Regiochemistry in alkylation, acylation and methoxycarbonylation of alkali salts from 2-substituted alkenylpropanedinitriles

Article abstract of DOI:10.1039/b009175h

Alkali salts (4^-) of several 2-substituted alkenylpropanedinitriles are prepared, isolated and characterised by ¹H and ¹³C NMR spectroscopy. By treating the salt 4a^- with 4a a dimer, 2-dicyanomethylene-6-methyl-4,6-diphenyl-1,2,5,6-tetrahydronicotinonitrile 6, is formed, for which a single-crystal X-ray structure is presented. Alkylation of the ambident salts 4^- with MeX (X = I, Tf, or Br) leads to regioisomers, 1-and 3-alkylation, and comparison with O vs. C-alkylation of ketones is made. Double alkylation is also observed, 3-alkylation followed by 1-alkylation. Thus from 4a^- a mixture of (E)-and (Z)-2-methyl-2-(1-phenylprop-1-enyl)propanedinitrile 9b is formed, and the X-ray structure of the former is presented. Acylation of 4a^- with benzoyl chloride gives only the 3-regioisomer; using equimolar amounts of reactants gives a ring-closure product, 2-oxo-4,6-diphenyl-2H-pyran-3-carbonitrile 12, while excess of benzoyl chloride gives the double-acylation product, (Z)-4,4-dicyano-1,3-diphenylbuta-1,3-dienyl benzoate (Z)-13, for which the X-ray structure is presented. Acylation of 4a^- with acetyl chloride gives both regioisomers; the 1-acylated product evaded isolation, and the 3-acylation product reacted further to give the (E)- and (Z)-form of 4,4-dicyano-1-methyl-3-phenylbuta-1,3-dienyl acetate 15. The X-ray structure of the latter isomer, (Z)-15, is presented. Methoxycarbonylation of salts 4^- with methyl chloroformate gives both regioisomers, while use of methyl cyanoformate gives only the 3-regioisomer, in addition to a Michael adduct from reaction of the CN^- group with protonated salt 4^-.

Full text of DOI:10.1039/b009175h

View Article Online / Journal Homepage / Table of Contents for this issue
Regiochemistry in alkylation, acylation and methoxycarbonylation
of alkali salts from 2-substituted alkenylpropanedinitriles
Hege Karlsen,a Pål H. Songe,b Lise Kristin Sunsby,a Lise Cathrine Hagen,a Per Kolsaker*a and
Christian Rømminga
1
a Department of Chemistry, University of Oslo, P.O. Box 1033 Blindern, N-0315 Oslo, Norway
b Institute for Energy Technology, P.O. Box 40, N-2027 Kjeller, Norway
Received (in Cambridge, UK) 8th November 2000, Accepted 22nd December 2000
First published as an Advance Article on the web 16th February 2001
Alkali salts (4Ϫ) of several 2-substituted alkenylpropanedinitriles are prepared, isolated and characterised by 1H
and 13C NMR spectroscopy. By treating the salt 4aϪ with 4a a dimer, 2-dicyanomethylene-6-methyl-4,6-diphenyl-
1,2,5,6-tetrahydronicotinonitrile 6, is formed, for which a single-crystal X-ray structure is presented. Alkylation
of the ambident salts 4Ϫ with MeX (X = I, Tf, or Br) leads to regioisomers, 1- and 3-alkylation, and comparison
with O vs. C-alkylation of ketones is made. Double alkylation is also observed, 3-alkylation followed by 1-alkylation.
Thus from 4aϪ a mixture of (E)- and (Z)-2-methyl-2-(1-phenylprop-1-enyl)propanedinitrile 9b is formed, and the
X-ray structure of the former is presented. Acylation of 4aϪ with benzoyl chloride gives only the 3-regioisomer; using
equimolar amounts of reactants gives a ring-closure product, 2-oxo-4,6-diphenyl-2H-pyran-3-carbonitrile 12, while
excess of benzoyl chloride gives the double-acylation product, (Z)-4,4-dicyano-1,3-diphenylbuta-1,3-dienyl benzoate
(Z)-13, for which the X-ray structure is presented. Acylation of 4aϪ with acetyl chloride gives both regioisomers;
the 1-acylated product evaded isolation, and the 3-acylation product reacted further to give the (E)- and (Z)-form
of 4,4-dicyano-1-methyl-3-phenylbuta-1,3-dienyl acetate 15. The X-ray structure of the latter isomer, (Z)-15, is
presented. Methoxycarbonylation of salts 4Ϫ with methyl chloroformate gives both regioisomers, while use of methyl
cyanoformate gives only the 3-regioisomer, in addition to a Michael adduct from reaction of the CNϪ group with
protonated salt 4Ϫ.
characterised,5 bears obvious resemblance to the enolate salt of
acetophenone (5). The ambident nature of such salts should in
principle lead to different products when they react with
electrophiles. In base-promoted reactions of ketones where
metal enolates are considered to be intermediates, they are
defined as O- or C-products. In our system such terminology
is meaningless, thus we prefer to call the analogous products 1-
Introduction
Our observation of ring–chain tautomerism in 2-(2,2-dicyano-
1-methylethyl)benzoic acid1 1a, similar to the behavior of
2-acetylbenzoic acid 1b in solution,2 supports the qualitative
similarity between an oxygen atom and the C(CN)2 group
expressed by K. Wallenfels et al. in their review article on ‘Die
O–C(CN)2-analogie’.3 Structural comparisons between 3-(N,N-
or 3-products, respectively.† The different reaction parameters
dialkylamino)propenones 2 and the analogously substituted
which influence the O/C-alkylation or -acylation ratio for
1,1-dicyanobuta-1,3-dienes 3 were made using dynamic 1H
ketones are discussed in extensive reviews.7 As we in the present
NMR, X-ray crystallography and quantum chemical calcu-
study on reactions of the salts 4Ϫ with electrophiles also experi-
lations.4 The latter compounds were synthesised in reactions
enced a competition between 1- and 3-products, it was of
interest to see how some of these parameters influenced the
competition.
between the potassium salt of (1-phenylethylidene)propane-
dinitrile 4a and appropriate amidinium salts,5 and the kinetics
of this reaction, involving a tetrahedral intermediate, were
studied.6
The above potassium salt 4aϪ, which could be isolated and
Results and discussion
Preparation of potassium salts
The general procedure for salt preparations has been
described,5 using solvents with the lowest polarity possible
in order to precipitate out the salt. Furthermore, to avoid
dimerisation, which occurs when the salt and the carbon acid
are both present at the same time (see below), the carbon
acid was added to the base, potassium tert-butoxide. This
1
procedure gave us in hand very pure salts (>95%, H NMR)
in excellent yields (>90%), which were used in the alkylation
studies (Scheme 1).
Unfortunately, it has not been possible, so far, to obtain
crystals suitable for single-crystal X-ray crystallography.
† The terms ‘1- or 3-products’ will only be used to indicate competition
in regiochemistry. In naming compounds IUPAC recommendations
will be used.
DOI: 10.1039/b009175h
J. Chem. Soc., Perkin Trans. 1, 2001, 497–507
This journal is © The Royal Society of Chemistry 2001
497

View Article Online
Table 1 Methylation of salts of 4 (4Ϫ) with MeXa
Expt
Compd
Solvent
X
Cation
4
7
8
9(E)
9(Z)
10
1-Me/3-Me
Remarks
1
2
3
4
5
6
7
8
9
4aϪ
4aϪ
4aϪ
4aϪ
4bϪ
4cϪ
4dϪ
4dϪ
4dϪ
4eϪ
4fϪ
CH3CN
DMSO
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
I
I
K
K
K
K
K
K
K
Li
Li
K
K
K
26
25
22
35
28
50
75
6
5
2
30
35
23
2
2
2
2
1
0.88
0.65
1.85
3.57
∞
Same results with 18-crown-6
Same results with 18-crown-6
Same results with 18-crown-6
Starting material: (E)/(Z) ≈ 75/25
Br
Tf
Tf
Tf
Tf
I
Tf
Tf
Tf
Tf
21
91
9
6
4
8
1
4
4
4
93
73
59
73
∞
17
10
16
3
10
1
2
4
3.32
2.46
4.29
∞
5.79
3.33
10
11
12
73
17
81
70
14
21
4gϪ
a Yields in percent.
Scheme 2
Alkylation of the salts
Fig. 1 ORTEP plot of 2-dicyanomethylene-6-methyl-4,6-diphenyl-
1,2,5,6-tetrahydronicotinonitrile 6. (For clarity, a molecule of acetone
present in the unit cell has been removed.)
The alkylation of the different 2-substituted alkylidenepropane-
dinitriles 4 was done by first preparing the pure salts 4Ϫ; fol-
lowed by reaction in selected solvents with electrophiles of the
type MeX. As can be seen from Scheme 3 and Table 1, double
alkylation is quite common, thus excess of MeX had to be used.
