

Ca(NO₃)₂(3.9 M) is sharply reduced primary sub-

NH₃

strate isotope effect (k^N_D^H/k_T= 1.1 at 25 C and

1.4 at 15 C). According to the data in Tables 1 and 3,

hydrogen exchange in -D(T)-acetophenone with pure

liquid ammonia is characterized by the maximal

isotope effect. It decreases as the basicity of the

reagent increases: NH₃piperidine >> sodium amide.

Unfortunately, the difference in the isotope effects

in going from ammonia to sodium amide can be

estimated only roughly, for kinetic experiments with

different bases cannot be performed at the same tem-

perature. Nevertheless, taking into account antiparallel

3

NH₃

variations of the isotope effect k^N_D^H/k_T

and the

3

basicity of the catalyst for the same CH acid (aceto-

phenone), we can state that the position of the transi-

tion states on the reaction coordinate corresponds to

the descending part of the curve (except for the reac-

tion in the presence of nitrate ions). Upon addition

of nitrate ions, a small increase in k_Dis accompanied

NH₃

by sharp decrease of k^N_D^H/k_T, which contradicts the

Leffler Grunwald postulate. Presumably, the forma-

tion of a solvate complex with equivalent C O bonds

leads to disappearance of positive charge and appear-

ance of a negative charge on the carbonyl carbon

atom. As a result, the transition state moves from

the product-like region (descending branch of the

3

NH₃

k^N_D^H/k_TpK_acurve) through the maximum of the

3

NH₃

k^N_D^H/k_TpK_adependence to the reagent-like region

3

(ascending branch). In other words, the transition

RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 73 No. 3 2003

366

TUPITSYN, ZATSEPINA

It should be emphasized that there is a close

amide in liquid ammonia were studied in [5], and the

kinetic isotope effect at 0 C was measured. However,

the contribution of internal ion pair return to the

rate of hydrogen exchange was not estimated quanti-

tatively (see above). More accurate results can be

obtained by performing experiments on joint deter-

mination of both the rate constants of protodedeutera-

analogy between the pattern of variation of deuterium

exchange rates in going from H₂O to D₂O and the

observed more pronounced catalytic effect of amide

ions on the rate of hydrogen exchange in the above

hydrocarbons in liquid deuterated ammonia compared

to NH₃. Namely, the rate constant of ionization of

deuteroacetone in water (CD₃COCD₃+ H₂O + OH )

is lower by a factor of 11 than the rate of the reverse

reaction of nondeuterated acetone in heavy water

containing alkali [22].

tion k^N_D^H

)

and protodetritiation (k_T^NH

)

(reverse

3

exchange) and the rate constants of deuterodeprotona-

ND₃

tion (k^N_H^D) and tritiodeprotonation (k_T) (direct

exchange where the solvent acts as a donor of heavy

isotope). In order to determine the whole set of rate

constants related to different temperatures, as well as

the ratios of these constants), we used various H/D

and H/T isotopomers of toluene and ammonia. The

conditions and results of experiments are collected in

Tables 5 and 6. Some comments should be given to

3

The considerable acceleration of the reaction by the

action of ND₂or OD ions in going from non-

deuterated solvent to deuterated one and the absence

of acceleration in the reaction catalyzed by piperidine

lead us to conclude that the accelerating effect is

a specific property of ions (ND₂, OD ). As concerns

the effect of organic bases on the rate of deuterium

exchange in going from a light solvent to heavy, their

contribution to the overall acceleration is a matter

of discussion. In terms of the above stated, the inter-

mediate degree of acceleration of hydrogen exchange

by Ca(NO₃)₂in liquid ammonia should be treated as

the result of superposition of two reaction paths (1

and 2), when protophilic species are simultaneously

the neutral base and NH₂ion.

