Selective mono-C-methylations of arylacetonitriles and arylacetates with dimethylcarbonate: A mechanistic investigation

Article abstract of DOI:10.1021/jo0057699

The very high mono-C-methylation selectivity (>99%) of arylacetic acid derivatives (ArCH₂X; X = CN, CO₂Me) with dimethyl carbonate (DMC) is due to a mechanism that involves consecutive methoxycarbonylation, methylation, and demethoxycarbonylation steps. Important aspects of this mechanism are clarified herein by a kinetic investigation. In the case of arylacetonitriles, at 140°C, the comparison of the rate constants of model reactions involving 2-phenyl propionitrile, phenylacetonitrile, 2-methoxycarbonylphenylacetonitrile, and 2-methyl-2-methoxy carbonylphenyl acetonitrile (compounds 1a-4a, respectively) with DMC indicates that the methylation process is the fastest and the irreversible step, which drives the overall reaction to completion. The situation is reversed for arylacetic esters, where the methylation is more difficult than the demethoxycarbonylation reaction; therefore, a higher reaction temperature is required.

Full text of DOI:10.1021/jo0057699

J . Org. Chem. 2002, 67, 1071-1077
1071
Selective Mon o-C-m eth yla tion s of Ar yla ceton itr iles a n d
Ar yla ceta tes w ith Dim eth ylca r bon a te: A Mech a n istic
In vestiga tion
Pietro Tundo,* Maurizio Selva, Alvise Perosa, and Sofia Memoli
Dipartimento di Scienze Ambientali, Universit a` Ca’ Foscari, Dorsoduro 2137 - 30123 Venezia, Italy
tundop@unive.it
Received December 13, 2000
The very high mono-C-methylation selectivity (>99%) of arylacetic acid derivatives (ArCH
2
X; X )
CN, CO Me) with dimethyl carbonate (DMC) is due to a mechanism that involves consecutive
2
methoxycarbonylation, methylation, and demethoxycarbonylation steps. Important aspects of this
mechanism are clarified herein by a kinetic investigation. In the case of arylacetonitriles, at 140
°
C, the comparison of the rate constants of model reactions involving 2-phenyl propionitrile,
phenylacetonitrile, 2-methoxycarbonylphenylacetonitrile, and 2-methyl-2-methoxy carbonylphenyl
acetonitrile (compounds 1a -4a , respectively) with DMC indicates that the methylation process is
the fastest and the irreversible step, which drives the overall reaction to completion. The situation
is reversed for arylacetic esters, where the methylation is more difficult than the demethoxycar-
bonylation reaction; therefore, a higher reaction temperature is required.
In tr od u ction
objectives of Green Chemistry.6,8 In fact, the reaction
takes place in the presence of a weak base (usually K
CO ) as a catalyst, uses a nontoxic alkylating agent
DMC), requires no solvent, and produces neither organic
nor inorganic wastes to be disposed of. The only coproduct
methanol) can, in principle, be recycled for DMC produc-
2
-
Under both continuous-flow and batch conditions and
3
in the presence of K
2
CO
3
, arylacetic acid derivatives (2:
(
ArCH X; X ) CN, CO
2
2
Me) reacted with dimethyl carbon-
ate (DMC) to afford the corresponding mono-C-methyl
derivatives (2-aryl proprionitriles and methyl 2-arylpro-
pionates, 1) with unprecedented selectivity (>99%, eq
(
9
tion.
As far as the mechanism is concerned, our experimen-
tal investigation of DMC-mediated alkylations supports
1
-5
a).
For example, in an autoclave at 180 °C, phenyl-
acetonitrile was transformed quantitatively into 2-phen-
ylpropionitrile (>98% yield) in 6 h. The same reaction
carried out using methyl iodide under PTC conditions
yielded a ratio between the monomethylated and di-
methylated products that never exceeded 2.4.7
the hypothesis that the reaction does not proceed through
6
an S
CN, CO
N
2 displacement of the ArCH( X nucleophile (X )
-)
4
2
Me) on DMC (BAl2 mechanism). Rather, the
selectivity arises from two consecutive reactions involving
two intermediate species, that is, ArCH(CO Me)X (3) and
ArC(CH )(CO Me)X (4), which are observed during the
2
3
2
K CO
2
3
4
ArCH X + (CH O) CO
8
reaction (Scheme 1).
2
3
2
Initially, the carbanion [ArCH(-)X] undergoes a meth-
oxycarbonylation reaction by attacking the acyl carbon
of DMC (BAc2 mechanism); the resulting intermediate
2
ArCH(CH )X + CH OH + CO (a)
3
3
2
1
(
-)
[
2
ArCH(CO Me)X, 3] reacts through its anion [ArC -
X ) CN, COOCH3
(CO
Me)X, 3 ] with the alkyl carbon of DMC to yield the
)(CO Me)X, 4;
Al2 mechanism]. Finally, compound 4 is subjected to a
demethoxycarbonylation reaction to the final product
[ArCH(CH )X].