The doubly alkylated products and concurrently the re-
formation of 4 is a consequence of an acid–base equilibrium
involving 4Ϫ and the 3-alkylation product 8. The existence of
such an equilibrium was confirmed by mixing equimolar
amounts of 4aϪ and presynthesised 8a (= 4b), giving equal pro-
portions of (E)- and (Z)- salt 8aϪ, reflecting thermodynamic
equality. The alkylation of the salts 8aϪ is not an equilibrium
reaction, giving 9a with a ratio (E):(Z) ≈ 15:1, perhaps due to
different reactivity of the two salts 8aϪ. When the reaction
between 4aϪ and 8a was repeated in the presence of excess of
MeI, the (E):(Z) ratio of product 9a was 19:1. When salt 4bϪ
was made directly from 4b using potassium tert-butoxide, the
(E):(Z) ratio of salts 4bϪ was 91:9, and methylation with MeTf
gave 9b with (E):(Z) ≈ 91:9. The stereochemistry of 9b(E) was
established using single-crystal X-ray diffraction (Fig. 2).
Scheme 1
1
However, H NMR spectra of all salts (Experimental section)
indicate non-equivalence of the methylene protons making
resonance structure 4aϪ (ii) the more dominant one for describ-
ing the structure of the salts. A slight shift to higher field of the
methylene protons points to a somewhat increased electron
density at this site as compared with protons at standard sp2
carbons.
As mentioned above, to avoid dimerisation of 4, an excess of
base had to be used in salt preparations. When, intentionally,
equivalent amounts of 4a and its potassium salt were mixed
in acetonitrile, a dimer, 2-dicyanomethylene-6-methyl-4,6-di-
phenyl-1,2,5,6-tetrahydronicotinonitrile 6, could be isolated.
The formation of dimer 6 (Scheme 2) is initiated by Michael
addition of 4aϪ to the neutral carbon acid followed by a multi-
step ring closure as suggested.8 Since the spectroscopic param-
eters of this compound were not completely in line with those
reported, a single-crystal X-ray structure analysis was carried
out (Fig. 1), showing that their postulated structure was
correct.
One of the objectives of this study was to get some inform-
ation about the competition between 1- and 3-alkylation, and,
in principle, the product ratio 7:8 should be used as a measure.
However, the secondary reactions 8→9 and 8→10 are very fast
and, in fact, when only 0.5 equivalent of MeI was used as
electrophile in the reaction with 4aϪ, 21% of the total products
was 9a(E). The reason for this is probably (1) that the acid–base
equilibrium 4Ϫ ϩ 8
4 ϩ 8Ϫ is very quickly established,
and (2) that carbanion 8aϪ is more reactive than 4aϪ due to its
higher alkylation. Consequently, one must therefore classify
the secondary products 9 and 10 as ‘3-methylation’ products.
However, in the alkylation of 4bϪ and 4eϪ where the start-
ing materials already have a methyl group in the 3-position
(Scheme 3), products 9 must be considered as 1-methylation
products.
498
J. Chem. Soc., Perkin Trans. 1, 2001, 497–507

View Article Online
Scheme 3
specific K-complexing agent,14 does not have any effect on the
1-Me:3-Me ratio. Changing the cation to Liϩ does not have
a significant effect on the 1-Me:3-Me ratio (Expt. 7 vs. 9). All
these findings point to a looser Coulombic attraction between
the ions in our system. As most of the negative charge of the
anion is located on the carbon atom in the C(CN)2 group, cf.
the 1H NMR spectra of 4Ϫ, a steric effect exerted by the cyano
groups could result in a larger distance between the ions.
Another, and perhaps better, explanation could be that the
charge of the anion may be intrinsically stabilised by the induct-
ive effect of the cyano groups. Similar arguments were used to
explain the more efficient contribution of the C(CN)2 compared
with the C᎐O group in stabilising conjugated systems.4
᎐
Increasing alkylation at the α-position in ketones leads to
higher O:C ratios,15,16 results also obtained with our system
where only 1-alkylation was observed (Expts. 5 and 6 vs. 4 and
10 vs. 7). The high preference for O- or 1-alkylation in these
reactions may be a result of increased steric hindrance; how-
ever, another explanation could be of electronic nature. The
left-hand resonance structures of both 4Ϫ and enolate 5, see
above, should be energetically destabilised, while the right-
hand structures should be stabilised upon increased α-alkyl-
ation, increasing the ‘hardness’ of the oxygen atom or the
1-position of the anions in question.
In terms of the HSAB principle, in ambident nucleophiles
like enolates 5 the oxygen atom represents the ‘hard’ and the
carbon atom the ‘soft’ site, and consequently in 4Ϫ the C(CN)2
group should be the ‘hard’ site. We have never observed any
attack at the nitrogen atom of the cyano group, which should
be the consequence of increased electron density on the nitro-
gen if the cyano group is part of a conjugated system. We have
previously suggested that the cyano group stabilises a negative
charge on the carbon atom by its ϪI effect, based on the
bond lengths in the C(CN)2 group in some highly conjugated
4-amino-1,1-dicyanobuta-1,3-dienes.4 The absence of any
N-alkylated compounds in our study does not weaken this
proposal. The ‘hardness’ of the electrophile MeX is a function
of the ‘hardness’ of the leaving group X. In the alkylation
Fig. 2 ORTEP plot of (E)-2-methyl-2-(1-phenylprop-1-enyl)propane-
dinitrile (E)-9b.
As mentioned in the Introduction section, one of the aims
of this study was to see whether and how the different reaction
parameters influencing the O:C alkylation of ketones also play
a similar role in determining the regiochemistry in our system.
In general the alkylation of ambident nucleophiles is controlled
according to the principle of ‘hard’ and ‘soft’ acids and bases
(HSAB); ‘hard’ alkylation reagents attack the ‘hard’ site of the
nucleophile and vice versa.9.10 In alkylation of the enolates from
ketones the O:C ratio seems to be controlled by several reaction
parameters.11 An important factor is the solvent’s capacity to
assist the dissociation of the salt. The dipolar aprotic solvent
DMSO is considered to be a good cation solvator by its
capacity as an electron-pair donator (EPD).12 This is demon-
strated in the alkylation of the enolate of cyclohexanone where
the O:C ratio changes from 24:44 in 1,2-dimethoxyethane
(DME) to 71:18 in DMSO.13 In this study we do not find a
similar solvent effect, the 1-methylation:3-methylation ratio
(1-Me:3-Me) being 88:100 in acetonitrile and in DMSO
65:100. Moreover, the use of 18-crown-6, considered to be a
J. Chem. Soc., Perkin Trans. 1, 2001, 497–507
499

View Article Online
Table 2 Acylation of potassium salts 4Ϫ with RCOCla
Molar ratio
4:RCOCl
(A:B)
Order of
mixing
Compd
R
12
13
14
15(E)
15(Z)
4
Remarks
4aϪ
4aϪ
4aϪ
4aϪ
Ph
Ph
Me
Me
1:1
1:5
1:1
1:1
B to A
A to B
80
0
0
>80
49
0
6
6
19
17
26
77
NMR-expt.
14 dec. to 4
on chrom. column
B to A
A to B
4cϪ
Me
1:5
>80
a Yields in percent.
of the enolate of propiophenone with n-pentyl halides the
O/C-quotient is 1.2, 0.64 and 0.23, respectively, on going from
Cl to Br to I.16 We observed a similar trend in 1-Me/3-Me: Tf
3.57, Br 1.85, I 0.88 (expts. 4, 3, 1).
1-Methylation products
The base peak in the MS spectra of all 1-methylated products 7
and 9 (except 7g) is obviously an example of a fragmentation
route which is more governed by the stability of the radical and
not as usual by the fragment stability (Scheme 4).
Scheme 4
Fig. 3 ORTEP plot of (Z)-4,4-dicyano-1,3-diphenylbuta-1,3-dienyl
benzoate (Z)-13.
As discussed in the Experimental section, problems in
obtaining pure samples of the 1-methylated compounds were
encountered. As the 3-methylated compounds 8 or 10 contain
an acidic proton, Expt. 4 (Table 1) was repeated on a larger
scale and a diethyl ether solution of the products 7a and 8a
was extracted with dil. sodium hydroxide, resulting in a rapid
decomposition of the 1-methylated product 7a. A chromato-
graphically pure sample of 7a was therefore treated with potas-
sium tert-butoxide, for details see Experimental section, leading
to a mixture of (E)- and (Z)-methyl-3-phenylbut-2-ene nitrile
11 (Scheme 5). The reaction may be depicted as a nucleophilic
Acylation and methoxycarbonylation
It should be expected that acylation with acid chlorides and
methoxycarbonylation with α-substituted formates, in prin-
ciple, should follow similar reaction pathways as in alkylation
and that the regiochemical outcome should be governed by the
principles of the HSAB theory. However, this theory is based
on reactivity, a kinetic parameter, which implies that precise
application of this theory requires that the products formed
are stable under the reaction conditions. For the alkylation
reactions discussed above, this condition is considered to be
fulfilled.17 In contrast, in acylation and, probably to a lesser
degree, also methoxycarbonylation reactions it appears to be
difficult to achieve kinetic conditions. This is mainly due to the
observed fact that unless the applied electrophile is used in
very large excess, the O-acylation-product (1-acylation-product
in our system) itself can compete with the applied electro-
phile to give the thermodynamically more stable C-acylation
products (3-acylation-product here).18 We will therefore not try
to apply the HSAB theory in this part of the presentation, but
merely present the preparation of some new acylation and
methoxycarbonylation products in the reactions of our salts 4Ϫ
with selected electrophiles.
Scheme 5
attack of the alkoxide ion on one of the nitrile carbon atoms
with the rest of the molecule as the leaving group in a stereo-
electronic manner as shown in Scheme 6.
