Effect of pressure on sterically congested cyanoalkylation reactions of alcohols

Article abstract of DOI:10.1016/S0040-4020(02)00347-2

The pressure effect on the phosphine-catalyzed nucleophilic addition of alcohols to unsaturated nitriles is examined. As a general result, pressure promotes these reactions. Their sensitivity to pressure increases with increasing steric congestion of either the alcohol or the nitrile. Activation volumes are found to be very negative pointing not only to a late transition state, but essentially to a considerable electrostriction contribution depending on steric hindrance to ionization. This means that pressures favors formation of the carbanion and attack of the nitrile. The results highlight the synthetic utility of high pressure to remove steric inhibition.

Full text of DOI:10.1016/S0040-4020(02)00347-2

TETRAHEDRON
Pergamon
Tetrahedron 58 ꢀ2002) 4311±4317
Effect of pressure on sterically congested cyanoalkylation
reactions of alcohols
p
G. Jenner
Laboratoire de Pi e zochimie Organique ꢀUMR 7123), Facult e de Chimie, Universite Louis Pasteur,
1
rue Blaise Pascal, 67008 Strasbourg, France
Received 22 January 2002; revised 6 March 2002; accepted 25 March 2002
AbstractÐThe pressure effect on the phosphine-catalyzed nucleophilic addition of alcohols to unsaturated nitriles is examined. As a general
result, pressure promotes these reactions. Their sensitivity to pressure increases with increasing steric congestion of either the alcohol or the
nitrile. Activation volumes are found to be very negative pointing not only to a late transition state, but essentially to a considerable
electrostriction contribution depending on steric hindrance to ionization. This means that pressures favors formation of the carbanion and
attack of the nitrile. The results highlight the synthetic utility of high pressure to remove steric inhibition. q 2002 Elsevier Science Ltd. All
rights reserved.
1
. Introduction
mediate alkoxy anion. The base must be strong and consists
usually of aqueous solutions of alkali hydroxides, alkoxides,
tetraalkyl ammonium hydroxides. New heterogeneous basic
catalysts have been recently proposed to improve the
catalytic activity in the cyanoethylation of alcohols by
The process traditionally known as cyanoethylation
involves cyanoethylating agents such as acrylonitrile.
1
Cyanoethylation of monohydric alcohols with acrylonitrile
is an old reaction already studied in the 1920s. A compre-
2
4
acrylonitrile. However, alkyl substituted acrylic nitriles
3
hensive review appeared a long time ago. In spite of its
react with alcohols less easily. The two methyl substituted
unsaturated nitriles, methacrylonitrile and crotononitrile,
require much longer reaction times than acrylonitrile in
ancientness the reaction deserves a wide and real interest for
the synthesis of number of compounds e.g. drug inter-
mediates. The general equation can be written as ꢀEq. ꢀ1)):
3
their reaction with methanol.
Some time ago we investigated the pressure effect in some
nucleophilic additions such as Michael and Henry reactions
and found that they are fairly promoted by pressure. We
ꢀ
1†
5
The reaction is apparently a Michael type nucleophilic
addition and occurs only in presence of a base via the inter-
felt of interest to examine the high pressure synthesis of
hindered cyanoethers which can be forbiddingly dif®cult
Scheme 1.
Keywords: high pressure kinetics; steric hindrance; nucleophilic addition; cyanoether synthesis.
Tel.: 133-3-90241679; fax: 133-3-90241739; e-mail: jenner@chimie.u-strasbg.fr
p
0040±4020/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020ꢀ02)00347-2

4312
G. Jenner / Tetrahedron 58 ꢀ2002) 4311±4317
Table 1. Effect of the base on cyanoalkylation of 1-propanol by methacryl-
onitrile
they lead to heterogeneization of the reaction medium, they
require stirring which cannot be achieved in our high
pressure device. Organic bases such as tertiary amines are
ineffective as is also the case of weak ꢀDMAP) or even
strong bases ꢀDABCO, DBU, guanidine). In the runs carried
out with the above bases high pressure ꢀ300 MPa) does not
signi®cantly enhance the reactivity. Good results, however,
were obtained with phosphorous bases ꢀphosphines and
a
Base
Yield ꢀ%)
0
.1 MPa
300 MPa
b
Aqueous NaOH
2
0
0
No run
No run
0
1
b
O
Aqueous Bu
DABCO
DMAP
4
NF´3H
2
7
0
phosphazenes). Phosphazenes are extremely strong bases,
8
DBU
Tetramethylguanidine
0
0
18
2
48
No run
3
67
No run
99
one of the strongest being phosphazene Base P -tBu: a 48%
4
yield of cyanopropylether is produced at ambient pressure
whereas the reaction is quantitative at 300 MPa. Tri-n-
butylphosphine though much less basic is revealed as a
convenient catalyst leading to 18 and 67% yields, respec-
tively. The availability and cheapness of the phosphine
prompted us to use it as the base in most reactions examined
in this paper.