the value of k^N_T^Dpresented in Table 6. Because of

3

difficulties in the synthesis of the initial C₆H₅CD₂T

isotopomer we failed to perform kinetic measurements

in the system C₆H₅CD₂T + ND₂+ ND₃(liq.). Instead,

an approximate value of the corresponding rate con-

stant was determined for the system C₆H₅CH₂T +

ND₂+ ND₃(liq.). Here, the process involving decrease

of the tritium concentration in the initial toluene is

accompanied by deuterium transfer from the solvent

to toluene molecule. The resulting secondary isotope

We thus conclude that the difference in the rates of

hydrogen exchange in different CH acids on variation

of isotope composition of the solvent does not

originate from change of the reaction mechanism but

reflect specific features of thermodynamic interaction

between CH acids and basic reagents.

effect of the substrate somewhat distorts the k_T^ND

3

value corresponding to tritium exchange in the system

C₆H₅CD₂T + ND₂+ ND₃(liq.). As noted above, the

NH₃

ND₃

set of k^N_D^H/k_Tand k^N_H^D/k_Tvalues is required for

3

calculations by Eq. (4) which gives a rough estimate

of the deviation of y from the maximal value y_max

=

3.34, when the k₁stage of scheme (1) (which is sen-

MECHANISM OF HYDROGEN ISOTOPE

EXCHANGE IN THE METHYL GROUP

OF TOLUENE

sitive to isotope replacement) is fully rate-controlling.

NH₃

ND₃

Substituting the values of k^N_D^H, k_T, k_H, and k_T

for hydrogen exchange in toluene into formula (4)

gives y = 3.3 ( 25 C). Taking into account errors in

the determination of rate constant, as well as the

approximate character of Eq. (4), the difference

between the experimental exponent y and the theore-

tical value (y_max) is negligible. This indicates that

internal ion pair return does not affect the reaction

under study. The absence of a diffusion limit of the

reaction rate (a_L= 0) makes it possible to estimate

the true kinetic isotope effect for hydrogen exchange

in toluene using formula (6): k^H₁/k₁^T= 21. This value

approaches the theoretical maximum, indicating

a symmetric structure of the transition state.

3

Primary substrate isotope effect in hydrogen

exchange in toluene with a solution of potassium

amide in liquid ammonia. This section describes the

relation between the structure of CH acid, specifically

the nearest environment of the labile methyl group

(with acetophenone and toluene as examples), and fine

details of the hydrogen exchange mechanism in liquid

ammonia solution.

Toluene is a fairly weak CH acid (pK_a43), which

determines its kinetic behavior. Hydrogen isotope

exchange in the methyl group of toluene occurs at

a kinetically measurable rate at reduced temperature

and only in the presence of a very strong proton

acceptor, amide ion [5]. The kinetics of isotope

exchange in toluene containing deuterium and tritium

in the methyl group with a solution of potassium

Using the k^H₁/k₁^Tvalues for the two CH acids,

toluene and acetophenone, we can consider the effect

of pK_aof CH acids on the mechanism of hydrogen

exchange. It is seen that, contrary to expectations,

change of pK_aby 17 units in going from one CH acid

RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 73 No. 3 2003

ISOTOPE EFFECTS AND HYDROGEN EXCHANGE MECHANISM

367

Table 5. Average rate constants of protodedeuteration

Table 6. Average rate constants of deuterodeprotonation

NH₃

NH₃NH₃

ND₃

ND₃ND₃

(k^N_D^H) and protodetritiation (k_T) and k_D/k_Tratios

for hydrogen exchange between -D(T)-toluene and

a solution of potassium amide in liquid ammonia^a

(k^N_H^D) and tritiodeprotonation (k_T) and k_H/k_Tratios

for hydrogen exchange in light toluene and -T-toluene

in a solution of potassium amide in liquid ammonia^a

3

Tempera- k^N_D^H10⁵,^b

k^N_T^H10⁵,

Tempera-

ture,

k^N_H^D10⁵,^b

k^N_T^D10⁵,^b

3

NH₃c

ND₃

k^N_D^H/k_T

k^N_H^D/k_T

3

1

ture,

C

s

C

s

25

12

0

7.8 0.4

18 2

61 3

3.1 0.2

6.4 0.3

23 2

2.5

2.8

2.7

45

18 1

69 3

35

25

110 10

5.0 0.2

22

a

c

b

Concentration of potassium amide 0.05 M. Tentative param-

eters of the Arrhenius equation: E = 8.0 kcal/mol, logA = 3;

the value of k

Average values from 8 12 runs at each temperature.

a

Concentration of potassium amide 0.05 M.