The outcome of acylations of 4Ϫ with acetyl and benzoyl
chloride is sketched out in Scheme 7 and Table 2.
Only 3-acylation products (or secondary products 12 from
them) were isolated with benzoyl chloride, even when using
excess of electrophile to avoid thermodynamic reaction condi-
tions, which only led to the double acylation product 13(Z)
instead of the cyclic product 12. A single-crystal X-ray struc-
ture of 13(Z) is presented in Fig. 3.
On the other hand, use of acetyl chloride led to substantial
yields of the 1-acylation product 14, which unfortunately
evades isolation since it decomposes on chromatographic
columns (SiO2) or in contact with aqueous solutions. The
Scheme 6
500
J. Chem. Soc., Perkin Trans. 1, 2001, 497–507

View Article Online
3-acylation product could not be isolated as it reacts further
with acetyl chloride, and when using excess of reagent followed
by aqueous work-up, these double-acylation products, 15(E)
and (Z), were the only products, together with substantial
amounts of starting material, partially formed according to the
lower equilibrium in Scheme 7 and mostly from the hydrolysis
of the 1-acylation product 14. The stereochemistry of 15(Z)
was established by single-crystal X-ray structure determination
(Fig. 4).
Increased alkylation at the 3-position of the reacting salt
(4cϪ) led to exclusive acylation in the 1-position to give 14c, a
result in line with alkylation reactions, see above.
The methoxycarbonylation reagents, methyl chloro- and
cyanoformate, also react differently with the salts 4Ϫ,
reflecting their different ‘hardness’ as electrophiles. The
chloride ion is the much better leaving group, making the
chloroformate the ‘harder’ acid in the HSAB sense and the
outcome of these reactions is summarised in Scheme 8 and
Table 3.
Double methoxycarbonylation products are not observed
with these reagents, and as the acidity of 3-product 17 probably
will be higher than that of starting material 4,‡ the equilibria
involving 17 and 4Ϫ will be shifted to the right. This property
has been utilised to separate the two methoxycarbonylation
products 16 and 17 (Experimental section).
Fig. 4 ORTEP plot of (Z)-4,4-dicyano-1-methyl-3-phenylbuta-1,3-
dienyl acetate (Z)-15.
‡ Estimated to have a pKa-value ≈11.5.19
Scheme 7
Scheme 8
J. Chem. Soc., Perkin Trans. 1, 2001, 497–507
501

View Article Online
Table 3 Methoxycarbonylation of potassium salts 4Ϫ with methyl chloro- and cyanoformatea
Molar ratio
Compd
X
4Ϫ :XCO2Me
16
17
4
18
4aϪ
4cϪ
4dϪ
4aϪ
4dϪ
Cl
Cl
Cl
CN
CN
1:5
1:5
1:5
1:5
1:5
40
>90
22
0
26
0
32
51
65
36
Traces
42
49
31
0
a Yields in percent.
<0.01%, stored under nitrogen over molecular sieves). All
experiments involving the salts were carried out in an inert
atmosphere (argon or nitrogen).
Materials
The 2-substituted ethylidenepropanedinitriles 4 were syn-
thesised in good yields from the corresponding ketones and
malononitrile using the Cope method:20 4a, mp 94–95 ЊC (from
aq. EtOH) (lit.,21 95 ЊC); 4b, mp 67–68 ЊC (from aq. EtOH)
(lit.,22 69–70 ЊC); 4c, mp 56–58 ЊC (from aq. EtOH) (lit.,22 60–
62 ЊC); 4d mp 69–70 ЊC (from aq. EtOH) (lit.,23 64–65 ЊC); 4e,
mp 36–38 ЊC (from CH2Cl2–pentane) (lit.,24 36–38 ЊC); 4f, bp
103–104 ЊC/14 mmHg (lit.,25 104 ЊC/11 mmHg). 2-[Cyclopropyl-
(phenyl)methylene]propanedinitrile 4g was synthesised with
reasonable yields (53%) from cyclopropyl phenyl ketone and
malononitrile using the Cope method.20 Mp 118–119 ЊC (from
aq. EtOH); νmax/cmϪ1 3022w, 2920m, 2223s, 1559s; δH(CDCl3)
0.8–0.9 (2 H, m), 1.3–1.4 (2 H, m), 2.4–2.5 (1 H, m), 7.2–
7.3 (2 H, m), 7.5–7.6 (3 H, m); δC(CDCl3) 10.4, (C-4, C-5),
19.5 (C-3), 85.0 (C-2), 112.1–112.5 (2 × CN), 127.2, 128.7,
130.4, 131.8 (Ar-C), 183.7 (C-1); MS (EI, 70 eV) m/z 194
Scheme 9
In the methoxycarbonylation reactions chloride or cyanide
ion are the leaving groups which in turn might react as
nucleophiles in Michael reactions with the protonated salts
formed in the equilibria sketched out in Scheme 8 for 3-attacks.
Only cyanide ions react in this fashion as these ions are better
nucleophiles, but also because the reversibility of the analogous
chloride reaction should be more pronounced, due to the
much better leaving-group capacity of the latter ion. That this
Michael reaction is a secondary reaction was confirmed in the
following way: shortly after the reaction with methyl cyano-
formate was finished, a sample was withdrawn and analysed by
GLC, showing only small amounts of the Michael product
and substantial amounts of protonated salts, a situation which
changed rapidly to give the product composition given in
Table 3.
(71%, [M]ϩ ), 193 (100, [M Ϫ H]ϩ), 179 (74), 167 (74), 77 (23);
᝽
HRMS: Found: Mϩ, 194.08504. C13H10N2 requires M,
194.08440 (Ϫ3.3 ppm).
Other electrophiles
Preparations of potassium salts of 4
Bromine, considered to be a ‘soft’ acid,10 reacts accordingly
with 4aϪ, 4bϪ and 4cϪ to give exclusive attack at the 3-position.
On the other hand, a proton is a very ‘hard’ acid10 which pref-
erentially should give 1-attack, and, indeed, when an ethereal
To stirred, powdered potassium tert-butoxide suspended in
diethyl ether was added dropwise tert-butyl alcohol until a
homogenous solution was obtained. A compound 4, dissolved
in diethyl ether, was added dropwise while vigorous stirring
was maintained. In order to avoid reaction between salts
and starting material, see below, a small excess of potas-
sium tert-butoxide was always used. The corresponding
white precipitate (4Ϫ) was filtered off under nitrogen and
stored at Ϫ10 to Ϫ15 ЊC (argon). Yields >90%. Purity
solution of 4cϪ was stirred with aq. HCl, H NMR spectro-
1
scopy showed the presence of both 1- and 3-protonated
products (≈70% of 1-protonation) in the organic phase.
Attempts to purify the products led to tautomeric rearrange-
ment of the 1-isomer to the more stable product 3-isomer, 4c.
On the other hand, 4aϪ, unsubstituted at the 3-position, gave
only 3-protonation upon the same treatment.
1
>95% by H NMR. Spectroscopic parameters are entered in
Table 4.
Lithium salt of 4d
Experimental
General
To tert-BuLi (1.5 M; 7.3 ml, 11 mmol) dissolved in pentane was
added diethyl ether (15 ml); 2-(1-cyclopropylidene)propane-
dinitrile 4d (1.3 g, 10 mmol) was dissolved in diethyl ether (25
ml) and added dropwise at Ϫ15 ЊC. The precipitate formed was
collected (under nitrogen) after 5 min, washed with diethyl
ether and dried (SiO2) for 30 min in a desiccator (yield 1.1 g,
78%); δH(DMSO-d6) 0.3–0.4 (2 H, m), 0.5–0.6 (2 H, m), 1.1–1.3
(1 H, m), 3.6 (1 H, m), 3.8 (1 H, m).
Mps were measured on a Reichert Thermopan (Wien) appar-
atus and are uncorrected. IR spectra were recorded on a Nicolet
Magna 550 FTIR spectrometer using an attenuated total
reflectance (ATR) ZnSe plate for solid samples (unless other-
wise noted); GLC-FTIR experiments were accomplished using
a Varian 3300 gas chromatograph in front of the FTIR instru-
ment. High-resolution NMR spectra (1H and 13C) were
obtained using Bruker Spectrospin Avance DMX 200 and
DMX 300 spectrometers where the reference compound SiMe4
is software controlled; J-values are given in Hz. UV spectra are
recorded using a Shimadzu UV-260 spectrophotometer, and
mass spectra were obtained using a Fison Instrument VG
ProSpec Q. NMR peak assignments were done using 2D
spectroscopy or other suitable pulse programs.