Tri-n-butylphosphine
Tri-n-butylphosphine
Phosphazene
b
c
Nitrile ꢀ1.8 mmol), propanol ꢀ13.3 mmol), T ꢀ323 K), t ꢀ4 h).
Base ꢀ0.18 mmol). DABCO ꢀdiazabicyclo[2.2.2]octane), DMA ꢀ4-dimethyl-
a
aminopyridine), DBU ꢀ1,8-diazabicyclo[5.4.0]undec-7-ene).
Water ꢀ28 mmol) is added. No agitation is provided.
Phosphazene base P -tBu ꢀfrom Fluka).
4
b
c
The reactions presented throughout were found not rever-
sible under our conditions. This was checked for some
cyanoalkylethers which were exposed to the same con-
ditions as for the forward reactions ꢀtemperature, concen-
tration of phosphine, time) or worked up upon addition
of aqueous chlorhydric acid. The cyanoalkylethers were
recovered unchanged in the same amount as initial.
or impossible to synthesize under ordinary conditions. A
previous note reported the considerable yield improvements
by application of high pressure.
6
2. Results
2.2. Addition of primary linear alcohols
The study was made essentially with methacrylonitrile and
crotononitrile. The nitriles are two isomers only differing by
the position of the methyl group at the a or b position. The
reaction sequence for a classic Michael reaction is depicted
in Scheme 1. Clearly, the attack on the b position in the rate
determining step, obviously, should be more dif®cult when
In a second step, we investigated the pressure effect in the
cyanoalkylation of linear primary alcohols by both unsatu-
rated nitriles ꢀTable 2). The results show that in agreement
with Ref. 3, crotononitrile reacts easier than methacrylo-
nitrile in the cyanoalkylation of methanol. In the case
of ethanol, reactions yields are similar. However, with
increasing chain length of the alcohol, the addition reaction
becomes more sluggish and it does not proceed with
R ±H, meaning that the addition of an alcohol would occur
1
more easily on methacrylonitrile.
1
-heptanol. The reactivity of alcohols with crotononitrile
2
.1. Selection of the base
is signi®cantly depressed, in larger extent than in the
reactions involving methacrylonitrile.
In the preliminary stage we were seeking an appropriate
base for our purpose. The traditional bases employed in
cyanoethylationÐaqueous alkali hydroxides or tetraalkyl-
ammonium saltsÐare not suitable ꢀTable 1). In addition, as
Considering the pressure effect, the reaction becomes
gradually more pressure sensitive independently of the
structure of the nitrile when the chain of the alcohol is
made longer. Another item worth noting in Table 2 includes
the fact that it becomes clear that at 300 MPa crotononitrile
reactions are more accelerated than the corresponding
methacrylonitrile reactions. This is a contrasting though
highly interesting result since the reaction center in
crotononitrile is much less accessible.
Table 2. Cyanoalkylation of linear primary alcohols ꢀEq. ꢀ1)). Effect of
pressure
R
R
1
R
2
Time ꢀh)
Yields ꢀ%)
