Solvent effects on barrier to rotation of enaminonitriles using inversion transfer (1)H NMR spectroscopy and FTIR spectroscopy.

Article abstract of DOI:10.1021/jo025516a

The barrier to rotation and hydrogen bonding interactions of 2,2-dicyano-1-(N,N'-dimethylamino)vinylbenzene (1) were studied in a range of solvents. The barrier to rotation of 1 in chloroform was 14.8 kcal/mol and increased by 1.7 kcal/mol in a protic solvent, trifluoroethanol. FTIR studies showed a shift in the cyano stretch of 1 to a higher wavenumber in trifluoroethanol, which is consistent with 1 participating in a hydrogen bonding interaction at the vinyl carbon (C(v)) of the enaminonitrile group.

Full text of DOI:10.1021/jo025516a

J . Org. Chem. 2002, 67, 3937-3940
3937
Ch a r t 1
Solven t Effects on Ba r r ier to Rota tion of
En a m in on itr iles Usin g In ver sion Tr a n sfer
1H NMR Sp ectr oscop y a n d F TIR
Sp ectr oscop y
Sarjit Kaur,* Eric. S. Eberhardt,* Amanda Doucette,
Adrien Chase, and Charlie Dalby
tions caused by the presence of the bulkier and more
polarizable dicyanovinylidene group in the polymer.3
Thus, unlike the strong intermolecular amide associa-
tions in polyamides, PEANs may be less self-associated
and solvents are able to compete for hydrogen bonding
sites in PEANs, leading to solubilization.
Department of Chemistry, Vassar College,
Poughkeepsie, New York 12604-0287
eberhardt@vassar.edu
Received J anuary 11, 2002
The enaminonitrile linkage retains many of the fea-
tures of the amide linkage, including a high barrier to
rotation about the C-N bond and the ability to partici-
pate as both a hydrogen bond donor and acceptor. It is
our interest to study more closely the amide-like char-
acter of enaminonitriles. Solvent effects on the barrier
to rotation for amides have been previously reported4,5
and will be compared to enaminonitriles. The barrier to
rotation in enaminonitriles has been measured using
dynamic NMR spectroscopy, but these studies have a
(1.0 kcal/mol margin of error.6 As a result, subtle
differences in the barrier to rotation between the amide
and enaminonitrile groups have been difficult to discern.
Previous studies on amide compounds have shown that
both electronic effects and sterics influence the barrier
to rotation, with a decrease in the barrier to rotation
being observed with the increasing bulk of amide sub-
stituents.7 Molecular modeling studies on enaminoni-
triles, however, suggest that the presence of the bulkier
(CN)2C moiety is not likely to impact the barrier to
rotation to any significant extent, and electronic effects
are expected to be the predominant factor in determining
the barrier to rotation.8
Compound 1 was selected as the enaminonitrile model
compound for this study because it can participate as a
hydrogen bond acceptor but not as a hydrogen donor
(Scheme 1). The barrier to rotation about the C-N bond
for 1 was determined in solvents using inversion transfer
1H NMR spectroscopy. FTIR spectroscopy was used to
follow changes in the cyano vibrational mode of 1 in
aprotic and protic solvents of varying polarity.
The behavior of enaminonitriles in solvents of different
polarity can be evaluated using resonance theory. The
expected resonance structures for an enaminonitrile
group are outlined in Scheme 2, with three primary
structures illustrated because of the extended conjugation
possible with the cyano groups. Two of these resonance
structures indicate potential hydrogen acceptor sites: the
Abstr a ct: The barrier to rotation and hydrogen bonding
interactions of 2,2-dicyano-1-(N,N′-dimethylamino)vinylben-