Deuterium concentration in the initial deuterated ammonia

98% of the theoretical value.

NH₃

5

1

b

derived therefrom ( 35 C) is 5 10

s

.

D

to the other is not accompanied by the corresponding

change of the transition state structure. The observed

insensitivity of k₁^H/k^T₁to CH acidity may be under-

stood on the assumption that hydrogen exchange in

acetophenone and toluene involves protophilic species

of different natures, namely these may be ammonia

molecule and amide ion, respectively. Presumably,

the high degree of symmetry of the transition states

in both reactions reflects similarity in electron density

distribution over the three-center transition bridge

2.6 0.1 and 2.9 0.1 at 25 and 12 C, respectively.

Comparison of the data for the pair toluene 2-methyl-

NH₃

pyridine shows a very weak tendency for k^N_D^H/k_T

to vary as the CH acidity of the substrate rises (unless

the lack of such tendency). Presumably, in our case

3

NH₃

the maximum on the k^N_D^H/k_TpK_acurve is so fuzzy

that no conclusion on the degree of symmetry of the

transition state can be drawn on the basis of this de-

pendence [25]. Streitwieser et al. [26] also noted the

absence of sensitivity of k_D/k_Tto pK_ain going from

toluene (pK_a= 43) to triphenylmethane (pK_a= 30.6)

while studying hydrogen exchange in a solution of li-

thium cyclohexylamide in cyclohexylamine (k_D/k_T=

2.8 0.15 at 25 C). Also, the insensitivity of isotope

effect in hydrogen exchange in the same solution to

the CH acidity of toluene and some its derivatives hav-

ing substituents in the benzene ring was reported [27].

3

/2

+

/2

C

H

N

formed, on the one hand, by

a medium-strength CH acid (acetophenone) and a rela-

tively weak base (ammonia) and, on the other, by

a very weak CH acid (toluene) and a powerful proton

acceptor (NH₂ion).

We have no possibilities for applying the set of

the above described techniques to analogous detailed

study of the mechanism of hydrogen exchange in

alike substrates differing from toluene by a smaller

number of pK_aunits. Therefore, in order to estimate

the sensitivity of isotope effect in the reaction in

a solution of potassium amide in liquid ammonia we

confined ourselves to measurement of the rate con-

Analysis of the mechanism of hydrogen exchange

in toluene is not limited to comparison of the data

on structural effects on the kinetic isotope effect

with those found for acetophenone and other related

CH acids. We also examined the temperature depend-

NH₃

ence of k^N_D^H/k_Tfor hydrogen exchange in toluene.

3

NH₃

stants k^N_D^Hand k^N_T^Hand their ratios k_D^NH/k_Tfor the

reaction of 2-methylpyridine with a solution of potas-

sium amide in liquid ammonia. 2-Methylpyridine is

a stronger CH acid than toluene [pK_a(THF) = 34]