0
.1 MPa
300 MPa
CH
CH
3
3
H
CH
H
CH
H
H
CH
H
CH
H
CH
H
CH
CH
H
CH
H
H
CH
H
CH
H
CH
H
CH
H
3
3
2
2
3.5
3.5
2
4
4
4
4
4
4
4
4
62
80
33
32
34
18
5
10
2
7
100
100
78
99
100
67
94
57
95
68
85
23
33
2.3. Addition of branched linear alcohols
3
C
C
C
C
C
C
C
C
C
C
C
2
H
2
H
3
H
3
H
3
H
4
H
4
H
5
H
5
H
7
H
7
H
5
5
3
7
7
3
3
3
3
7
3
3
3
3
9
ꢀ2†
9
11
11
15
15
2
0
0
Informed by the results listed in Table 2, we were expecting
that increasing bulkiness of the alkyl rest of the alcohol ꢀin
the case of branched primary alcohols) would induce low or
Nitrile ꢀ1.8 mmol), solvent ꢀROH), tri-n-butylphosphine ꢀ0.18 mmol), T
323 K).
ꢀ

G. Jenner / Tetrahedron 58 ꢀ2002) 4311±4317
Table 3. Cyanoalkylation of branched primary alcohols ꢀEq. ꢀ2)). Effect of pressure
4313
0
00
Entry
R
R
R
R
1
R
2
Time ꢀh)
Yields ꢀ%)
0
.1 MPa
300 MPa
1
2
3
4
5
6
7
8
9
CH
CH
CH
3
3
3
CH
CH
CH
CH
CH
CH
CH
CH
CH
3
3
3
3
3
3
3
3
3
H
H
CH
CH
H
H
CH
H
H
CH
H
H
CH
H
3
3
3
3
8
8
8
22
22
24
24
24
8
22
4
0
38
11
0
13
1
70
81
61
79
H
H
H
H
H
CH
CH
CH
H
3
C
2
H
5
5
5
C
C
C
2
H
2
H
2
H
5
5
5
H
CH
a
3
95
C
2
H
80
84
CH
CH
CH
3
3
3
3
3
3
H
CH
a
92
3
C
2
H
0
43
2
40
56
10
1
Cyclohexyl
Cyclohexyl
H
CH
24
24
a
58
1
H
3
Nitrile ꢀ1.8 mmol), solvent ꢀROH), tri-n-butylphosphine ꢀ0.18 mmol), T ꢀ323 K).
The Baylis±Hillman dimer of crotononitrile is also formed.
a
8
no reactivity at ambient pressure and, on the other hand, an
interesting pressure effect. We examined the cyanoalkyl-
ation of isobutanol ꢀentries 1±3), 2-methyl-1-butanol
signi®cant observation is that application of pressure
seemingly removes steric restrictions. However, the
pressure dependence varies with the type of nitrile
involved. Reactions of methacrylonitrile are fairly sensi-
tive. The yields at 300 MPa are increased by a factor
varying from 1.5 to 7 depending on the bulky nature of
R, R , R . The corresponding additions involving methyl
and ethyl b-substituted acrylic nitriles are revealed as
highly pressure dependent reactions. With increasing
steric complexity nearly quantitative yields are obtained
ꢀentries 5, 8). Some reactions occur only under 300 MPa
pressure ꢀentries 3, 6, 8, 9).
ꢀ
entries 4±6), neopentanol ꢀentries 7±9), cyclohexylmetha-
nol ꢀentries 10, 11) with methacrylonitrile, crotononitrile
and cis-2-pentenenitrile ꢀTable 3). Clearly, the results of
Table 3 are completely in agreement with our expectations
and observations detailed in the preceeding paragraph. We
take note of the following features:
0
00
Placing bulky substituents on the b-center of the acrylic
nitrile create steric inhibition and cause considerable rate
penalty at ambient pressure. In fact, such encumbered
nitriles exhibit only marginal activity ꢀentries 2, 5, 11)
or are fully refractory to addition ꢀentries 3, 6, 8, 9). On
the other hand, methacrylonitrile shows some reactivity
in these reactions ꢀentries 1, 4, 10) even with the crowded
neopentanol ꢀentry 7).
At last, with the less reactive bulky alcohols, the Baylis±
Hillman dimer of crotononitrile was also formed as
described in one of our former papers. The occurrence
9
of this reaction largely depends on the feasibility of the
corresponding cyanoalkylation. For example, the dimeri-
zation extent at 300 MPa is respectively 4% in entries 5
and 8, up to 29% on entry 11.