-
2
corresponding methyl derivative [ArC(CH
B
3
2
Besides the excellent synthetic outcome, this meth-
ylation method is environmentally benign in character
and fits the requirements for preventing pollution of
chemical processes at the source, according to the general
3
With the aim of exploring this intriguing mechanism
and the unusual selectivity, a detailed kinetic study of
the reaction was begun: each step of Scheme 1 was
considered separately, and the related kinetic constants
were determined. This investigation proved to be funda-
mental to comprehending the reaction selectivity.
*
To whom correspondence should be addressed. Fax: +39-041-234
620.
1) Tundo, P. In Continuous Flow Methods in Organic Synthesis;
Ellis Harwood, Ltd.: Chichester, U.K., 1991.
2) Tundo, P.; Trotta, F.; Moraglio, G. J . Chem. Soc., Perkin Trans.
1989, 1070.
3) Carruthers, W. In Some Modern Methods of Organic Synthesis,
rd ed.; Cambridge University Press: Cambridge, 1989.
4) Selva, M.; Marques, C. A.; Tundo, P. J . Chem. Soc., Perkin Trans.
8
(
(
1
3
1
(
(8) (a) Tundo, P.; Anastas, P.; Black, D.; Breen, J .; Collins, T.;
Memoli, S.; Miyamoto, J .; Polyakoff, M.; Tumas, W. Pure Appl. Chem.
2000, 72, 1207. (b) Selva, M.; Tundo, P. In Green Chemistry: Designing
Chemistry for the Environment; Anastas, P., Williamson, T., Eds.; ACS
Symposium Series No. 626; American Chemical Society: Washington,
DC, 1996; Chapter 7, pp 81-91.
(
1994, 1323-1328.
(
5) Tundo, P.; Selva, M. Chemtech 1995, 25, 31-35
6) Tundo, P.; Selva, M.; Bomben, A. Org. Synth. 1999, 76, 169-
(
1
77.
(
7) Mikolajczyk, M.; Grzejszczak, S.; Zatorski, A.; Montanari, F.;
Cinquini, M. Tetrahedron Lett. 1975, 3757.
(9) Delledonne, D.; Rivetti, F.; Romano U. J . Organomet. Chem.
1995, 488, c15.
1
0.1021/jo0057699 CCC: $22.00 © 2002 American Chemical Society
Published on Web 01/30/2002

1
072 J . Org. Chem., Vol. 67, No. 4, 2002
Tundo et al.
Sch em e 1. Mech a n ism of Mon om eth yla tion s of CH
2
-Active Com p ou n d s (X ) CN, CO
2
CH
3
) w ith Dim eth yl
Ca r bon a te.
Ta ble 1. Rea ction Con d ition s for th e In vestiga tion of Step s 2, 5, a n d 6 of Sch em e 1a
K2CO3/substrate
(molar ratio)
MeOH/substrate
(molar ratio)
n-decane/substrate
entry
substrate (mmol)
T (°C)
140
140
140
140
140-160
140-160
(molar ratio)
1
2
3
4
5
6
PhCH(Me)CN, 1a (1.52)
PhCH2CN, 2a (1.71)
2
2
2
b
b
0.3
0.5
0.2
b
10
PhCH(CO2Me)CN, 3a (1.14)
PhC(Me)CO2Me)CN, 4a (1.01)
10
0.1-4c
10-30c
+
-
-
K PhC (CO2Me)CN, 3a (0.94)
PhCH(CO2Me)2, 3b (0.96)
2
10
a
All reactions were carried out by using DMC in a 200 molar equiv excess with respect to the substrate. b Internal standard was not
used. c Further details can be found in Table 2.
Resu lts a n d Discu ssion
Meth oxyca r bon yla tion of 2a : Step 2. A mixture of
the substrate, DMC, and K CO was made to react in a
:200:2 molar ratio. At 140 °C, the reaction proceeded
very slowly: after 25 h, only a 10% conversion was
reached with the formation of the methylmethoxycarbo-
nylated derivative PhC(Me)(CO Me)CN (4a ) as the sole
2
2
product. The corresponding kinetic constant (k ) was not
determined.
2
3
The methylation of phenylacetonitrile (Scheme 1, 2a :
Ar ) Ph, X ) CN) was used as the model reaction.
Although this reaction was generally carried out at T g
the study was done at a lower temperature
140 °C) in order to decrease the reaction rate and allow
a more accurate determination of the kinetic constants.
All experiments were run in an autoclave since the
alkylating agent (DMC) boils at 90 °C. DMC itself was
used in large excess (200 molar equiv) with respect to
1
-2,4
1
(
80 °C,1
However, under the same reaction conditions but at a
higher temperature (180-200 °C), a number of different
2
CH -acidic compounds such as aryloxiacetic acid deriva-
2
2 3
a and acted as the solvent. K CO was always utilized
1
0
tives (ArOCH X; X ) CN, CO Me) and alkylbenzyl
as the base. If not otherwise specified, the reaction
mixture was magnetically stirred at 650 rpm. Table 1
reports the conditions. Reactions were followed by GLC
and GC/MS, on samples withdrawn at intervals from the
autoclave. Steps 2, 5, and 6 of Scheme 1 were considered,
while very fast acid-base equilibria (eqs 1, 3, 4, and 7)
were not.