Reaction between potassium salt 4a؊ and 2-(1-phenylethylidene)-
propanedinitrile 4a in acetonitrile
Salt 4aϪ (0.500 g, 2.45 mmol) and 4a (0.410 g, 2.45 mmol) were
dissolved in acetonitrile (25 ml) and the solution was stirred
overnight under argon. 1 M HCl saturated with NaCl was
added and the mixture extracted with dichloromethane
(DCM). The organic layer was dried (MgSO4), and evaporated
under reduced pressure. The crude product (0.87 g, yellow–
Solvents used for salt preparations and the alkylation studies
were commercial (FLUKA’s brand, dried to a water content
502
J. Chem. Soc., Perkin Trans. 1, 2001, 497–507

View Article Online
Table 4 NMR data for the potassium salts of compounds 4a
1H
Compd
R1
R2
R3
13Cb
4aϪ
7.2–7.3 (3 H, m), 7.3–7.4 (2 H, m)
7.1–7.4 (5 H, m)
4.36 (1 H, d, J 1.8)
1.45 (3 H, d, J 7.0)
1.84 (3 H, d, J 7.0)
1.53 (3 H, s)
4.08 (1 H, d, J 1,8)
4.91 (1 H, q, J 7.0)
4.83 (1 H, q, J 7.0)
1.87 (3 H, s)
93.4, 128.0, 128.4, 128.5, 143.5,
148.7
14.7, 101.6, 126.3, 127.6, 129.2,
129.7, 139.4, 140.7
14.7, 101.6, 126.3, 127.6, 129.2,
129.7, 139.4, 140.7
24.3, 24.6, 115.7, 126.5, 128.1,
131.1, 131.2, 144.6
4b(E)Ϫ
4b(Z)Ϫ
4cϪ
7.1–7.4 (5 H, m)
7.1–7.2 (5 H, m)
4dϪ
0.3–0.4 (2 H, m), 0.4–0.6 (2 H, m),
1.3–1.4 (1 H, m)
0.2–0.3 (2 H, m), 0.3–0.4 (2 H, m),
1.2–1.3 (1 H, m)
0.2–0.3 (2 H, m), 0.3–0.4 (2 H, m),
1.2–1.3 (1 H, m)
1.03 (6 H, d, J 6.8), 2.3–2.4 (1 H, m,
J 6.7)
3.90 (1 H, s)
3.60 (1 H, s)
9.6 (2 C), 15.0, 85.2, 127.5, 133.6,
147.2
N. S.
4e(E)Ϫ
4e(Z)Ϫ
4fϪ
1.62 (3 H, d, J 7.0)
1.66 (3 H, d, J 7.0)
4.02 (1 H, s)
4.28 (1 H, dq, J 0.9, 7.0)
4.69 (1 H, dq, J 0.9, 7.0)
3.83 (1 H, s)
N. S.
Dec.
Dec.
4gϪ
6.9–7.8 (5 H, m)
0.3–0.5 (2 H, m)
0.5–0.7 (2 H, m)
a Spectra recorded in acetone-d6. b N. S.: too many overlapping signals from the other isomer for assignment. Dec.: compound decomposes too
quickly in the available solvents.
brown solid) was washed with a small quantity of diethyl ether
to give a yellow solid (0.53 g, 70%). After recrystallisation,
0.47 g of yellow crystals of 2-dicyanomethylene-6-methyl-4,6-
subjected to column chromatography (SiO2). The 3-methylated
products, 8 and 10, except 8g (see below), are equivalent to
some of the starting materials; e.g. 8a = 4b, 8d = 4e and
10a = 4c.
diphenyl-1,2,5,6-tetrahydronicotinonitrile
6 were obtained,
mp 206–208 ЊC (from acetone–pentane) (lit.,8 208–210 ЊC); νmax
/
Unfortunately, we were not able to obtain all new products,
i.e. the 1-methylated compounds, in a state pure enough
for elemental analysis or single-crystal X-ray analysis. How-
ever, using GLC-FTIR the IR spectra of most components
were recorded. Mass spectra, including HRMS, could be
obtained for all compounds using GLC-MS. NMR assign-
cmϪ1 3265w (NH), 2216m (CN), 2198m (CN), 1594s, 1573m,
1545m; δH(CDCl3) 1.84 (3 H, s), 3.20 (1 H, d, J 18.1), 3.53 (1 H,
dd, J 18.1, 0.7), 6.63 (1 H, br s), 7.3–7.4 (2 H, m), 7.4–7.6 (8 H,
m); δH(acetone-d6) 1.91 (3 H, s), 3.46 (1 H, d, J 18.2), 3.82 (1 H,
d, J 18.2), 7.3–7.6 (10 H, m); δC(CDCl3) 29.3, 45.7, 54.2, 56.8,
102.0, 112.6, 112.7, 114.6, 124.4, 125.9, 127.9, 128.3, 128.7,
129.2, 129.5, 132.2, 135.5, 141.3, 157.2, 166.5; δC(acetone-d6)
45.8,52.6,58.0,58.2,102.9,114.0,114.6,115.5,125.9,128.7,129.1,
129.7, 129.8, 132.7, 137.1, 143.7, 158.7, 168.7; MS (EI, 70
1
ments, both for H and 13C, could be made using spectra of
chromatographic fractions with different product composi-
tions, coupled with software-controlled difference spectra
if one or more of the components could be isolated in a
pure state.
eV) m/z 336 ([M]ϩ , 58%), 321 ([M Ϫ CH3]ϩ, 100); HRMS:
᝽
Found: Mϩ, 336.136 173. Calc. for C22H16N4: M, 336.137 497
(Ϫ3.1 ppm); Found: Mϩ, 337.139 208. Calc. for 12C2113CH16N4:
M, 337.140 852 (4.9 ppm); Found: m/z, 206.071 997. Calc.
for C13H8N3: m/z 206.071 822 (Ϫ0.8 ppm); Found: m/z,
130.065 610. Calc. for C9H8N: m/z, 130.065 674 (0.5 ppm);
λmax[MeOH (log ε)] 206 (4.29), 249 (3.99), 325 (4.03) nm. Single-
crystal X-ray analysis, see below.
The following compounds have been isolated: 2-Methyl-2-
(1-phenylvinyl)propanedinitrile 7a, chromatographically pure,
colorless oil (mp < Ϫ15 ЊC); (E)-2-methyl-2-(1-phenylprop-1-
enyl)propanedinitrile (E)-9b, mp 32–33 ЊC (from acetone–
pentane), X-ray analysis: see below; 2-methyl-2-(2-methyl-1-
phenylprop-1-enyl)propanedinitrile 7c, chromatographically
pure, colorless oil; 2-(1-cyclopropylvinyl)-2-methylpropane-
dinitrile 7d, bp 219 ЊC (Found: C, 73.6; H, 7.2; N, 19.0. C9H10N2
requires C, 73.9; H, 7.0; N, 19.2%); 2-[cyclopropylidene(phenyl)-
methyl]-2-methylpropanedinitrile 7g, chromatographically
pure, colorless oil. Their spectroscopic parameters, obtained on
pure compounds, are entered in Tables 5 and 6.
The rest of the 1-methylated compounds; viz. (Z)-2-
methyl-2-(1-phenylprop-1-enyl)propanedinitrile (Z)-9b; (E)-2-
(1-cyclopropylprop-1-enyl)-2-methylpropanedinitrile (E)-9e;
(Z)-2-(1-cyclopropylprop-1-enyl)-2-methylpropanedinitrile
(Z)-9e; 2-(1-isopropylvinyl)-2-methylpropanedinitrile 7f were
not obtained in a pure state; their spectroscopic parameters
were recorded according to the principles described above and
the results are entered in Tables 5 and 6.
Alkylation of the salts 4؊
General. MeI. After purging of the selected solvent (5 ml)
with argon, a salt 4Ϫ (0.485 mmol) was added, and argon
purging was continued for 5 min. Iodomethane (1.45 mmol)
was added and the reaction mixture was stirred overnight at
room temperature in the dark. Diethyl ether (15 ml) was added
and the solution was washed three times with 1 M HCl satur-
ated with NaCl (5 ml). After drying of the organic phase
(MgSO4), evaporation left a yellow–brown oil. The oil was
1
analysed by H and 13C NMR, GLC-MS and GLC-FTIR to
establish the identity of the components, and GLC and 1H
NMR were used to establish the relative amounts of the
individual components, which are listed in Table 1.
2-[1-Methylcyclopropyl(phenyl)methylene]propanedinitrile
8g was not obtained pure and its spectroscopic parameters were
obtained in the same way as described for the 1-methylated
compounds above: GLC-FTIR: νmax/cmϪ1 3088s, 2974s, 2230m
(CN), 1671s, 1561m; δH(CDCl3) 0.9–1.0 (4 H, m, 2 × CH2),
1.53 (3 H, s), 7.3–7.6 (5 H, m, Ph); δC(CDCl3) 16.8, 23.4,
24.8, 86.3 (C-1), 112.5–113.3 (2 × CN), 128.2, 128.8, 132.0,
MeTf. Similar procedure as above was carried out, but as
initial experiments with this electrophile revealed that only tiny
amounts of doubly alkylated products were formed, a much
smaller excess (≈10%) of electrophile was used. Moreover, as
the reaction with this electrophile was considerably faster,
work-up as above was started after 1 h of reaction time.
Analytical procedures as above. The results are given in Table 1.
135.4 (Ph), 182.6 (C-2); MS (EI, 70 eV) m/z 208 ([M]ϩ , 15%),
᝽
207 ([M Ϫ H]ϩ, 40), 193 ([M Ϫ Me]ϩ, 100), 166 (61); HRMS
(GLC-MS) Found: Mϩ, 208.10005. C14H12N2 requires M,
208.09905 (4.8 ppm).
Identification of the products. The reaction products were
J. Chem. Soc., Perkin Trans. 1, 2001, 497–507
503

View Article Online
Table 5 HRMS of molecular ions of 1-methylated compounds 7 and 9
Compd
Mol. formula
Mass found
Mass calc.