Pressure has a bene®cial effect in all reactions. The most
Table 4. Cyanoalkylation of secondary and tertiary alcohols ꢀEq. ꢀ3)). Effect of pressure
0
00
Entry
R
R
R
R
1
R
2
Time ꢀh)
Base
Yields ꢀ%)
300 MPa
0
.1 MPa
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
CH
3
3
3
3
3
3
3
3
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
3
3
3
3
3
3
3
3
3
3
H
H
H
H
CH
CH
CH
H
H
H
CH
H
CH
H
H
CH
H
H
H
H
H
H
24
24
24
20
20
24
20
20
20
20
24
22
22
24
24
24
24
24
24
24
DMAP
PꢀC
DMAP
PꢀC
Phosphazene
DMAP
0
90
0
3
78
0
0
0
2
0
0
0
0
16
7
0
0
0
0
81
100
0
CH
CH
CH
CH
CH
CH
CH
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
CH
CH
4
H
9
)
3
3
3
3
3
4
H
9
)
56
98
0
3
a
65
3
PꢀC
PꢀC
PꢀC
PꢀC
PꢀC
PꢀC
4
H
4
H
4
H
4
H
4
H
4
H
9
9
9
9
9
9
)
)
)
)
)
)
3
3
3
3
3
3
C
2
H
5
22
35
C
C
2
H
H
5
5
H
CH
H
CH
CH
H
CH
CH
CH
H
3
3
a
38
2
3
iC
3
iC
3
iC
3
H
H
H
7
7
7
iC
3
iC
3
iC
3
H
H
H
7
7
7
2
15
a
3
a
3
Phosphazene
27
47
C
C
6
H
H
5
CH
CH
3
3
PꢀC
PꢀC
PꢀC
4
H
4
H
4
H
9
9
9
)
)
)
3
3
3
a
6
5
3
3
3
3
54
24
a
Cyclohexyl
Cyclohexyl
CH
Phosphazene
PꢀC
Phosphazene
PꢀC
Complexmi xt ure
CH
CH
3
3
3
3
3
CH
CH
CH
3
3
3
4
H
9
)
3
10
15
0
CH
CH
H
H
C
2
H
5
4
H
9
)
3
0
Nitrile ꢀ1.8 mmol), solvent ꢀROH), tri-n-butylphosphine ꢀ0.18 mmol), T ꢀ323 K).
The Baylis±Hillman dimer of crotononitrile is also formed.
a

4314
G. Jenner / Tetrahedron 58 ꢀ2002) 4311±4317
Table 5. Addition of allylcyanide to alcohols ꢀstandard conditions)
19). A striking example highlighting the reactivity dif-
ference between methacrylonitrile and crotononitrile
under pressure is disclosed in entries 22 and 23. A 2 vs
15% yield is obtained at 300 MPa in the addition of
diisopropylalcohol, a highly encumbered alcohol, to
methacrylonitrile and crotononitrile respectively. This
difference is noticeably enhanced under 700 MPa: the
yields of cyanoethers are 4 and 54%, respectively.
Nitrile
Alcohol
Yields ꢀ%)
.1 MPa 300 MPa
0
Allyl cyanide
Crotononitrile
Allyl cyanide
Crotononitrile
Allyl cyanide
Crotononitrile
Ethanol
Ethanol
Propanol
Propanol
Isopropanol
Isopropanol
40 ꢀ4 h)
32 ꢀ3.5 h)
26 ꢀ6 h)
5 ꢀ4 h)
4 ꢀ6 h)
4 ꢀ8 h)
100 ꢀ4 h)
99 ꢀ3.5 h)
100 ꢀ6 h)
94 ꢀ4 h)
80 ꢀ6 h)
81 ꢀ8 h)
Tertiary alcohols are relatively inert even under pressure.
Only tert-butanol is able to deliver by combination with
acrylonitrile the cyanoether albeit in very modest yields
ꢀentries 29, 30). It does not react with methyl substituted
acrylic nitriles. tert-Amylalcohol is unreactive ꢀentry 31).
2.4. Addition of secondary and tertiary alcohols Table 4,
Eq. 3))
2
.5. Addition of alcohols to allyl cyanide
A cursory examination of Table 4 indicates that secondary
and tertiary alcohols behave like primary alcohols in their
addition to acrylic nitriles. Most of the reactions listed do
not proceed at all at ambient pressure. Again, the crotono-
nitrile dimer is produced when hindered alcohols are
involved as reactants ꢀ13% in entry 18, 18% in entry 21).
With highly hindered alcohols the dimer is the major
product ꢀ45% in entry 24).
In place of crotononitrile, we examined the possibility of
adding its isomer, allyl cyanide, to alcohols. The allylic
nitrile should not show any particular reactivity. However,
we observed that it reacted with alcohols in the presence of
tributylphosphine like other acrylic nitriles, apparently at
the same rate as crotononitrile leading to similar yields
under identical conditions. The resulting cyanoethers
showed the same structures as those derived from crotono-
nitrile ꢀTable 5). In fact, we suspected isomerization of allyl
cyanide prior to nucleophilic addition. This was veri®ed
in independent runs demonstrating that in the presence of
tributyl phosphine allyl cyanide was quantitatively
converted to crotononitrile ꢀ508C, 15 h) whereas under
similar conditions, crotononitrile remained unaffected.