2
2
sulfones (ArCH
2
SO
2
R; R ) Ar, alkyl) were selectively
1
1
monomethylated with DMC. In all cases, reactions
proceeded through methoxycarbonylated intermediates,
but importantly, the direct methylation of the monom-
2
ethyl derivatives [ArOCH(Me)X and ArCH(Me)SO R] was
never observed unless at temperatures over 220 °C.
R-Methylated benzyl ketones [ArCH(Me)COR] behaved
similarly.12
(
10) Although we already demonstrated that K
act as a catalyst (in a 5% molar amount with respect to ArCH
2 3
CO could certainly
4
2
X), in
The reaction of alcohols with DMC was even more
exclusive: only methoxycarbonylated products (ROCO -
2
order to follow the reactions more conveniently, they were all performed
using an excess of the base (2 molar equiv).

Mono-C-methylations of Arylacetonitriles and Arylacetates
J . Org. Chem., Vol. 67, No. 4, 2002 1073
conditions of reactions 2 and 5 comparable to those of
reaction 6, which requires this reagent (see later).
From a kinetic standpoint, reactions 2 and 5 are a set
of competitive reactions (eq b):
k2
k5
2
a y z 3a
98 4a
(b)
k-2
2 5
where k , k-2, and k are the corresponding kinetic
constants.13 However, since the methoxycarbonylation
reaction of 2a (2a f3a ) was much slower than its reverse
reaction (3a f 2a ), eq b was simplified by placing k
-2 (eq c):
2
,
k
k
k5
-2
2
a 7
3a
9
8 4a
(c)
F igu r e 1. Reaction of 2-methoxycarbonylphenylacetonitrile
3a ) with DMC. Determination of first-order rate constant.
(
Eq c yields a first-order rate law expression for the
disappearance of 3a with time (eq d):
Me, with no trace of methyl derivatives) were obtained
at 220 °C and above.1
Moreover, it was observed that the reaction of DMC
-
d[3a ]/dt ) k-2[3a ] + k [3a ] )
5
(
k-2 + k )[3a ] ) k[3a ] (d)
5
with 2-phenylpropionitrile [PhCH(CH
yield the doubly methylated derivative [PhC(CH
even operating at 160 °C and for prolonged reaction times
18 h).
In light of these results, structural and acidity factors
3
)CN, 1a ] did not
3 2
)
CN]
5
where k ) k-2 + k . According to eq d, the substrate 3a
follows a pseudo-first-order rate law (Figure 1). This
(
reaction was repeated three times, and an average kobs
-2
-1
of 1.1 × 10 min was calculated.
The formations of 2a and 4a [PhC(CH
are given, respectively, by (Scheme 1):
strongly suggested that the direct methylation of 2a (via
a BAl2 mechanism) with DMC could also be reasonably
ruled out. Therefore, the formation of the product 4a in
the investigated reaction (step 2) had to occur through
the initial methoxycarbonylation of the reagent (2a )
followed by a subsequent methylation.
3
2 3
)(CO CH )CN]
d[2a ]/dt ) k-2[3a ] ) k-2[3a ]oe-kt or [2a ] )
[2a ] + (k /k)[3a ] (1 - e ) (e)
-kt
o
-2
o
Presently, there are no clear reasons why the anion
-
undergoes methylation while 2a - does not. A possible
and
3
a
explanation could lay in the hard-soft character of the
d[4a ]/dt ) k [3a ] - k [4a ] + k-6[1a ]
(f)
)-
involved moieties: the methoxycarbonylation of the more
5
6
-
stabilized anion 3a is probably much slower (and
reversible) than that of the less hindered (and more
where 1a was the monomethyl derivative [PhCH(CH
3
-
reactive) 2a . Therefore, the observed behavior would be
CN].
due to the high acidity of 3a rather than to a high
Eq f was simplified by applying the following consid-
erations. (i) The reaction of 3a with DMC showed that
the concentration of 1a was low compared to that of 3a
until the latter was present, being [3a ] ≈ 7[1a ] through-
out all the reaction. (ii) An independent investigation of
step 6 indicated that the methoxycarbonylation of the
monomethyl derivative (1a f4a ) occurred very slowly:
-
reactivity of 3a . Also, a factor that should account for
-
-
the different nucleophilicity of 2a and 3a is that the
reaction takes place in the presence of a solid base;
therefore, surface phenomena may affect the reaction
outcome (see later). Indirect evidence comes from the
DMC-methylation of arylacetonitriles, promoted by ho-
4
mogeneous strong bases (phosphazenes): in this case, the
k
2
-6 has to be very small in comparison to k .
reaction is no longer selective since the double methyla-
tion of the substrates occurs to a large extent.