Dev. (ppm)
Remarks
12
7a
C H10N2
182.084 637
183.087 609
196.100 116
197.103 756
196.100 936
197.102 045
210.115 209
211.118 387
146.084 382
160.100 820
181.088 83
182.084 398
193.087 753
196.100 048
197.103 403
196.100 048
197.103 403
210.115 699
211.119 053
146.084 398
160.100 048
181.089 15
Ϫ1.3
ϩ0.3
Ϫ0.3
Ϫ1.8
Ϫ4.5
ϩ6.9
ϩ2.3
ϩ3.2
ϩ0.1
Ϫ4.8
ϩ1.8
12
12
C
13CH10N2
11
12
12
12
12
12
12
9b(E)
9b(Z)
7c
C H12N2
13
C
13CH12N2
12
C H12N2
13
C
13CH12N2
12
C H14N2
14
C
13CH14N2
13
7d
9e(E)
7g
12C9H10N2
12
C H12N2
10
C H11N ([M Ϫ HCN]ϩ)
([M]ϩ too small to measure)
12
13
Table 6 Spectroscopic parameters of 1-methylated products 7 and 9a
1H NMR
13C NMR
FTIRb
MS
7a
1.86 (3 H, s), 5.50 (1 H, s), 5.91
(1 H, s), 7.4 (5 H, m)
25.2 (1-Me), 36.7 (C-1), 115.1
(2 × CN), 120.1 (C-3), 128.3,
128.5, 129.0, 135.4 (Ar C), 140.9
(C-2)
14.9 (3-Me), 25.4 (1-Me), 37.9
(C-1), 115.4 (2 × CN), 128.8
(C-3), 129.7, 129.9, 132.9 (Ar
C), 133.2 (C-2)
15.0 (3-Me), 25.8 (1-Me), 31.3
(C-1), 115.5 (2 × CN), 128.3
(C-3), 128.6, 128.8, 133.0 (Ar C),
133.9 (C-2)
22.5–24.9 (2 × 3-Me), 26.9
(1-Me), 32.1 (C-1), 115.8 (2 ×
CN), 125.8, 127.5, 129.1, 136.1
(Ar C), 128.1 (C-3), 140.5 (C-2)
7.3–12.4 (3 × Cyclopropyl C),
25.4 (1-Me), 37.6 (C-1), 112.3
(C-3), 115.2 (2 × CN), 142.3
(C-2)
2250w, 1495m, 1451m,
1444m
182 ([M]ϩ , 30%), 103 ([M Ϫ C(C-
᝽
N)2Me]ϩ, 100)
9b(E)
9b(Z)
7c
1.54 (3 H, d, J 6.8), 1.77 (3 H, s),
6.42 (1 H, q, J 6.8), 7.1–7.2 (2
H, m), 7.4–7.5 (3 H, m)
2243w, 1494m, 1451m,
1443m
196 ([M]ϩ , 39%), 181 ([M Ϫ Me]ϩ, 4),
᝽
117 ([M Ϫ C(CN)2Me]ϩ, 100)
1.78 (3 H, s), 2.13 (3 H, d, J 7.5),
5.95 (1 H, q, J 7.5), 7.1–7.2
(2 H, m), 7.3–7.4 (3 H, m)
N. S.
196 ([M]ϩ , 36%), 117 ([M Ϫ C(CN)2-
᝽
Me]ϩ, 100)
1.55 (3 H, s), 1.57 (3 H, s), 2.21
(3 H, s), 7.0–7.1 (2 H, m), 7.3–
7.4 (3 H, m)
2243m, 1649m, 1491s,
1451s, 1444s
210 ([M]ϩ , 61%), 195 ([M Ϫ Me]ϩ, 6),
᝽
131 ([M Ϫ C(CN)2Me]ϩ, 100)
7d
0.6–0.7 (2 H, m), 0.8–0.9 (2 H,
m), 1.4–1.5 (1 H, m), 1.96 (3 H,
s), 5.02 (1 H, d, J 1.1), 5.38 (1 H,
d, J 1.1)
0.7–0.9 (2 H, m), 0.8–0.9 (2 H,
m), 1.3–1.4, (1 H, m), 1.83 (3 H,
dd, J 1.9, 7.2), 1.96 (3 H, s), 6.11 115.9 (2 × CN), 129.9 (C-3),
(1 H, dq, J 1.9, 7.29)
0.4–0.5 (2 H, m), 0.7–0.8 (2 H,
m), 1.4–1.5, (1 H, m), 1.89 (3 H,
dd, J 1.9, 7.5), 2.01 (3 H, s), 5.70 115.3 (2 × CN), 127.6 (C-3),
3098m, 3016s, 2955w,
1645m
146 ([M]ϩ , 7%), 131 ([M Ϫ Me]ϩ, 9),
᝽
67 ([M Ϫ C(CN)2Me]ϩ, 100)
9e(E)
9e(Z)
7f
5.9, 9.3, 13.8 (3 × Cyclopropyl
C), 24.8 (1-Me), 36.9 (C-1),
160 ([M]ϩ , 12%), 145 ([M Ϫ Me]ϩ,
᝽
13), 118 (86), 81 ([M Ϫ C(CN)2Me]ϩ,
100)
130.6 (C-2)
5.5, 13.8, 15.8 (3 × Cyclopropyl
C), 24.1 (1-Me), 32.4 (C-1),
N. S.
N. S.
160 ([M]ϩ , 23%), 159 ([M Ϫ H]ϩ, 11),
᝽
145 ([M Ϫ Me]ϩ, 26), 81 ([MC(CN)2-
Me]ϩ, 99), 66 (100)
(1 H, J 1.9, 7.5)
131.8 (C-2)
1.19 (6 H, d, J 6.8), 1.90 (3 H, s),
2.4–2.5 (1 H, m, J 6.8), 5.32
(1 H, d, J 1.4), 5.55 (1 H, d,
J 1.4)
1.25 (2 H, dd, J 10.5, 9.0), 1.70
(2 H, J 10.5, 9.0), 1.96 (3 H, s),
7.3–7.4 (5 H, m)
24.2 (2 × Me), 25.8 (1-Me), 30.6
(2-CMe2), 37.5 (C-1), 113.9
(C-3), 115.2 (2 × CN), 147.9
(C-2)
2.9–6.3 (2 × Cyclopropyl C),
36.2 (C-1), 115.6 (2 × CN),
120.9 (C-3), 128.3, 128.4, 128.5,
130.7 (Ar C), 135.9 (C-2)
148 ([M]ϩ , 1%), 133 ([M Ϫ Me]ϩ, 9),
᝽
106 ([M Ϫ C(Me)2]ϩ, 28), 69 ([M Ϫ
C(CN)2Me]ϩ, 100)
7g
3059s, 2984s, 2247s,
1445s
207 ([M Ϫ H]ϩ, 3%), 181 (81), 166
(100), 128 ([M Ϫ C(CN)2Me]ϩ, 61)
a Numbering of the carbon atoms refers to Scheme 3, compound 7, cf. footnote †. b N. S. Not sufficient separation from other products to give
individual IR spectra.
Reaction between 2-methyl-2-(1-phenylvinyl)propanedinitrile 7a
and potassium tert-butoxide
[M]ϩ ), 156 (63, [M Ϫ H]ϩ), 142 (32, [M Ϫ Me]ϩ); HRMS:
᝽
Found: Mϩ, 157.088 886. Calc. for C11H11N: M, 157.089 149
(1.7 ppm); Found: Mϩ, 158.091 880. Calc. for 12C1013CH11N: M,
158.092 504 (3.9 ppm).
Potassium tert-butoxide (68 mg, 0.60 mmol) was suspended in
diethyl ether (5 ml) and tert-butyl alcohol was added dropwise
to obtain an almost homogeneous solution. A solution of com-
pound 7a (0.10 g, 0.55 mmol) in diethyl ether (5 ml) was added
dropwise to the stirred mixture (room temperature). After 5 min
diethyl ether (5 ml) was added, the reaction mixture was washed
three times with HCl saturated with NaCl (5 ml), dried with
MgSO4, and evaporated under reduced pressure to give a yellow
oil (96 mg). The crude product was purified through a silica gel
column with ethyl acetate–hexane (1:9). Fraction 1: Clear oil,
23 mg [(E)-2-methyl-3-phenylbut-2-enenitrile (E)-11, major
isomer]; δH(CDCl3) 1.82 (3 H, d, J 1.6), 2.34 (3 H, d, J 1.6), 7.1
(2 H, m), 7.3–7.5 (3 H, m); δC(CDCl3) 117.6, 24.7, 105.7, 119.8,
127.1, 128.3, 128.5, 139.2, 154.2; MS (EI, 70 eV) m/z 157 (100%,
Fraction 2: 0.44 g, mixture of starting material and
(Z)-2-methyl-3-phenylbut-2-enenitrile (Z)-11 (minor isomer).
Spectroscopic data for (Z)-11: GLC-FTIR: νmax/cmϪ1 2212s
(CN), 1493m, 1442m; NMR (difference spectra) δH(CDCl3)
2.05 (3 H, d, J 1.0), 2.15 (3 H, d, J 1.0), 7.2–7.4 (5 H, m);
δC(CDCl3) 17.5, 20.6, 105.2, 127.2, 128.4, 140.9, 154.3.