ꢀ3†
Among the three bases used in the runs, the strong basic
phosphazene leads to increased yields ꢀabout twice the
yields obtained when tributylphosphine serves as base).
Electrostriction is addressed with entry 12. Isopropanol
does not add to acrylonitrile in the presence of DMAP at
ambient pressure ꢀno formation of the isopropoxy anion)
whereas the reaction proceeds to 81% at 300 MPa.
However, DMAP as catalyst is unable to promote any
reaction with substituted acrylic nitriles even at
2.6. Kinetic studies and interpretation
The mechanism of the phosphine-catalyzed addition is
different to the mechanism outlined in Scheme 1. It can be
rationalized by the proposed mechanism depicted in
3
00 MPa ꢀentries 14, 17).
7
,10
Pressure is a powerful activation mode in promoting the
nucleophilic addition, particularly with acrylic nitriles in
which R ˆCH ꢀentries 18, 21, 27) and R ˆC H ꢀentry
Scheme 2.
Pressure may affect each step. As yields are
strongly dependent on steric demand both from the nitrile
and the alcohol side, the rate determining step would be
1
3
1
2
5
Scheme 2.

G. Jenner / Tetrahedron 58 ꢀ2002) 4311±4317
4315
Figure 1. Addition of linear primary alcohols CH
T is 303 K).
3
2
ꢀCH )
n
OH to nitriles. Dependence of rate constant ratio on n ꢀk100 is constant at 100 MPa and k
0
at 0.1 MPa,
deprotonation of the alcohol and addition of the alkoxy
moiety on the b carbon with concomittant elimination of
the phosphine.
the pressure effect on the reactivity difference between both
acrylic nitriles in their addition to such alcohols. Visibly, the
reactions involving crotononitrile are con®rmed to be more
pressure sensitive than those involving the other isomer.
With lengthening of the chain of the alcohol ꢀincreasing
of n), both reaction series become also more pressure
dependent, particularly the crotononitrile reactions.
The pressure effect on rate constants yields the volume of
activation DV . DV is usually a composite quantity and
±
±
1
1
must be written as:
X DVi±
±
±
DV
ˆ
The values of DV extracted from experimental kinetic data
are listed in Table 6 which embodies also some DV -values
R
The intrinsic volume change for the associative nucleophilic
addition process relates basically to the formation of the
O±C bond. This geometric modi®cation during the pro-
ꢀreaction volume based on partial molar volumes cf.
experimental part). These DV -values are all lower than
±
3 21
20 cm mol which is the typical average value for the
2
formation of one covalent bond in a late transition state.
1
2
gression towards the transition state is characterized by
a structural activation volume DV
±
of about 215 to
0
This is corroborated by the DV -values which do not show
R
3
21
2
20 cm mol as it was shown in the Bu NF-catalyzed
4
very signi®cant variations regardless of the type of either the
alcohol or the nitrile. It is, therefore, necessary to consider
additional volume terms in Eq. ꢀ1) especially for reactions
5
Michael addition of nitromethane to methyl vinyl ketone.
However, the whole body of our results is, at the evidence,
not uniquely traceable to bond-making effects. In fact, a
complete description of the kinetic behavior would entail
an analysis of several factors. To this respect, we were
prompted to determine kinetic rate constants as a function
of pressure. This was done by measuring the pseudo-®rst-
order constants k in the addition of primary linear alcohols
±
featured by the most negative DV -values. Such values refer
to considerably pressure dependent reactions which, to our
knowledge, have no precedence ꢀin the Baylis±Hillman
±
13
reaction DV have slightly higher values ). We must
concede that such negative values are somewhat enigmatic,
even beyond reasonable comprehension. They certainly
cloud the occurrence of other phenomena which will be
examined now.
ꢀC ±C ) to methacrylonitrile and crotononitrile, respec-
2 5
tively, in the pressure range ꢀ0.1±150 MPa). Fig. 1 pictures
Table 6. Apparent resulting activation volumes for the addition of primary
3
21
ꢀa) The transition state in these cyanoalkylations must be
highly dipolar ꢀScheme 2) and charge pair generation should
dominate so that bond making represents only a minor
contribution. The pressure effect on ionogenic reactions,
generally, depends largely on the medium since it is directly
linear alcohols to methyl acrylic nitriles ꢀvolumes are given in cm mol
)
ROH
Methacrylonitrile
Crotononitrile
±
±
DV
DV
R
DV
DV
R
1
4
Ethanol
240
240
245
252
218.5
216.5
219.0
nd
252
256
268
275
215.5
215.5
217.0
nd
related to the pressure effect on its dielectric constant
resulting in a volume of electrostriction. Since this phenom-
enon can be critical to rate enhancement under pressure,
1-Propanol
1-Butanol
1-Pentanol
15
we examined the kinetic solvent effect ꢀTable 7). It must be
emphasized that introducing an apolar medium depressed
the yields.