Therefore, eq f gives eq g by omitting the term
k
-6[1a ]:
These aspects are, however, beyond the scope of the
present work, the aim of which is to understand and
describe the reaction selectivity through a kinetic inves-
tigation.
d[4a ]/dt ) k [3a ] - k [4a ]
(g)
5
6
Since [4a ]
o
) 0, one may assume that the condition k
[4a ] can be verified for a certain reaction time (see
later). If so, eq g is further simplified to the following:
5
[3a ]
.
k
6
Dem et h oxyca r b on yla t ion a n d Met h yla t ion of
3
a : Step s 2 a n d 5. 2-Methoxycarbonylphenylacetonitrile
(
3a ), the key intermediate of both equilibrium 2 and
d[4a ]/dt ) k [3a ]
(h)
5
reaction 5, was used for the investigation of these two
steps. Experiments were carried out by reacting a
mixture of 3a , DMC, K
and considering that also [2a ]
eqs e and h yields:
o
) 0, the combination of
2 3
CO , and methanol in a 1:200:2:
0 molar ratio. Methanol was used to make experimental
1
(
13) Scheme 1 should dictate the use of carbanion concentrations
-
-
(
11) (a) Bomben, A.; Marques, C. A.; Selva, M.; Tundo, P. Tetrahe-
(2a , 3a ) rather than [2a ] and [3a ] in kinetic equations. However,
whatever the concentrations of 2a and 3a - are, ratios 2a /2a - and
3a /3a - must be constant throughout the reaction (steps 2 and 5); the
apparent simplification of eq b, which serves as a working hypothesis,
does not alter kinetic results.
-
dron 1995, 51, 11573. (b) Selva, M.; Bomben, A.; Tundo P. J . Chem.
Res., Synop 1997, 448-449.
(12) Selva, M.; Marques, C. A.; Tundo P. Gazz. Chim. It. 1993, 123,
5
15-518.

1
074 J . Org. Chem., Vol. 67, No. 4, 2002
2a ]/[4a ] ) k-2/k5
Tundo et al.
[
(i)
Eq i means that the ratio between 2a and 4a should
remain constant and independent from the concentration
of the reagent 3a . Eq i was validated experimentally. As
can be seen in Figure 2a (solid line), when 3a was made
to react with DMC, the ratio [2a ]/[4a ] remained unaltered
(
0.3). Moreover, the plot of the concentration of the
-
kt
product 2a versus (1 - e ) showed a good linear
correlation (Figure 2b), which proved the pseudo-first-
order assumed by eq e.
Since the investigated reactions take place in the
presence of a heterogeneous catalyst (K CO ), it should
2 3
be noted that the rate may also depend on diffusion
phenomena of reactants/products on the solid surface. We
have already reported this in the particular case of
methylation of phenylacetonitrile, for which it was shown
that conversions were dependent on the stirring rate of
the reaction mixture.6 To account for this fact, the
reaction of 3a with DMC was also investigated by
applying different stirring rates (vstir). Results are re-
ported in Table 2.
A mass transport limitation seemed to occur at low vstir
(compare conversion/time in entries 1-3). However, the
ratio [2a ]/[4a ] was substantially the same (0.3) in all
cases, meaning that the ratio between rates of demethoxy-
carbonylation and methylation of 3a was also free from
-
a diffusion control. Even if the amount of the anion 3a
(from which both compounds 2a and 4a are derived) did
depend on the stirring, it was not necessary to measure
its actual concentration in the reaction mixture (and
hence, the equilibrium constant of eq 4 in Scheme 1).
Therefore, for kinetic calculations, the reagent concentra-
F igu r e 2. (a) Concentration ratio [2a ]/[4a ] during the reaction
of 3a with DMC, at 140 °C. Solid line, experimental course;
dashed line, predicted curve from eq n. (b) Molar concentration
-
tion measured (by GC) as [3a ]meas ) {[3a ]sol + [3a ]}
-
kt
of 2a versus 1 - e during the reaction of 3a with DMC.
(
where [3a ]sol was the reagent concentration in solution)
1
4
appeared to be adequate. Bearing this in mind, we were
Ta ble 2. Rea ction of 3a w ith DMC Ca r r ied Ou t a t 140 °C
able to determine the rate constants k-2 and k
5
from the
w ith Differ en t Stir r in g Ra tesa
combination of eqs d and i:
products,
stirring reaction
(
% by GC)
1a 2a 4a
14 44
rate
time
conversion
(%)
ratio of
2a /4a
k ) k-2 + k5
(j1)
(j2)
entry
(rpm)
(min)
{
[
2a ]/[4a ] ) k-2/k5
1
2
3
4
5
80
150
210
120
240
67
83
97
92
9
0.31
0.30
0.32
0.28
0.31
3
00
13 16 54
47 12 38
28 14 50
58 10 32
650
-
3
The solution of this system gave k-2 ) 2.5 × 10 ( 3 ×
-
4
-1
-3
-4
-1 15
1100
1
0
min and k
5
) 7.9 × 10 ( 2 × 10 min
.
100
These data show that k
5
≈ 3k-2, i.e., the methylation
a
All reactions were carried out with a mixture of 3a , DMC,
K2CO3, and MeOH in a 1:30:2:10 molar ratio.
of 3a (step 5) occurs faster than its demethoxycarbon-
ylation (step 2). This also proved that kinetic analyses
of the reactions of DMC with 2a and 3a were in good
agreement: when compound 3a was formed from 2a , it
readily reacted to give 4a ; again, the direct reaction of
methoxycarbonyl intermediate [ArCH(CO
2
Me)CN, 3] was
always much lower than that of the methylmethoxycar-
16
2
bonyl intermediate [ArC(Me)(CO Me)CN, 4].