Acylation
Potassium salt 4a؊ and benzoyl chloride. Two modes of
reaction were run: One (1:1) where salt 4aϪ was kept in excess
throughout the reaction and one (1:5) where a large excess of
benzoyl chloride was used.
504
J. Chem. Soc., Perkin Trans. 1, 2001, 497–507

View Article Online
Table 7 Spectroscopic parameters for products from acylation and methoxycarbonylation of salts 4Ϫ a
1H NMR
13C NMR
FTIR
MS
12b
13c
6.98 (1 H, s), 7.5–7.6 (6 H, m),
7.7–7.8 (2 H, m), 7.9–8.0 (2 H,
m)
7.1–7.2 (1 H, m), 7.2–7.3 (2 H,
m), 7.32 (1 H, s), 7.4–7.5 (2 H,
m), 7.5–7.6 (5 H, m), 7.6–7.7
(3 H, m), 7.7–7.8 (2 H, m)
2.35 (3 H, s), 5.82 (1 H, d,
J 1.3), 6.04 (1 H, d, J 1.3),
7.3–7.5 (5 H, m)
1.49 (3 H, d, J 0.8), 2.18 (3 H,
s), 6.53 (1 H, q, J 0.8), 7.4–7.5
(5 H, m)
1.32 (3 H, s), 2.15 (3 H, d,
J 0.9), 6.36 (1 H, q, J 0.9),
7.3–7.4 (2 H, m), 7.4–7.5
(3 H, m)
3.82 (3 H, s), 5.74 (1 H, d,
J 1.0), 6.05 (1 H, d, J 1.0),
7.3–7.4 (5 H, m)
1.69 (3 H, s), 2.11 (3 H, s), 3.61
(3 H, s), 7.1–7.2 (2 H, m), 7.3–
7.4 (3 H, m)
0.6–0.7 (2 H, m), 0.8–0.9 (2 H,
m), 1.4–1.5 (1 H, m), 4.00
(3 H, s), 5.2 (1 H, m), 5.6 (1 H,
m)
3.68 (3 H, s), 3.95 (2 H, s), 7.4–
7.5 (5 H, m)
95.9, 103.7, 105.0, 126.9, 128.3, 129.6,
129.7, 130.4, 132.3, 133.0, 134.8, 159.4,
163.6, 164.4
85.2, 113.7, 113.8, 114.1, 127.2, 128.3,
129.0, 129.2, 129.5, 130.1, 130.7, 131.6,
132.4, 134.6, 135.0, 135.3, 158.7, 163.5,
168.9
2226w, 1726s, 1610m,
1513s, 1495m
273 ([M]ϩ , 96%), 245 (100)
᝽
2224s, 1742s, 1608s,
1575m, 1523m, 1449m
376 ([M]ϩ , 4%), 105 (100)
᝽
14a
24.9, 55.1, 111.2, 123.4, 127.7, 128.9,
129.8, 134.3, 136.9, 188.4
2163w, 2127w, 1767s
210 ([M]ϩ , 0.2%), 168 (4),
᝽
140 (8)
15a(E)
15a(Z)
19.5, 21.0, 83.5, 112.6, 113.1, 116.9,
128.7, 129.4, 132.1, 134.1, 162.6, 168.1,
169.0
19.7, 22.2, 83.6, 112.5, 112.9, 113.2,
127.9, 128.9, 130.9, 134.5, 160.8, 166.8,
168.1
2227w, 1782m, 1640m,
1200s, 1154s
252 ([M]ϩ , 1.0%), 237 (16),
᝽
183 (12), 140 (6), 43 (100)
2227w, 1783m, 1646m,
1194s, 1165s
252 ([M]ϩ , 2.0%), 237 (17),
᝽
183 (14), 140 (6), 43 (100)
16a
16c
16d
48.6, 55.7, 110.6, 122.9, 127.9, 128.7,
129.6, 134.5, 137.1, 175.5
1782s, 1228s
226 ([M]ϩ , 7%), 195 (1),
᝽
167 (12), 140 (35), 59 (100)
21.5, 24.2, 46.0, 55.2, 111.2, 123.4,
128.5, 128.6, 129.8, 136.8, 142.5, 160.9
254 ([M]ϩ , 27%), 209 (100),
᝽
195 (84), 59 (84)
7.4, 11.0, 13.1, 55.7, 110.7, 116.1,
138.7, 160.6
2256vw, 1781s, 1237s
190 ([M]ϩ , 0.5%), 150 (22),
᝽
105 (27), 59 (100)
17a
17d
42.1, 52.9, 87.9, 112.1, 112.2, 127.6,
129.2, 132.4, 134.4, 166.8, 169.8
10.8, 19.2, 35.3, 53.1, 87.2, 111.8,
111.9, 166.7, 176.6
2231m, 1741s, 1587m,
1566m
3017w, 2956w, 2230m,
1739s
226 ([M]ϩ , 100%), 195 (25),
᝽
140 (69), 59 (76)
1.0–1.1 (2 H, m), 1.3–1.4 (2 H,
m), 2.3–2.4 (1 H, m), 3.10 (2 H,
s), 3.70 (3 H, s)
2.15 (3 H, s), 5.47 (1 H, s), 7.5–
7.6 (3 H, m), 7.7–7.8 (2 H, m)
0.6–0.9 (4 H, m), 1.1–1.2
(1 H, m), 1.71 (3 H, s), 4.09
(1 H, s)
190 ([M]ϩ , 35%), 175 (10),
᝽
130 (91), 59 (100)
18a
18d
24.5, 35.8, 45.7, 111.7, 119.6, 127.0,
130.4, 130.9, 134.9
3.2, 4.0, 17.1, 23.2, 33.0, 43.5, 110.0,
110.1, 116.7
2253w, 1493m, 1449s
195 ([M]ϩ , 6%), 168 (14),
᝽
130 (100), 103 (29)
3016m, 2909s, 2259w,
2247w, 1463s
159 ([M]ϩ , 0.8%), 132 (25),
᝽
131 (19), 94 (100)
a Solvent (NMR): CDCl3, unless otherwise noted. b CD2Cl2. c Acetone-d6.
Molar ratio 1:1. Salt 4aϪ (0.50 g, 2.42 mmol) was dissolved
in acetonitrile (12.5 ml) and a solution of benzoyl chloride (0.34
g, 2.42 mmol) in acetonitrile (12.5 ml) was added dropwise. 1 M
HCl saturated with NaCl (25 ml) was added after a few minutes.
The mixture was extracted with diethyl ether (25 ml), and the
organic phase was washed with saturated aq. NaCl (25 ml),
dried (MgSO4), and evaporated to give a brown oil/solid (0.62
g). Washing of the residue with diethyl ether gave a solid
compound (0.1 g), which was purified by passage through a
short SiO2 column using DCM as eluent to give bright yellow
crystals (80 mg) of 2-oxo-4,6-diphenyl-2H-pyran-3-carbonitrile
12, mp 210–212 ЊC (from DCM–pentane). Spectroscopic
parameters, see Tables 7 and 8.
Molar ratio 1:5. To benzoyl chloride (1.70 g, 12.1 mmol)
dissolved in acetonitrile (12.5 ml) was added dropwise a solu-
tion of salt 4aϪ (0.50 g, 2.42 mmol) in acetonitrile (12.5 ml).
Work-up was as described above, except that excess of benzoyl
chloride was removed from the residue by passage through a
SiO2 column using hexane as first eluent. Subsequent elution
with DCM gave a yellow oil (0.60 g) which after crystallis-
ation gave light yellow crystals (0.20 g) of (Z)-4,4-dicyano-
1,3-diphenylbuta-1,3-dienyl benzoate (Z)-13, mp 60–62 ЊC
(decomp.) (from DCM–pentane). Spectroscopic parameters,
Table 7 and 8. Single-crystal X-ray structure determination
(Fig. 3), see below.
phenylvinyl)propanedinitrile 14a (1-acylation product, 49%),
4,4-dicyano-1-methyl-3-phenylbuta-1,3-dienyl acetate 15a (3-
acylation products, 19% (Z)- and 6% (E)-form) and 2-(1-
phenylethylidene)propanedinitrile 4a (starting material, 26%).
Since compound 14a decomposes (deacetylation) upon
chromatography and in aqueous solution, its NMR spectro-
scopic parameters were obtained directly from the above
spectra, and using GLC-FTIR, GLC-MS and HRMS, the
other parameters were obtained (see Tables 7 and 8).
Preparative experiment. Salt 4aϪ (0.30 g, 1.45 mmol)
was dissolved in acetonitrile (7.5 ml) and a solution of acetyl
chloride (0.114 g, 1.45 mmol) in acetonitrile (7.5 ml) was
added dropwise. 1 M HCl saturated with NaCl (25 ml) was
added after a few minutes. The mixture was extracted with
diethyl ether (50 ml), and the organic phase was washed with
saturated aq. NaCl (25 ml), dried (MgSO4), and evaporated
to give a brown oil (0.28 g). Column chromatography (SiO2),
using first DCM to elute (1-phenylethylidene)propanedinitrile
4a [formed by deacetylation of 2-acetyl-2-(1-phenylvinyl)-
propanedinitrile 14a, and as reaction product, see above] and
then hexane–ethyl acetate (4:1) as eluent gave two main
fractions of the (Z)- and (E)-form of 4,4-dicyano-1-methyl-3-
phenylbuta-1,3-dienyl acetate 15. Fraction 1 (37 mg) was the
pure (Z)-form, while fraction 2 (33 mg) contained a 1:1 mix-
ture of both isomers. (Z)-4,4-Dicyano-1-methyl-3-phenylbuta-
1,3-dienyl acetate (Z)-15 had mp 102–103 ЊC. Spectroscopic
parameters, Tables 7 and 8. Single-crystal X-ray structure
determination of (Z)-15 (Fig. 4), see below.