±
Estimated precision for DV ꢀ10%) ꢀdetermined at Tˆ303 K) and
3 21
0.5 cm mol for DV
^
R
ꢀdetermined at Tˆ298 K).

4316
G. Jenner / Tetrahedron 58 ꢀ2002) 4311±4317
Table 7. Solvent effect in the addition of 1-propanol to crotononitrile
Tˆ303 K)
the primary steric effect of a compressive group R charac-
terizing the degree of accessibility of reaction centers due to
R and the secondary steric effect involving the moderation
ꢀ
2
6
10 k ꢀs
21
±
3
DV ꢀcm mol )
21 a
Solvent
d
)
2
0
of a polar or resonance effect by non-bonded compression.
Tetrahydrofuran
Dichloromethane
Acetonitrile
1-Propanol
Formamide
83
0.92
1.47
4.02
18
246
246
250
256
253
Steric hindrance to ionization is often brought up to
rationalize the kinetic alteration. If approach to ionic sites
is hindered, solvation might suffer interference and DV is
104
141
142
369
±
expected to depend on electrostriction. There is a known
example, namely the Menshutkin reaction which is a iono-
genic reaction. It was found that the reaction can be very
sensitive to pressure as both electrostriction and steric
17
We checked the absence of reaction between the solvent and the acrylic
nitrile ꢀexcept obviously with propanol).
Estimated precision ꢀ10%).
a
21
effects are involved.
The results of Table 7 indicate a relatively moderate depen-
dence of the rate constant on solvent polarity ꢀd : cohesion
This is a very interesting facet of the study. It is in relation
with a yet largely undeciphered phenomenon. The pressure
effect on the rate of sterically hindered reactions has
revealed an apparently strange aspect of piezochemistry i±n
the sense that the volume of activation decreases ꢀe.g. DV
is more negative) with increasing steric congestion of the
2
energy density). The k-values differ by a factor of 20
from the least polar medium to formamide. However, the
DV -values in polar as well as in apolar media are not
±
3
21
altered signi®cantly ꢀaverage value: 250 cm mol ). The
result is at variance with the usual trend e.g. a decrease of
electrostrictive effects with increasing polarity of the
2
2
transition state. In other words, apart from electrostatic
considerations, steric hindrance would shift the transition
state towards products. This could be viewed as a conse-
1
4,15
medium.
The result contrasts with other Michael like
reactions such as the addition of amines to nitriles which
2
3
quence of the Hammond postulate or its corollar that for a
given type of reaction, increased exothermicities imply
earlier transition states or the reverse, hindered ꢀendergonic)
reactions correspond to later transition states ꢀfor a more
detailed comment on energy vs volume considerations,
1
6
do not need any added base. In this case, the formation of
zwitterionic species is assisted by pressure producing
variable DV -values comparable to those shown in Table
ꢀ,255 cm mol ) in non polar media, but higher in the
most polar solvents ꢀ,225 cm mol ). We have no
rational explanation for the dichotomy relative to the
±
3
21
7
3 21 15
24
see Ref. 21). In a comprehensive paper, we proposed to
take into account a steric component of DV termed DV
s
±
±
±
solvent dependence of DV . An earlier paper reported the
solvent effect on Michael reactions involving acrylates and
which would acknowledge the progression of the transition
state along the reaction axis due to steric congestion. In the
present case, steric interactions are manifested differently
owing to the probable lateness of the transition state.
Pressure helps the process by promoting the formation of
the alkoxy carbanion ꢀfor the most hindered alcohols) and
by facilitating the steric accessibility of the reaction center
of the nitrile. The present results con®rm the magni®ed
pressure effect on steric hindrance to ionization and can
1
7
aminoalcohols at ambient pressure. The authors suggested
that the solvent effect might be due to a change of nucleo-
philicity of amine and alcohols according to the solvent.
Considering Scheme 2, pressure should provide strong
assistance to deprotonation of the alcohol with formation
of the alkoxy carbanion with subsequent addition on the b
carbon. Both events should be sensitive to the bulkiness of R
and the steric accessibility of the b carbon. As the yields are
considerably depressed at 0.1 MPa with increasing steric
congestion we are inclined to suspect steric inhibition to
ionization even for polar media, which is relieved by
application of pressure.