3
a with DMC afforded 4a faster than 2a . The methyla-
It should be stressed, however, that all the above
considerations as well as the evaluation of constants k-2
tion step being more rapid than the methoxycarbonyla-
tion and the demethoxycarbonylation ones pushed the
reaction onward.
and k
[4a ]. Under which conditions does this relation hold?
Obviously, the answer requires the determination of the
5 5
rely on the occurrence of the inequality k [3a ] .
k
6
This behavior may also explain some results previously
4
reported: for a set of different acetonitriles (ArCH
2
CN;
; X ) H, p-Me, o-Me, p-F, p-Cl, o-MeO, p-MeO,
m-MeO), it was observed that the concentration of the
k
6
constant related to the demethoxycarbonylation of 4a .
Dem eth oxyca r bon yla tion of 4a : Step 6. The reac-
tion of 2-methoxycarbonyl-2-phenylpropionitrile (4a ) with
6 4
Ar: XC H
methanol was investigated by using a mixture of the
substrate, DMC, K CO , and methanol in a 1:200:2:10
2 3
(
14) Before GC analysis, all samples of the reaction mixture were
acidified.
(
15) Since the reaction of 3a with DMC was repeated three times,
molar ratio. The disappearance of the substrate with time
three different k values were available. The system i was solved by
coupling the corresponding k value to a set of six values of the [2a ]/
[
4a ] ratio (calculated for each sample analyzed during such reactions).
(16) However, electron-withdrawing substituents (i.e., p-Cl and p-F)
seemed to speed up the overall reaction, probably through a stabiliza-
tion of the involved carbanions.
5
Thus, the determination of both k-2 and k was averaged over a total
of 18 solutions of i.

Mono-C-methylations of Arylacetonitriles and Arylacetates
J . Org. Chem., Vol. 67, No. 4, 2002 1075
Ta ble 3. Dem eth oxyca r bon yla tion of
2
-Meth oxyca r bon yl-2-p h en ylp r op ion itr ile (4a ): In flu en ce
a
of Meth a n ol a n d P ota ssiu m Ca r bon a te
-
3
MeOH:4a
(molar ratio)
K2CO3:4a
(molar ratio)
no. of
runs
kobs x 10
-1
entry
(min ) b
1
2
3
4
5
6
10
30
10
10
10
10
2
2
0.1
1
3
3
3
2
3
2
3
3.0
7.6
1.5
1.9
3.1
3.1
4
a
All reactions were carried out at 140 °C using a DMC/substrate
molar ratio of 200. b kobs values were averaged over the number
of runs.
However, the further point is why the ratio [2a ]/[4a ]
seems to be constant throughout the reaction (Figure 2a,
solid line). This can be discussed through the combination
of eqs e and g:
F igu r e 3. Demethoxycarbonylation reaction of 4a (step 6 of
Scheme 1).
[
2a ]/[4a ] ) [(k-2/k-2 + k )/(k /k - k )] ×
5
5
5
6
-
(k-2+k5)t
-k6t
- e-k5t) (n)
(
1 - e
)/(e
whose corresponding plot (using the k-2, k
5
, and k values)
6
is the dashed line of Figure 2a. Eq n exhibits a rather
good overlap to the experimental curve. However, at a
high reaction time (>250 min), the two plots tend to
diverge. Equation n correctly predicts that [2a ]/[4a ]
should increase with time: in fact, 4a is consumed via
the demethoxycarbonylation process (step 6, see also
Figure 4), while the amount of 2a remains constant since
its methoxycarbonylation does not take place at 140 °C
(step 2, Scheme 1). This behavior appears less evidently
from the experimental dots probably because of the
smooth slope of the curve.19
F igu r e 4. Molar concentration of 4a during the reaction of
3
a with DMC. As calculated from eq m.
5 6
In summary, the comparison between k-2, k , and k
showed that the methylation step (3a f4a ) took place
faster than both demethoxycarbonylation reactions
followed a pseudo-first-order law (Figure 3) from which
the rate constant, determined by repeating the reaction
(3a f2a and 4a f1a ), which in turn had comparable
-
3
-1
three times, was kobs ) 2.9 × 10 min
.
kinetic constants. Therefore, the methylation step was
responsible for driving both equilibria 2 and 6 to the right
and accounted for the observed selectivity. The meth-
oxycarbonylation step (2a f3a ) was instead the rate
determining process of the overall reaction.
From Scheme 1, the rate of disappearance of 4a is
given by the following:17
-
d[4a ]/dt ) k [4a ] - k-6[1a ]
(l)
6
Step 6 was further investigated to explore whether the
reaction rate could be affected by both the base and the
coproduct methanol. Thus, in separate experiments, the
The methoxycarbonylation of 1a with DMC was ex-
tremely slow (only 8% conversion after 25 h), k-6 , k
Hence, kobs ) k
With both constants k
of validity for eq h (i.e., of the inequality k
4a ]) can be estimated through the integrated form of eq
g:
6
.