[24], and it more readily reacts with liquid ammonia

in the presence of potassium amide. The values of

3

According to the data in Table 5, in the temperature

NH₃

range from 25 to 0 C, the ratio k^N_D^H/k_Tslightly

3

increases as the temperature rises: The values of

NH₃

k^N_D^H/k_Tare 2.5, 2.8, and 2.7 at 25, 12, and 0 C,

respectively. Measurements of the isotope effects were

repeated for 8 12 times at each temperature; the dif-

3

k^N_D^H, k_T, and k_D/k_Tfor hydrogen exchange in

NH₃

NH₃NH₃

3

NH₃

ferences in the resulting k^N_D^H/k_Tvalues were so

large that they should be regarded as real. An ana-

logous weak sensitivity of the isotope effect to

temperature was observed for hydrogen exchange in

acetophenone. Using the temperature dependence of

3

-D(T)-2-methylpyridine with potassium amide

liquid ammonia (c_KNH= 0.05 M) at 35 C are as

2

5

1

follows: 1.6 10 , 0.56 10 s , and 2.85. The ratio

k^N_D^H/k_Talmost does not differ from the correspond-

ing value for toluene, especially if we take into

account that hydrogen exhange in 2-methylpyridine

NH₃

3

NH₃

k^N_D^H/k_T

we have substantiated the influence of

3

internal ion pair return on the process. It was assumed

that the normal course of the process implies gradual

was examined at a higher temperature: k^N_D^H/k_T

=

NH₃

3

RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 73 No. 3 2003

368

TUPITSYN, ZATSEPINA

Table 7. Conditions and results of protodedeuteration of -D₃-toluene in liquid ammonia in the presence of NH₂ions

(c_KNH= 0.05 M)

2

Relative intensity of ion peaks in the mass spectrum,^a

%

k^N_D^H10⁵,

3

Run

no.

Temperature,

C

Time, h

1

s

m/z = 92

m/z = 93

m/z = 94

m/z = 95

1

2

3

4

25

15

1.5

5

8

51.0

8.1

19.1

19.3

3.3

29.4

29.7

37.8

7.9

39.1

37.5

30.5

37.8

23.4

13.7

12.4

2.6

2.8

2.4

7.3

3

a

Initial intensity, I , %: run no. 1: m/z = 92, 49.5; m/z = 93, 0.5; m/z = 94, 0.0; m/z = 95, 50.5; run nos. 2 4: m/z = 92, 0.0; m/z =

rel

93, 1.6; m/z = 94, 1.8; m/z = 95, 96.6.

NH₃

decrease of k^N_D^H/k_Tas the temperature rises. There

are published data on other examples of a nonclassical

temperature dependence of isotope effect in protolytic

reactions; here, variation of the isotope effect is

assigned to variable temperature contribution of

transition state for hydrogen exchange in toluene is

characterized by two structural parameters. One of

these is linear; it is defined as the drift along the

reaction coordinate relative to the position of proton

in the symmetric model which corresponds to the

maximal isotope effect. The second parameter is

angular; it is related to rehybridization of the methyl

carbon atom during the reaction from sp³to near-sp².

The above results make it possible to estimate on

a qualitative level only the first parameter. To get

a deeper insight into the mechanism of hydrogen

exchange we should consider angular deformations of

the C D (C H) bonds at the carbanionic center in

the transition state. For this purpose, we determined

the secondary -deuterium substrate isotope effect by

comparing the rates of deuterium exchange in -D-

toluene and -D₃-toluene with a solution of potassium

amide under similar conditions. It should be kept in

mind that the proposed approach is applicable only

when the relative degree of exchange is low, i.e.,

when the overall rate constant of the reaction in the

methyl group of -D₃-toluene can be equated to k_III.

3

internal ion-pair return to the ratio k^N_D^H/k_T[13, 16].

NH₃

3

The results given in Table 5 do not allow us to

interpret the observed temperature anomaly in terms

of the same approach, for our analysis of the relations

between elementary steps of hydrogen exchange in

toluene indicates the absence of internal ion pair

return effect. We believe that the anomalous tempera-

ture dependence of isotope effect in the system under

study should be treated as originating from an addi-

tional thermodynamic isotope effect intrinsic to

metalation of the methyl C H bond with potassium

amide. This effect is concomitant to the kinetic

isotope effect on proton transfer from CH acid to

base. It is still impossible to separate the contributions

of thermodynamic and kinetic factors to the observed

k^N_D^H/k_Tvalue. Nevertheless, it seems to be improb-

able that the isotope effects would be similar for the

two processes and that they would show similar tem-

perature dependences. This assumption is supported

by the fact that some structuraly related compouds,

which are not considerably stronger CH acids than

toluene, react with potassium amide in liquid

ammonia to give alkaline organic derivatives. The

latter are detected by the appearance of carbanion

bands in the electronic spectra and IR spectra of solu-

tions of potassium amide in liquid ammonia contain-

ing compounds with both activated (2-methylpyridine,

fluorene, etc.), and nonactivated methyl or methylene

NH₃

3

k

II

III

C₆H₅CD₃

C₆H₅CD₂H

C₆H₅CH₃.