2
2,25,26
be framed in the context of our earlier papers.
However, at this stage it is evident that the phenomenon
requires further investigation.
±
3. Conclusion
ꢀ
b) The DV -values illustrate what appears to be a manifes-
tation of steric congestion. In these cyanoalkylations steric
effects are manifested in the blockage to the attack of
the nucleophile on the b carbon of the nitrile. More sur-
prisingly, whereas the degree of branching and substitution
of the alcohol may rationalize to some extent the yields
obtained, the size of the alkyl group in primary alcohols
seems also to be a critical parameter ꢀFig. 1 and Table 6).
In fact, the kinetic effect due to steric hindrance in classic
nucleophilic processes is known. During the attack of the
unsaturated nitrile by a nucleophile, steric effects increase
around the attacked carbon atom and, whence, the steric
compression of substituents in the transition state which
is, therefore, destabilized. The activation energy is
increased and, accordingly, the rate constant decreases.
The phosphine-catalyzed nucleophilic addition of alcohols
to acrylic nitriles is a fairly pressure dependent reaction. We
propose to view the accelerative effect of pressure in
cyanoalkylations of alcohols as the result of
±
ꢀ
2
ꢀ
a) geometric factors ꢀbond formation) ꢀDV ,215 to
3 21
20 cm mol )
b) electrostatic interactions. This contribution is
1
8
seemingly largely dependent on the total steric con-
straints including steric hindrance to ionization.
Considering these effects, the reactions become highly
pressure sensitive when both reactants harbour sterically
compressive groups. We feel that our ®ndings are of utmost
importance for the synthesis of hindered molecules. To this
respect, pressure activation appears as a highly valid tactic
as we reported previously.
Steric effects can be described by number of empirical
LFER relations. The most known is the Taft±Hammett
equation. It is common practice to distinguish between
1
9

G. Jenner / Tetrahedron 58 ꢀ2002) 4311±4317
4317
4
. Experimental
References
Experiments were made with commercially available
chemicals used as received. In a typical run, 1,2,3-tri-
methoxybenzene ꢀinternal standard) was weighed and
placed in a ¯exible PTFE tube ꢀ1 mL). The tube was ®lled
half of its volume with the considered alcohol. Tri-n-butyl-
phosphine ꢀ0.18 mmol) was introduced via syringe followed
by the nitrile ꢀ1.8 mmol). The volume was then completed
with the alcohol. The tube was placed in the high pressure
vessel and pressurized to 300 MPa for the desired time.
After reaction pressure was released. Diethyl ether was
added to the mixture which was washed with a 2N HCl
1. Butskus, P. F. Russ. Chem. Rev. 1961, 30, 583±598.
2. Cited by Bayer, O. Angew. Chem. 1949, 61, 229±241.
3. Burson, H. A. Org. React. 1949, 5, 79±135.
4. Khumbar, P. S.; Sanchez-Valente, J.; Figueras, F. J. Chem.
Soc., Chem. Commun. 1998, 1091±1092.
5. Jenner, G. New J. Chem. 1999, 23, 525±529.
6. Jenner, G. Tetrahedron Lett. 2001, 42, 4807±4810.
7. Tertiary phosphines were used in the Michael reaction of
2-nitropropane to activated ole®ns including both acrylic
nitriles: White, D. A.; Baizer, M. M. Tetrahedron Lett.
1973, 14, 3597±3600.
solution, an aqueous NaHCO solution and water ꢀtwice)
3
8. Schwesinger, R.; Schlemper, H. Angew. Chem., Int. Ed. 1987,
26, 1167±1169. Schwesinger, R.; Hasenfratz, C.; Schlemper,
H.; Walz, L.; Peters, E. M.; Peters, K.; von Schnering, H. G.
Angew. Chem., Int. Ed. 1993, 32, 1361±1363.
successively. The organic fractions were collected and
volatile compounds removed on a rotary evaporator. The
residue was analyzed by H NMR spectroscopy ꢀBrucker
1
AC 200) and the yield determined from the relative inten-
sities of characteristic protons of the cyanoether vs methoxy
groups of the internal standard. For analytical data, the
cyanoether was puri®ed by chromatography. We report
9. Jenner, G. Tetrahedron Lett. 2000, 41, 3091±3094.
10. Trost, B. M.; Dake, G. R. J. Org. Chem. 1997, 62, 5670±5671.
11. Jenner, G. In High Pressure Molecular Science; Winter, R.,
Jonas, J., Eds.; Kluwer Academic: Dordrecht, 1999; pp 291±
311.