6
.
2 3
amount of K CO and the volume of methanol were
5
and k
6
determined, the range
6
[3a ] . k -
changed. Reactions were all of pseudo-first order with
respect to the substrate. Results are reported in Table 3
and show the following. (i) The rate constants doubled
5
[
when the K
and 3, kobs ) 1.5 and 3 × 10 min , respectively); a
further increase of the base amount (K CO /4a > 2) did
2 3
CO /4a ratio ranged from 0.1 to 2 (entries 1
-
3
-1
-k6t
- e-k5t] (m)18
[
4a ] ) k /(k - k ) [3a ] [e
5
5
6
o
2
3
The comparison of this concentration profile (eq m, Figure
) with the one of the reagent 3a (Figure 1) shows that,
as far as the reaction of 3a with DMC proceeds to a 40%
[4a ] e 0.2 k [3a ]
is verified. Within this limit, it is assumed that eq h is
validated as well.
(
18) The concentration [3a ]
reagent.
19) To gauge this discrepancy, the ratio [2a ]/[4a ] can also be
o
was known, being compound 3a the
4
(
expressed with the assumption that k-2 ∼ k
6
) (1/3)k
5
6
. From this
conversion (∼100 min), the condition k
6
5
-4k
t
-k
t
6 6
-2k t -k t
assumption, it follows that [2a ]/[4a ] ) (1/6)(1 - e
6
)/(e
- e
e
)
-k6t
-2k6t
(eq o) and [2a ]/[4a ] ) (1/6)(2/e )(1 + e
6
) ) (1/3) cosh k t (eq p).
The Taylor expansion of cosh k t finally yields [2a ]/[4a ] ) (1/3) cosh
6
2
2
6 6 6
k t ) 1/3 [1 - (1/2)(k t) + ...] (eq q). If the term (1/2)(k t) ) 0.1, it can
be plausibly neglected. Hence: [2a ]/[4a ] ) (1/3) (eq r), which is the
experimentally determined value (Figure 2a). Eq r implies that t )
(
17) Both reaction 6 and its reverse reaction are bimolecular, and
-
1/2
-1
6
the correct rate law should be: -d[4a ]/dt ) k′
k′-6[1a ][DMC]. However, once the steady state is reached, since the
base is catalytic and DMC is used in a large excess, both [MeO ] and
DMC] should not vary with time. Accordingly, k
6
[4a ][MeO ] -
(0.2) (k
)
∼ 150 min: below this reaction time and with the initial
-
approximation for rate constants, the two plots in Figure 2a, overlap
within 10% error. The same reasoning also indicates that an
acceptable overlap (within a 20% error) still holds until 220 min of
reaction.
-
a
-
[
k
6
) k′
6
[MeO ] and
-6 ) k′-6[DMC].

1
076 J . Org. Chem., Vol. 67, No. 4, 2002
Tundo et al.
was made to react with DMC under the same conditions
as for PhCH(CO
Me)CN (Substrate:DMC:MeOH:K CO
2 2 3
in a 1:200:10:2 molar ratio). At 140 °C, a slow demethoxy-
carbonylation of 3b was observed (step 2): after 2 h,
conversion was 4% and phenyl methyl acetate (2b) was
the only product. Instead, at 160 °C, both methylation
(step 5) and demethoxycarbonylation of the substrate
took place: after 3 h (40% conversion), the amounts of
methyl derivative PhC(Me)(CO
and 34%, respectively.
2 2
Me) (4b) and 2b were 6
Although kinetic constants were not determined (reac-
tion times were too long), these results draw an impor-
tant difference between esters and nitriles: in the case
of nitriles, the reactivity of 3a proved that the methyla-
tion step was the fastest reaction (k
esters, the demethoxycarbonylation was the faster pro-
cess by far.
5
≈ 3k-2), while for
F igu r e 5. Logarithmic Eyring plot for the methylation of 3a -
with DMC.
Con clu sion s
The kinetic analysis of the mechanism of Scheme 1
allows the following general considerations. (i) The
similarity of k-2 and k reveals that both the starting
6
reagent 2a and its methyl derivative 1a undergo
demethoxycarbonylation reactions at comparable rates,
while the methylation step of the intermediate 4a is the
fastest reaction. (ii) The rate determining step of the
overall transformation is the methoxycarbonylation reac-
tion, which is so slow for both 2a and 1a that the
not produce any further rate enhancement (entries 1, 5,
and 6), which meant that the catalytic effect of K
had a cutoff point probably due to its solubility in DMC.