Potassium salt 4a؊ and acetyl chloride. NMR experiment.
Equimolar amounts of salt 4aϪ and acetyl chloride were
1
dissolved in CD3CN and analysed by H NMR spectroscopy,
The spectroscopic parameters for (E)-4,4-dicyano-1-methyl-
3-phenylbuta-1,3-dienyl acetate (E)-15 were obtained using
and the following compounds were observed: 2-Acetyl-2-(1-
J. Chem. Soc., Perkin Trans. 1, 2001, 497–507
505

View Article Online
Table 8 HRMS of molecular ions of acylated and methoxycarbonylated compounds 12–18
Compd
Mol. formula
Mass found
Mass calc.
Dev. (ppm)
12
C18H11NO2
273.079 219
274.082 544
376.119 527
377.123 576
252.087 998
254.106 709
255.110 251
190.074 235
191.077 262
226.073 758
227.076 347
190.074 598
195.080 565
196.083 176
159.079 896
273.078 979
274.082 334
376.121 178
377.124 533
252.089 878
254.105 528
255.108 883
190.074 228
191.077 583
226.074 228
227.077 583
190.074 228
195.097 647
196.083 002
159.079 647
Ϫ0.9
Ϫ0.8
ϩ4.4
ϩ2.5
ϩ7.5
Ϫ4.6
Ϫ5.4
0
ϩ1.7
ϩ2.1
ϩ5.4
Ϫ1.9
Ϫ4.7
Ϫ0.9
Ϫ1.6
12
C
13CH11NO2
17
13(Z)
C25H16N2O2
12
C
13CH16N2O2
24
15a(Z)
16c
C15H12N2O2
C15H14N2O2
12
C
13CH14N2PO2
14
16d
17a
C10H10N2O2
12C913CH10N2O2
C13H10N2O2
12
C
13CH10N2O2
12
17d
18a
C10H10N2O2
C12H9N3
12
C
13CH9N3
11
18d
C9H9N3
1
GLC-FTIR, GLC-MS, and software-controlled difference H
dicyano-3-cyclopropylbut-3-enoate 17d and 31% 3-cyano-2-
cyclopropyl-2-methylsuccinonitrile 18d. The latter compound
decomposed to 4d when submitted to GLC. However, prepar-
ative GLC could be used to obtain product 17d in a pure
state, as crystals, mp 39–40 ЊC. Authentic samples of 18d were
obtained by reaction of 4d with tetrabutylammonium cyanide
in a standard way; purification by elution (DCM) through a
short acid Al2O3 column gave white crystals, mp 53–55 ЊC.
Salt 4aϪ and methyl chloroformate. To the reaction solution
(before work-up as described in the standard procedure above)
was added diethyl ether (25 ml), causing precipitation of KCl
and salt 17aϪ. After filtration, the ethereal solution was washed
successively with 1 M HCl saturated with NaCl (25 ml) and
saturated aq. NaCl (25 ml) and dried (MgSO4) to give a blue-
green oil (0.30 g) which was shown (1H NMR) to consist of a
mixture of 4a (40%) and methyl 2,2-dicyano-3-phenylbut-3-
enoate 16a (60%). Compound 16a decomposes on SiO2 gel
and its spectral parameters were therefore obtained using
GLC-FTIR, GLC-MS and computer-controlled difference
NMR spectra.
and 13C NMR spectra, see Tables 7 and 8.
Potassium salt 4c؊ and acetyl chloride. Acetyl chloride (0.837
g, 0.758 ml, 10.7 mmol) was dissolved in acetonitrile (11 ml)
and a solution of potassium 1,1-dicyano-3-methyl-2-phenyl-
butenide (0.500 g, 2.13 mmol) in acetonitrile (11 ml) was
added dropwise under rigorous stirring. 30 min after addi-
tion the reaction mixture was evaporated under reduced
pressure. The crude product (0.71 g) was dissolved in diethyl
ether and washed successively with 1 M HCl and water,
dried (MgSO4), and evaporated under reduced pressure to
give an orange oil (0.43 g). Pentane was added and, after
stirring, the pentane layer was removed. This procedure was
repeated and the combined pentane fractions were evaporated
under reduced pressure to give a yellow oil (0.3 g) which was
2-acetyl-2-(2-methyl-1-phenylprop-1-enyl)propanedinitrile 14c.
Methoxycarbonylation
Standard procedure. Methyl chloro- or cyanoformate (10
mmol) was dissolved in acetonitrile (11 ml) and the appropriate
potassium salt 4Ϫ (2 mmol) as a solution in acetonitrile (11 ml)
was added dropwise under vigorous stirring. After 30 min the
solution was evaporated at reduced pressure to leave an oil,
which was dissolved in diethyl ether and washed successively
with 1 M HCl (saturated with NaCl), and water, dried
(MgSO4), and evaporated to give a yellow oil. Product distribu-
The above mixture of salts was treated with diethyl ether–
1 M HCl, and the organic phase was washed with water, dried
(MgSO4), and evaporated to give 0.12 g of chromatographically
pure methyl 4,4-dicyano-3-phenylbut-3-enoate 17a.
Salt 4cϪ and methyl chloroformate. The above yellow oil
(0.70 g) was dissolved in diethyl ether (25 ml) and washed
successively with 1 M HCl (saturated with NaCl), and water,
dried (MgSO4), and evaporated to give an orange oil (0.45 g).
Pentane (20 ml) was added and after being stirred for some
minutes, the pentane layer was removed. After repetition of this
operation, the combined pentane layer was evaporated to give a
light yellow oil (0.30 g), namely methyl 2,2-dicyano-4-methyl-3-
phenylpent-3-enoate 16c.
1
tion (Scheme 8) was estimated using H NMR and/or GLC
and entered in Table 3. Details of the separation procedures
given below. All spectroscopic parameters are entered in Tables
7 and 8.
Isolation of the products. Salt 4aϪ and methyl cyanoformate.
To the above yellow oil was added diethyl ether (150 ml) and the
mixture was extracted with saturated aq. NaHCO3 (100 ϩ 50
ml). The combined aqueous solution was acidified (HCl) and
extracted with diethyl ether. After drying of the organic phase
(MgSO4), evaporation gave 0.88 g of a brownish oil, which was
eluted with DCM through a short SiO2 column to give 0.72 g
of a chromatographically pure light yellow oil, methyl 4,4-
dicyano-3-phenylbut-3-enoate 17a. Spectroscopic parameters,
Tables 7 and 8.
Salt 4dϪ and methyl chloroformate. The above oil (0.32 g)
was shown by NMR (1H and 13C) analyses to consist of 4d
(42%), 17d (32%) and 22% methyl 2,2-dicyano-3-cyclopropyl-
but-3-enoate 16d, which could be separated by preparative
GLC.
Hydrolysis of 17a
Compound 17a was dissolved in aq. acetonitrile and after
stirring overnight the solution was evaporated to dryness to
leave a solid, which after crystallisation gave yellow needles
of methyl 4-carbamoyl-4-cyano-3-phenylbut-3-enoate 19, mp
124–125.5 ЊC (from CHCl3–pentane); νmax/cmϪ1 (FTIR) 3425m
(NH), 3331m (NH), 3191m, 2218w (CN), 1736s (CO), 1685s
(CO); δH(CDCl3) 3.66 (3 H, s), 4.17 (2 H, s), 5.89 (1 H, br), 6.46
(1 H, br), 7.4–7.5 (5 H, m); δC(CDCl3) 40.8, 52.4, 108.5, 117.1,
127.4, 128.9, 130.6, 139.2, 162.6, 164.4, 169.1; MS (EI 70 eV)
The organic phase after extraction was dried (MgSO4) and
evaporated to give 1.2 g of a brown oil, which after elution
(DCM) through a short SiO2 column and crystallisation gave
0.68 g of white crystals of 3-cyano-2-methyl-2-phenylsuccino-
nitrile 18a, mp 122–123 ЊC (from CHCl3–pentane), identical
to the Michael product synthesised from 4a and tetrabutyl-
ammonium cyanide.
Salt 4dϪ and methyl cyanoformate. The oil (0.32 g) was
shown by 1H NMR analysis to consist of 65% methyl 4,4-
m/z 244 ([M]ϩ , 15%), 227 (42), 199 (51), 184 (100); HRMS
᝽
Found: Mϩ, 244.085 756. C13H12N2O3 requires M, 244.084 792
506
J. Chem. Soc., Perkin Trans. 1, 2001, 497–507

View Article Online
(Ϫ3.9 ppm); Found: Mϩ, 245.088 165. 12C1213CH12N2O3 requires
M, 245.088 147 (Ϫ0.1 ppm).
12 776 reflections in 2θ-range 15.8–66.3Њ, Rint = 0.021. 326
Parameters refined against 5897 F2, R = 0.049 for Io > 2σ(Io)
and 0.062 for all data.