1
some selected H NMR data for unprecedented cyano-
alcohols ꢀCDCl , 200 MHz, d ppm):
12. le Noble, W. J.; Kelm, H. Angew. Chem., Int. Ed. Engl. 1980,
1
3
9, 841±946.
13. Hill, J. S.; Isaacs, N. S. Tetrahedron Lett. 1986, 27, 5007±
5010.
cyanoether from methacrylonitrile and neopentanol
ꢀ
entry 7): 3.47 ꢀbr, 2H, OCH ), 3.05 ꢀs, 2H, OCH ),
2 2
2.77 ꢀm, 1H, CHCN), 1.25 ꢀd, 3H, CH ), 0.85 ꢀs, 9H,
3
CH3)
14. Fleischmann, F. K.; Kelm, H. Tetrahedron Lett. 1973, 14,
3773±3776.
cyanoether from crotononitrile and neopentanol ꢀentry 8):
15. Jenner, G. J. Phys. Org. Chem. 1999, 12, 619±625.
16. Jenner, G. New J. Chem. 1995, 19, 173±178.
17. Samui, K.; Ogata, N. Bull. Chem. Soc. Jpn 1968, 41, 1968±
1970.
3
2
.58 ꢀbr, 1H, OCH), 2.99±3.10 ꢀm, 2H, OCH ), 2.43 ꢀbr,
2
H, CH CN), 1.20 ꢀd, 3H, CH ), 0.85 ꢀs, 9H, CH )
3
2
3
cyanoether from crotononitrile and isopropanol ꢀentry
1
2
8): 3.70 ꢀm, 1H, OCH), 3.60 ꢀm, 1H, OCH), 2.40 ꢀd,
H, CH CN), 1.20 ꢀd, 3H, CH ), 1.09 ꢀd, 6H, CH )
18. Br uÈ ckner, R. Reaktionmechanismen.Organische Reaktionen,
Stereochemie, moderne Synthesemethoden; Spektrum
Akademischer, 1996.
2
3
3
cyanoether from crotononitrile and 2-butanol ꢀentry 21):
.73 ꢀm, 1H, OCH), 3.38 ꢀm, 1H, OCH), 2.42 ꢀbr, 2H,
CH CN), 1.38 ꢀbr, 2H, CH ), 1.22 ꢀd, 3H, CH ), 1.10 ꢀd,
3
19. Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165±
195.
20. Shorter, J. Quart. Rev. 1970, 24, 433±453.
2
2
3
3
H, CH ), 0.88 ꢀt, 3H, CH ).
3
3
21. le Noble, W. J.; Ogo, Y. Tetrahedron 1970, 26, 4119±4124.
22. Jenner, G.; Kellou, M. Tetrahedron 1981, 37, 1153±1160.
23. le Noble, W. J.; Asano, T. J. Am. Chem. Soc. 1975, 97, 1778±
Kinetic experiments were carried out at 303 K with 1 mmol
nitrile and 0.1 mmol phosphine in 2.5 mL alcohol according
1
5
to the method previously described. Activation volumes
were determined mathematically from the initial slope
of the smoothed kinetic curve ln kˆf ꢀP). The values
were compared to those calculated by means of El'yanov's
1
782.
4. Jenner, G. J. Chem. Soc., Faraday Trans. 1 1985, 81, 2437±
460.
2
2
2
2
5. Jenner, G. High Press. Res. 1992, 11, 21±32.
6. Jenner, G.; Kellou, M. J. Chem. Soc., Chem. Commun. 1979,
2
7
procedure.
8
51±852. Jenner, G. Tetrahedron Lett. 2001, 42, 243±245.
Partial molar volumes serving to calculate DVR were
determined at 25.08C in the given alcohol with a digital
densimeter ꢀParr DMA 602). For full details about the
determination of partial molar volumes see Ref. 28.
Jenner, G. Tetrahedron Lett. 2002, 43, 1235±1238.
7. El'yanov, B. S.; Gonikberg, E. M. J. Chem. Soc., Faraday
Trans. 2 1979, 75, 172±191.
2
2
8. Jenner, G. Tetrahedron 1998, 54, 2771±2776.
Acknowledgements
We thank Dr B. Michels for providing access to the digital
densimeter.