zene (1) were studied in a range of solvents. The barrier to
rotation of 1 in chloroform was 14.8 kcal/mol and increased
by 1.7 kcal/mol in a protic solvent, trifluoroethanol. FTIR
studies showed a shift in the cyano stretch of 1 to a higher
wavenumber in trifluoroethanol, which is consistent with 1
participating in a hydrogen bonding interaction at the vinyl
carbon (Cv) of the enaminonitrile group.
The chemical properties of the amide bond make a
significant contribution to the formation of protein struc-
ture and bulk properties of commercially important
polymers such as Kevlar. In both of these examples, the
amide properties of interest include the ability to par-
ticipate in hydrogen bonding interactions and the “slow”
rotation observed about the C-N bond. Decades of
studies have revealed that Pauling’s resonance theory
explains much of the behavior of amides,1 and in the
design of novel amide mimics, these properties are
important features to be considered. In this study, we
report our findings on the characterization of the amide-
like character of enaminonitriles to evaluate its potential
as a peptidomimic. The enaminonitrile group was first
described by Wallenfel2 who suggested that the C(CN)2
group is inductively equivalent to an oxygen atom. Moore
et al. developed synthetic approaches to make high
molecular weight poly(enaminononitriles) (PEANs) where
the oxygen atom of the amide bond was replaced with
the C(CN)2 moiety to give the enaminonitrile group
(Chart 1).3 PEANs were originally developed to overcome
the inherent insolubility of a commercially important
aromatic polymer, Kevlar, that otherwise exhibits excel-
lent thermal and mechanical properties. With the incor-
poration of enaminonitrile linkages, PEANs became
significantly more soluble in polar aprotic solvents than
their aromatic polyamide counterparts. The improved
solubility of PEANs is believed to be the result of lower
crystallinity and/or weaker hydrogen bonding interac-
(4) (a) Drakenberg, T.; Dahlqvist, K. I.; Forsen, S. J . Phys. Chem.
1972, 76, 2178. (b) Eberhardt, E. S.; Loh, S. N.; Hinck, A. P.; Raines,
R. T. J . Am. Chem. Soc. 1992, 114, 5437.
(5) Wiberg, K. B.; Rablen, P. R.; Rush, D. J .; Keith, T. A. J . Am.
Chem. Soc. 1995, 117, 4261.
(6) (a) Mehta, P. Ph.D. Dissertation, Rensselaer Polytechnic Insti-
tute, Troy, NY, 1989. (b) Kaur, S. Ph.D. Dissertation, Rensselaer
Polytechnic Institute, Troy, NY, 1995. (c) Moore, J . A.; Mehta, P. G.
Breneman, C. M. Struct. Chem. 1997, 8 (1), 21.
(1) (a) Pauling, L. The Nature of the Chemical Bond, 3rd ed.; Cornell
University Press: Ithaca, NY, 1960. (b) Stewart, W. E.; Siddall, T. H.
III. Chem. Rev. 1970, 70, 517 and references therein. (c) Drakenberg,
T.; Dahlqvist, K. J .; Forsen, S. J . Phys. Chem. 1972, 76, 2178.
(2) (a) Wallenfels, K. Chimia 1966, 29, 303; Chem. Abstr. 1967, 66,
2109. (b) Wallenfels, K.; Friedrich, K.; Rieser, J .; Ertel, W.; Thieme,
K. Angew. Chem., Int. Ed. Engl. 1976, 15, 261.
(3) (a) Moore, J . A.; Robello, D. R. Macromolecules 1989, 22, 1084.
(b) Moore, J . A.; Robello, D. R. Macromolecules 1986, 19, 2667. (c)
Moore, J . A., Robello, D. R. Polym. Prepr. (Am. Chem. Soc., Div. Polym.
Chem.) 1986, 27 (2), 127; 1987, 28 (1), 39.
(7) (a) Kessler, H. Angew. Chem., Int. Ed. Engl. 1970, 9, 213. (b)
Yoder, C. H.; Wunderlich, M. D.; Leung, L. K.; Sandberg, J . A.; Meyer,
K. D. J . Am. Chem. Soc. 1975, 97, 2.
(8) Moore, J . A.; Breneman, C. M. Struct. Chem. 1997, 8, 13.
10.1021/jo025516a CCC: $22.00 © 2002 American Chemical Society
Published on Web 05/07/2002