2
CO
3
4
(ii) The rate constant increased by a factor of ap-
proximately 3 when the molar ratio MeOH/4a was tripled
-
3
-1
(
entries 1 and 2, kobs ) 3 and 7.6 × 10
min ,
respectively). Both the base and methanol concentrations
had a clear influence on the demethoxycarbonylation
reaction, as Scheme 1 predicted (eq 6).
determination of the related kinetic constants (k
2
and k-6)
Meth yla tion of th e P ota ssiu m Sa lt of 2-Meth oxy-
is impracticable at 140 °C. This is probably due to the
(
+)
(-)
ca r bon ylp h en yla ceton itr ile [K P h C (CO
3
2
CH
3
)CN,
-
-
low concentration of nucleophiles 2a and 1a attainable
in the presence of a weak base such as K
-
a ] w ith DMC: Activa tion En er gy. With the aim of
2
CO3.
determining the activation energy of the reaction, com-
On the whole, the comparison of the kinetic behavior
of the investigated steps reveals that the nonequilibrium
methylation reaction is crucial for driving the overall
process to completion. In fact, the higher rate of step 5
allows both the rapid consumption of 3a and the ac-
cumulation of 4a , which serves as a reactant for step 6;
in other words, both equilibria 2 and 6 are controlled by
the irreversible reaction 5.
-
pound 3a was reacted with DMC at three different
temperatures (140, 150, and 160 °C) in the absence of
the heterogeneous base. To avoid the parallel demethoxy-
-
carbonylation of 3a , no methanol was added. The
reaction gave 4a as the sole product, with a pseudo-first-
order rate law whose rate constant (kobs) was determined
by repeating the reaction five times at each of the chosen
temperatures. The corresponding logarithmic plot of the
The described mechanism evinces the crucial action of
the methoxycarbonyl group, which by increasing the
acidity of 3 increases the concentration of the correspond-
Eyring equation [ln(kobsh/k T) ) ∆S*/R - ∆H*/RT] gave
B
a good linear correlation (Figure 5), yielding an activation
enthalpy of 112 kJ /mol.20 This value, compared for
example to the activation energy of the hydrolysis of
(-)
-
ing anion, ArC (CO
2 3
CH )R (3 ), thereby allowing step
5
to occur rapidly. This being established, a key aspect
2
0
methyl halides (≈100 kJ /mol),
is in agreement with
-
still remains unclear: why 3 undergoes direct meth-
the elevated temperatures (>140 °C) required for the
DMC methylation reactions to proceed at appreciable
rates.
-
-
ylation, while 1 or 2 does not. Steric and stability
factors of the involved anions can be considered, but a
more in-depth investigation is necessary. Also, the spe-
cific influence of the heterogeneous catalyst must be
addressed.
However, the different reaction environment (absence
of methanol and K CO ) does not allow strict comparison
2 3
of the kinetic constant to that obtained for step 5 that
Arylacetoesters behave differently with respect to the
nitriles. The reaction of dimethyl phenylmalonate with
DMC proves that the demethoxycarbonylation step is
easier than the methylation reaction. This fact could be
the reason esters need higher temperatures for the
reaction to be completed. However, the monomethyl
selectivities are comparable (>99%) for both of the two
classes of arylacetic acid derivatives, even using a large
excess (200 molar equiv) of the methylating agent.
was previously discussed (in fact, at 140 °C, kobs values
-
3
-2
-1
were 7.9 × 10 and 3.2 × 10 min , respectively).
Meth yla tion of Ar yla cetic Ester s by DMC. Methyl
aryl acetates with DMC give mono-C-methyl derivatives
through the mechanism outlined in Scheme 1 as well;
however, at comparable reaction times, they required a
higher temperature (200-220 °C) than nitriles (180-190
,8,11
°
C).4
2 2
Dimethyl phenylmalonate [PhCH(CO Me) , 3b]
(
20) Maskill, H. In The Physical Basis of Organic Chemistry; Oxford
University Press: Oxford, 1989; p 258. h and k are Planck and
Boltzmann constants. The activation energy was calculated from the
Exp er im en ta l Section
B
All compounds used were ACS grade and were used without
a
gradient of Figure 4 (∆H*/R) using the equation E ) ∆H* + RT (by
placing T ) 423 K, as the mean absolute temperature of the range
over which kobs values were determined).
1
13
further purification. H and C NMR spectra were recorded
using a Bruker Ac 200 spectrometer (200 and 50 MHz,

Mono-C-methylations of Arylacetonitriles and Arylacetates
respectively) using CDCl as the solvent with TMS as the
internal standard. GC analyses were performed on a Varian
GC 3400 CX apparatus using a 30 m, DB5 capillary column.
GC/MS analyses were performed at 70 eV on an HP 5971 mass
detector coupled to an HP 5890-Series II gas chromatograph
fitted with a 30 m, DB5 capillary column. Melting points were
determined on a B u¨ chi 535 melting point apparatus and are
uncorrected.