Reaction with other electrophiles
Crystal data for compound (Z)-15. C15H12N2O2, M = 252.27,
orthorhombic, P212121, a = 7.325(1), b = 9.440(1), c = 19.246(1)
Reaction of salts 4a؊, 4b؊ and 4c؊ with Br2. In separate
experiments the salts were suspended in CCl4 and equivalent
amounts of Br2 added. After being stirred for 3 h in the dark,
the organic phases were washed with water, dried (MgSO4), and
evaporated to give high yields (>90%) of the following bromo
Å, V = 1330.8(2) Å3, Z = 4, Dx = 1.259 Mg mϪ3, µ = 0.085 mmϪ1
,
T = 150(2) K, measured 24 377 reflections in 2θ-range 10.3–
66.3Њ, Rint = 0.026. 220 Parameters refined against 4872 F2,
R = 0.038 for Io > 2σ(Io) and 0.043 for all data.
compounds:
(2-bromo-1-phenylethylidene)propanedinitrile,
mp 120–121 ЊC (from EtOH) (lit.,26 122 ЊC); (2-bromo-1-phenyl-
propylidene)propanedinitrile, mp 103–105 ЊC (from MeOH)
(lit.,26 104.5–105.5 ЊC) and (2-bromo-2-methyl-1-phenyl-
propylidene)propanedinitrile, mp 89–91 ЊC (from benzene)
(lit.,26 91–92 ЊC).
Acknowledgements
P. H. S. is grateful to the Norwegian Research Council for
financial support.
Reaction between salt 4c؊ and aq. HCl. To a suspension of
salt 4cϪ (0.88 g, 3.76 mmol) in diethyl ether (50 ml) was added
HCl (50 ml; 1 M) under vigorous stirring. After separation of
the two layers, the aqueous layer was extracted with diethyl
ether (50 ml). The combined organic layers were dried (MgSO4)
and evaporated, to leave 0.70 g of a yellow oil. 1H NMR
analysis of this oil showed that the major product (≈70%) was
2-(2-methyl-1-phenylprop-1-enyl)propanedinitrile; δH(CDCl3)
1.65 (3 H, s), 2.01 (3 H, s), 4.91 (1 H, s), 7.2 (2 H, m), 7.3–7.4
(3 H, m). Efforts to purify this 3-protonated product led to
tautomerisation to the more stable isomer, viz. 2-(2-methyl-1-
phenylpropylidene)propanedinitrile 4c which also constituted
the minor constituent of the yellow oil.
References
1 P. Kolsaker, J. Arukwe, J. Barcóczy, A. Wiberg and A. K. Fagerli,
Acta Chem. Scand., 1998, 52, 490.
2 R. E. Valters and W. Flitsch, Ring-Chain Tautomerism, Plenum
Press, New York, 1985.
3 K. Wallenfels, K. Friedrich, J. Rieser, W. Ertel and H. K. Thieme,
Angew. Chem., 1976, 88, 311.
4 H. Karlsen, P. Kolsaker, C. Rømming and E. Uggerud, Acta Chem.
Scand., 1998, 52, 391.
5 P. Kolsaker, H. Karlsen and C. Rømming, Acta Chem. Scand., 1996,
50, 623.
6 H. Karlsen and P. Kolsaker, Acta Chem. Scand., 1999, 53, 487.
7 (a) O. A. Reutov, I. P. Beletskaya and A. L. Kurts, Ambident Anions,
Consultants Bureau, New York, 1983, pp. 10–146; (b) D. Kaine,
in Carbon-Carbon Bond Formation, ed. R. L. Augustine, Marcel
Dekker, New York, 1979, vol. 1, pp. 85–264; (c) H. O. House,
Modern Synthetic Reactions, W. A. Benjamin, Menlo Park, CA,
2nd edn., 1972, pp. 492–570; (d) C. R. Hauser, F. W. Swamer and
J. T. Adams, Org. React., 1954, 8, 59.
X-Ray crystallographic analysis data for compounds 6, (E)-9b,
(Z)-13 and (Z)-15§
X-Ray data were collected on a Siemens SMART CCD dif-
fractometer27 using graphite-monochromated Mo-Kα radiation
(λ = 0.710 73 Å). Data-collection method: ω-scan, range 0.6Њ,
crystal-to-detector distance 5 cm. Data reduction and cell
determination were carried out with the SAINT and XPREP
programs.27 Absorption corrections were applied by the use of
the SADABS program.28 The structure was determined and
refined using the SHELXTL program package.29 The non-
hydrogen atoms were refined with anisotropic thermal param-
eters; hydrogen atoms were located from difference Fourier
maps and refined with isotropic thermal parameters.
8 Y. U. Abramenko, A. V. Ivaschchenko, K. A. Nogaeva, N. A.
Andronova and E. B. Putsykina, J. Org. Chem. USSR (Engl.
Transl.), 1986, 22, 230.
9 G. Klopman, J. Am. Chem. Soc., 1968, 90, 223.
10 R. G. Pearson, J. Chem. Educ., 1968, 45, 581, 643.
11 D. Kaine, in Carbon-Carbon Bond Formation, ed. R. L. Augustine,
Marcel Dekker, New York, 1979, vol. 1.
12 C. Reichardt, Solvents and Solvent Effects in Organic Chemistry,
VCH, Weinheim, 2nd edn., 1988, pp. 17–25.
13 G. J. Heiszwolt and H. Koosterziel, Recl. Trav. Chim. Pays-Bas,
1970, 89, 1153.
14 F. Vögtle and P. Neumann, Chem.-Ztg., 1973, 97, 600.
15 H. D. Zook, T. J. Russo, E. F. Ferrand and D. S. Stotz, J. Org. Chem.,
1968, 33, 2222.
16 H. D. Zook and J. A. Miller, J. Org. Chem., 1971, 36, 1112.
17 O. A. Reutov, I. P. Beletskaya and A. L. Kurts, Ambident Anions,
Consultants Bureau, New York, 1983, p. 3.
18 O. A. Reutov, I. P. Beletskaya and A. L. Kurts, Ambident Anions,
Consultants Bureau, New York, 1983, pp. 80–81; cf. W. M. Muir,
P. D. Ritchie and D. J. Lyman, J. Org. Chem., 1966, 31, 3790.
19 H. Karlsen and P. Kolsaker, unpublished results.
20 A. C. Cope, C. M. Hofmann, C. Wyckoff and E. Hardenberg,
J. Am. Chem. Soc., 1941, 63, 3452.
21 D. T. Mowry, J. Am. Chem. Soc., 1945, 67, 1050.
22 E. Campaigne, G. F. Gulbenko, W. E. Kreighbaum and D. R.
Maulding, J. Org. Chem., 1962, 27, 4428.
Crystal data for compound 6. C22H16N4ؒC3H6O, M = 394.47,
monoclinic, P21/c, a = 7.099(1), b = 18.261(1), c = 16.496(1) Å,
β = 100.47(1)Њ, V = 2102.92(1) Å3, Z = 4, Dx = 1.246 Mg mϪ3
,
µ = 0.078 mmϪ1, T = 150(2) K, measured 30 823 reflections in
2θ-range 5.1–64.1Њ, Rint = 0.044. 359 Parameters refined against
7288 F2, R = 0.056 for Io > 2σ(Io) and 0.076 for all data.
Crystal data for compound (E)-9b. C13H12N2, M = 196.25,
monoclinic, P21/n, a = 10.414(1), b = 7.626(1), c = 14.178(1) Å,
β = 96.33(1)Њ V = 1119.1(1) Å3, Z = 4, Dx = 1.165 Mg mϪ3
,
µ = 0.070 mmϪ1, T = 150(2) K, measured 16 549 reflections in
2θ-range 6.8–66.2Њ, Rint = 0.026. 184 Parameters refined against
4145 F 2, R = 0.049 for Io > 2σ(Io) and 0.059 for all data.
23 A. J. Bellamy and J. B. Kerr, Acta Chem. Scand., Ser. B, 1979, 33,
370.
24 P. Kolsaker, P. Songe and C. Rømming, Acta Chem. Scand., 1997,
51, 1104.
25 K. Tortschanoff, H. Kirsch and O. E. Polansky, Liebigs. Ann.
Chem., 1975, 449.
26 A. S. Berg and P. Kolsaker, Acta Chem. Scand., Ser. B, 1980, 34, 289.
27 SMART and SAINT Area-detector Control and Integration
Software, Siemens Analytical X-ray Instruments Inc., Madison,
Wisconsin, USA, 1997.
Crystal data for compound (Z)-13. C25H16N2O2, M = 376.40,
triclinic, P-1, a = 8.157(1), b = 10.846(1), c = 12.121(1) Å, α =
68.57(1)Њ, β = 76.12(1)Њ, γ = 80.52(1)Њ, V = 965.6(1) Å3, Z = 2,
Dx = 1.295 Mg mϪ3, µ = 0.083 mmϪ1, T = 150(2) K, measured
28 G. M. Sheldrick, personal communication, 1996.
29 G. M. Sheldrick, SHELXTL, Version 5.1, Bruker AXS Inc.,
Madison, Wisconsin, USA, 1997.
§ CCDC reference number 207/504. See http://www.rsc.org/suppdata/
p1/b0/b009175h/ for crystallographic files in .cif format.
J. Chem. Soc., Perkin Trans. 1, 2001, 497–507
507