C₆H₅CDH₂

k

I

(12)

At a moderate degree of exchange, the rate constant

measured experimentally (Table 7) is an apparent

quantity which characterize at least two steps of the

complex reaction (k_IIIand k_II). This means that the

experimental rate constant ratio should be regarded

only as a lower limit of the -deuterium effect given

by the ratio k_D(C₆H₅CH₂D)/k_D(C₆H₅CD₃). According

group

(1-methylnaphthalene,

diphenylmethane,

3-methylpyridine) [21]. The reason for the anomalous

temperature dependence of primary isotope effect in

hydrogen exchange in toluene requires further study.

to the data in Tables

5

and 7, the ratio

k_D(C₆H₅CH₂D)/k_D^NH(C₆H₅CD₃) is equal to 3 at 25 C.

Using the approximate formula k_I/k_II= k_II/k_III, we can

readily calculate the intermediate rate constant k_II

k_D(C₆H₅CD₂H) = 4.5 10 ⁵s ¹and hence the -deute-

3

Utilization of -deuterium isotope effect for

studying the transition state structure. A model of

RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 73 No. 3 2003

ISOTOPE EFFECTS AND HYDROGEN EXCHANGE MECHANISM

369

rium effect due to replacement of two H atoms by D.

Such unit -deuterium effect k_I/k_II

inductive contribution remains unchanged regardless

of the nature of the medium. Therefore, hydrogen

exchange in the methyl group of toluene with a solu-

tion of potassium amide in liquid ammonia is charac-

terized by increased overall -deuterium isotope effect

as compared with the analogous reaction in cyclo-

hexylamine.

a

=

k(C₆H₅CH₂D)/k(C₆H₅CD₂H) is equal to 1.7. The

experimental value of k(C₆H₅CH₂D)/k(C₆H₅CD₂H) is

anomalously high. In keeping with the theoretical

views, according to which the main contribution to

-deuterium effect is the difference in zero-point

vibrations corresponding to deformation of carbon

hydrogen bonds with respective isotopes in the initial

molecules and transition states, the maximal unit

-deuterium effect in going from the sp³-hybridized

the methyl carbon atom in toluene to the sp²-hybridized

carbon atom in the benzyl carbanion should approach

a value of 1.4 [28]. According to our data, the

-deuterium effect caused by replacement of one

H atom by D (except for the atom subjected to isotope

substitution) exceeds not only the theoretical limit

but also the experimental value 1.3 obtained for

hydrogen exchange in toluene with a solution of

lithium cyclohexylamide in cyclohexylamine [29].

In order to interpret the above disagreements, we

presume simultaneous action of two different factors

(vibrational and inductive) which are responsible for

the value of -deuterium effect. The inductive factor

arising from a slightly stronger electron-donor effect

of deuterium as compared to protium upon abstraction

from toluene molecule does not depend on the mode

of coordination of the anionic part of the transition

state to the counterion. Unlike the inductive factor,

the influence of the vibrational factor on isotope effect

strongly depends on the polarity of the medium.