J . Org. Chem., Vol. 67, No. 4, 2002 1077
3
presence of 2-methoxycarbonylphenylpropionitrile as the sole
compound. Then, the mixture was cooled to room temperature;
K CO was filtered off and carefully washed with acetone. The
2 3
filtrate was distilled under vacuum to remove light solvents
and DMF; the remaining brown liquid residue was treated
3
with diethyl ether (200 cm ) and allowed to stand overnight
in a refrigerator (+4 °C). After filtration, a pale yellow solid
3
was obtained: it was dissolved with acetone (≈15 cm ),
3
Rea ction s Ca r r ied Ou t in Au tocla ve: Gen er a l P r oce-
d u r e. A stainless steel (AISI 316) autoclave (internal volume
precipitated with additional diethyl ether (150 cm ), and
-
finally, filtered. The salt 3a was obtained in a 59% yield (23.2
3
1
of 250 cm ) was loaded with a mixture of DMC, the substrate
g). H NMR (DMSO-d
6
): δ ppm 7.55 (d, 2H, J ) 7.4 Hz, ortho
(0.2 g), the base (K
2
CO
3
), methanol (where indicated), and
aromatic), 7.01 (t, 2H, J ) 7.4 Hz, meta aromatic), 6.81 (t,
13
n-decane as the internal standard in the molar ratios reported
below (Table 3). At room temperature and before each reaction,
1H, J ) 7.4 Hz, para aromatic), 3.39 (s, 3H, COOCH
3
).
C
NMR (DMSO, d ): δ ppm 168 (CO), 127, 121, 117 (C,
6
air was removed by a purging valve with a N
2
stream. The
aromatics), 48 (OCH ). The literature reported the formation
3
of the sodium salt Na( PhC (COOCH )CN obtained in a
+)
(-)
autoclave was placed in a tubular oven heated by an oil-
circulating jacket, while the mixture was kept under magnetic
stirring. A thermocouple and a needle valve were fixed onto
the autoclave head, the former dipping into the reaction
mixture and the latter connected to a 1/8 in stainless steel
suction pipe that, in turn, was immersed into the reaction
3
methanol solution of sodium methoxide and 3a .24
Ca libr a tion Cu r ves. The evaluation of rate constants for
the steps of Scheme 1, required a high degree of accuracy in
the GC analyses. Thus, calibration curves were determined
-
for both compounds 3a and 4a -3a by using n-decane as the
3
mixture. In this way, samples (0.3 cm ) could be withdrawn
internal standard. The general procedure is as follows. A 0.05
M solution of each substrate was prepared using DMC as the
during the reaction and analyzed by both GC and GC/MS.
2
-Meth oxycar bon ylph en ylaceton itr ile (3a) an d 2-Meth -
3
solvent. Four aliquots of 5, 10, 15, and 20 cm were taken and
oxyca r bon ylm eth yl-2-p h en ylp r op ion itr ile (4a ) were syn-
placed separately into four volumetric flasks. Each solution
4
,21
thesized according to established procedures;
characteriza-
was added with 20 cm3 of a 0.026 M solution of n-decane in
tion data were in agreement to those reported in the
3
DMC and brought to a final volume of 50 cm with additional
2
2,23
literature.
P ota ssiu m sa lt of 2-Meth oxyca r bon ylp h en yla ceton i-
DMC. Then, each solution was analyzed twice by GC. In the
-
3
case of 3a , analytical samples of 0.2 cm were acidified and
t r ile [K( P h C (CO
acetonitrile (20 g, 0.17 mol), K
100 g, 1.37 mol) was loaded in a 250 mL three-necked flask
equipped with a condenser, a dropping funnel filled with DMC
+)
(-)
CH
3
)CN, 3a ]. A mixture of phenyl-
CO (47 g, 0.34 mol), and DMF
-
2
extracted with diethyl ether as described in the above prepara-
tion of 3a . The calibration curves were obtained by plotting,
for each solution, the ratio of the chromatographic areas of
the substrate and the standard versus the concentration of
the substrate. Linear fits with r g 0.99 were obtained in all
cases.
2
3
-
(
3
(
50 cm , 0.52 mol), and a glass screw-capped tube (fitted with
a rubber silicon septum) used for sample withdrawal. The flask
was heated in an oil bath at the reflux temperature (155 °C)
3
while a gentle N
2
stream (0.5 cm /min) was allowed in and
Ack n ow led gm en t. Murst (Italian Ministry of Uni-
versity and Scientific and Technological Research) and
INCA (Interuniversity Consortium Chemistry for the
Environment) are gratefully acknowledged for financial
support. Prof. Achille Giacometti is also gratefully
acknowledged for his helpful discussion and suggestions
on the mathematical aspects of the work.
maintained for the whole reaction. At intervals (30 min), small
aliquots (1 cm ) of DMC were dripped into the reaction
3
mixture, which was magnetically stirred. The course of the
reaction was followed by GC and GC/MS: before each analysis,
3
3
samples (0.2 cm ) were added with water (1 cm ), acidified with
concd HCl (3 drops), and then extracted with diethyl ether (2
3
cm ). The completion of the reaction was indicated by the
(
21) Horning, E. C.; Finelli, A. F. Org. Synth. Vol. Coll. IV; Wiley:
967, pp 461-63.
22) Cox, R. J .; O’Hagan, D. J . Chem. Soc., Perkin Trans. 1 1991,
537-40.
23) Gompper, R.; Vogt, H.-H. Chem. Ber. 1981, 114, 6-83.
J O0057699
1
2
(
(24) Abbotto, A.; Bradamante, S.; Pagani, G. A. J . Org. Chem. 1993,
58, 449-455.
(