Presumably, in a weakly polar solvent such as cyclo-

hexylamine ( = 4.7), the carbanionic transition state

occurs as an ion pair with lithium cation. The electro-

static interaction therein is sufficiently strong to fix

sp³-like bond configuration at the anionic carbon atom

[29]. Thus, the low value of -deuterium effect

relative to the theoretical limit corresponding to the

vibrational approximation is explained by a combina-

tion of the two factors: (1) The degree of orbital re-

hybridization in the transition state is far from the

limiting one (which corresponds to its planar struc-

ture) and (2) the inductive effect of deuterium is

weak. By contrast, the role of counterion in the

formation of transition state for deuterium exchange

in a strongly polar solution of alkali metal amide in

liquid ammonia is likely to be insignificant. The

absence of interionic coordination favors transition to

a state characterized by a greater s-order of hybridized

orbitals than in the reaction in cyclohexylamine. This

is due to greater delocalization of p electron on the

anionic carbon atom via conjugation with the aromatic

-electron system. Such structural variation of the

transition state leads to increased contribution of the

vibrational factor to -deuterium effect, while the

EXPERIMENTAL

-D- and -D₃-Toluene isotopomers and deuterated

liquid ammonia were obtained from the pilot plant at

the Prikladnaya khimiya Research Center. Samples of

-D,T-toluene were prepared by decomposition of

benzylmagnesium bromide with water containing both

deuterium (95 at %) and tritium. -D,T-Acetophenone

was synthesized by exchange of commercial aceto-

phenone (preliminarily distilled under reduced pres-

sure) with heavy water containing an indicator amount

of HTO under catalysis by KOD (12 h, 100 C, c_KOD

=

0.2 M). The procedures for preparation of KNH₂solu-

tions in liquid ammonia and kinetic measurements

were described in [1]. The purity of all the compounds

was checked by gas chromatography. In all experi-

ments, the amount of solvent was 100 150 mol per

mole of acetophenone; in experiments with toluene,

the amount of solvent was 70 mol or more per mole

of toluene. The rate constants did not change on varia-

tion of the solvent amount by a factor of 1.5 2. The

position of deuterium in samples of acetophenone

obtained in experiments with ND₃was determined

from increase in the intensity of vibration bands of

1

aliphatic C D bonds in the IR spectra (2200 cm ).

Mass spectrometric analysis of benzoic acid prepared

by oxidation of toluene with potassium permanganate

showed the absence of deuterium in the aromatic ring

of toluene samples obtained in direct-exchange

experiments. The concentration of deuterium in the

methyl groups of toluene and acetophenone was deter-

mined by low-voltage mass spectrometry using

an MI-1301 instrument. The concentration of tritium

in water obtained by combustion of substances before

and after exchange was determined using a scintilla-

tion counter (in a solution of liquid scintillator). The

rate constants of hydrogen isotope exchange were

calculated by the first-order equation using average

values from 5 6 runs. In the determination of the

NH₃

temperature dependence of k^N_D^H/k_T8 12 parallel

3

experiments were run.

REFERENCES

1. Tupitsyn, I.F. and Zatsepina, N.N., Russ. J. Gen.

Chem., 2002, vol. 72, no. 3, p. 405.

2. Shatenstein, A.J., Adv. Phys. Org. Chem., 1965,

vol. 1, p. 155.

RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 73 No. 3 2003

370

TUPITSYN, ZATSEPINA

3. Reutov, O.A., Beletskaya, I.P., and Butin, K.P.,

Izotopnye effekty v skorostyakh reaktsii, Moscow:

Mir, 1964, p. 188.

CH-Kisloty (CH Acids), Moscow: Nauka, 1980,

p. 247.

16. Thibben, A., Onyido, I., and Ahlberg, P., Chem. Scr.,

1982, vol. 119, no. 3, p. 145.

17. Gold, V. and Grist, S., J. Chem. Soc., Perkin

Trans. 2, 1972, no. 1, p. 89.

4. Alikhanov, P.P., Yakovleva, E.A., Isaeva, G.G., and

Shatenshtein, A.I., Zh. Obshch. Khim., 1969, vol. 39,

no. 7, p. 1433.

18. Kabachnik, M.I., Usp. Khim., 1979, vol. 48, no. 9,

p. 1523.

5. Shatenshtein, A.I., Yakushin, F.S., and Arshino-

va, M.I., Kinet. Katal., 1964, vol. 5, no. 6, p. 1000.

19. Meot-Ner, M. and Smith, S.C., J. Am. Chem. Soc.,

1991, vol. 113, no. 3, p. 862.

6. Reeves, R.L., The Chemistry of the Carbonyl Group,

Patai, S., Ed., London: Wiley, 1966, p. 354.

20. Dewar, M.J.S. and Dieter, K., J. Am. Chem. Soc.,

7. Kane, A.A. and Tupitsyn, I.F., Raboty po termodi-

namike i kinetike khimicheskikh protsessov (Studies

on Thermodynamics and Kinetics of Chemical

Processes), Leningrad: Khimiya, 1970, no. 66,

pp. 132 138.

1986, vol. 108, no. 25, p. 8075.

21. Kane, A.A. and Tupitsyn, I.F., Raboty po termo-

dinamike i kinetike khimicheskikh protsessov (Studies

on Thermodynamics and Kinetics of Chemical

Processes), Leningrad: Khimiya, 1970, no. 66,

pp. 107 113.

8. Bell, R.P., The Proton in Chemistry, London:

Chapman and Hall, 1973. Translated under the title

Proton v khimii, Moscow: Mir, 1977, p. 381.

22. Pocker, Y., Chem. Ind., 1959, no. 44, p. 1383.

9. Boerth, D.W. and Streitwieser, A., J. Am. Chem.

Soc., 1981, vol. 103, no. 21, p. 6443.

23. Swain, S., Stivers, E., Reuwer, J.H., and Schaad, L.J.,

J. Am. Chem. Soc., 1958, vol. 80, no. 21, p. 5885.

10. Tupitsyn, I.F., Semenova, N.K., and Avdulov, G.I.,

Radioisotopes in the Physical Sciences and Industry,

Vienna: Int. Atomic Energy Agency, 1962, pp. 255

257.

24. Fraser, R.R. and Mansur, T.S., J. Org. Chem., 1984,

vol. 49, no. 18, p. 3443.

25. Tornton, H.R., J. Org. Chem., 1962, vol. 27, no. 4,

p. 1943.

11. Melander, L. and Saunders W.H., Reaction Rates of

Isotopic Molecules, New York: Wiley, 1980, p. 245.

26. Streitwieser, A., Owens, P.H., Sonninchsen, G.,

Smith, W.K., and Niemeyer, H.M., J. Am. Chem.

Soc., 1973, vol. 95, no. 13, p. 4254.

12. Suhnel, J., Isotopen Praxis, 1986, vol. 22, no. 1,

p. 73.

27. Streitwieser, A. and Marls, F., J. Am. Chem. Soc.,

13. Washbaugh, M.N. and Jencks, W.P., J. Am. Chem.

Soc., 1989, vol. 111, no. 2, p. 683.

1967, vol. 89, no. 15, p. 3767.

28. Streitwieser, A., Jagow, R.H., Fahey, R.C., and

Suzuki, S., J. Am. Chem. Soc., 1958, vol. 80, no. 9,

p. 2326.

14. Cram, D.J. and Kollmeyer, W.B., J. Am. Chem. Soc.,

1968, vol. 90, no. 5, p. 1791.

15. Melander, L.C.S., Isotope Effects on Reaction Rates,

29. Streitwieser, A. and van Sickle, D.E., J. Am. Chem.

Soc., 1962, vol. 84, no. 2, p. 254.

New York: Ronald, 1960. Translated under the title

RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 73 No. 3 2003

Article Doi

Isotope Effects and Hydrogen Exchange Mechanism in Compounds Having Labile Aliphatic C-H Bonds in Basic Liquid Ammonia Solutions

DOI: 10.1023/A:1024997601298

Source and publish data:

Authors:

Article abstract of DOI:10.1023/A:1024997601298

Full text of DOI:10.1023/A:1024997601298

Products guided by the article

Relevant to this article

Hot Product