3938 J . Org. Chem., Vol. 67, No. 11, 2002
Notes
N,N-dimethylbenzamide (15.3-15.6 kcal/mol in CDCl3),13
suggesting that the C(CN)2 group may have a slightly
lower electron-withdrawing effect than the oxygen atom
in amides. These results, however, demonstrate that
enaminonitriles, like amides, have a fairly high barrier
to rotation and that there is significant double-bond
character about the C-N bond in enaminonitriles, as
predicted by the resonance delocalization description
(Scheme 2). Also like amides, Eyring plots of different
solvents showed a linear relationship, an indication that
there is no significant heat capacity of activation for this
isomerization reaction.
Sch em e 1
Sch em e 2. P r im a r y Reson a n ce Str u ctu r es of
En a m in on itr ile
The data presented in Table 1 report the barrier to
rotation for 1 in solvents of different polarity. The
activation parameters show that the barrier to rotation
is mostly enthalpic in origin with a minor entropic
contribution. This thermodynamic signature is similar
to that reported for amides4 and carbamates.14 In the
absence of hydrogen bonding, the barrier to rotation
appears to increase slightly with increasing polarity of
aprotic solvents; there is a small 0.3 kcal/mol increase
in the barrier to rotation as the dielectric constant of the
solvent goes from 4.81 in chloroform to 35.94 in aceto-
nitrile. While this change is subtle, the trend is consistent
with the expectation that greater polarity of the solvent
should stabilize the resonance structures b and c that
have dipolar character (Scheme 2). Acyclic tertiary
amides also respond to solvents of greater polarity with
a higher barrier to rotation, with increases on the order
of 0.3-1.0 kcal/mol reported in the literature.4 Thus, the
enaminonitrile group appears to respond to solvent
polarity in a fashion similar to that of amides, albeit with
a smaller change in the barrier to rotation. The barrier
to rotation of 1 increased significantly when studied in
TFE (trifluoroethanol), most likely the result of hydrogen
bonding interactions between the protic polar solvent and
the enaminonitrile group. A similar increase in the
rotational barrier was observed for tertiary amides in the
presence of a protic solvent.4
While the barriers to rotation of amides and enami-
nonitriles show similar trends in protic polar solvents,
the barrier to rotation for N,N-dimethylaminoacrylo-
nitrile, NC-CHdCH-N(CH3)2, in methanol or water was
reported to be insensitive to hydrogen bonding interac-
tions.9 Also, the barrier to rotation of 12.6 kcal/mol in
chloroform for N,N-dimethylaminoacylonitrile is signifi-
cantly lower than that observed for 1. This result is not
unexpected because one cyano group in N,N-dimethyl-
aminoacrylonitrile is likely to have a lower inductive
effect than C(CN)2 in enaminonitriles. This observation
is supported by the difference in 13C NMR chemicals
shifts observed for the two vinyl carbons in the enami-
nonitrile group (a difference of 110-120 ppm).5 The vinyl
carbons of N,N-dimethylaminoacrylonitrile show a smaller
difference (95 ppm), which is consistent with a smaller
charge separation that can be expected for a lower
inductive effect and may explain the insensitivity of the
barrier to rotation for N,N-dimethylaminoacrylonitrile to
the presence of protic solvents.
Ta ble 1. Solven t Effects on th e Activa tion P a r a m eter s
of Com p ou n d 1
a
solvent
ꢀr
∆G‡ b
∆H‡ c
∆S‡ d
CDCl3
CD2Cl2
CD3CN
DMF-d6
TFEe
4.81
8.93
35.94
36.71
25.41
14.8
14.9
15.1
15.1
16.5
19.9
18.6
16.9
18.4
19.7
17
12
6
11
11
a
b
Dielectric constants.11 At 25 °C: At 25 °C: (0.2 kcal/mol.
c At 25 °C: (0.3 kcal/mol. At 25 °C: (4 cal/(mol K). e See ref 12.
d
negatively charged carbon in resonance structure b and
the negatively charged nitrogen in resonance structure
c. With protic polar solvents, both dielectric constant
effects and hydrogen bonding interactions are possible,
although the latter are expected to predominate.9 Hy-
drogen bonding should lead to a higher contribution from
the more polar resonance structures b and c, and,
accordingly, a higher barrier to rotation is expected.
There is some controversy in the literature about the
validity of using the resonance model to describe amides.10
One contention is that the resonance description does not
adequately explain changes in the bond length deter-
mined experimentally and from theoretical calculations.
This controversy also has been considered for enamino-
nitriles where the overall charge density distribution for
an enaminonitrile compound has been determined using
ab initio quantum mechanical calculations. In the case
of enaminonitriles, the theoretical calculations appear to
be in agreement with the resonance description outlined
in Scheme 2.8
The barrier to rotation (∆G‡) around the C-N bond in
1 was determined using inversion transfer 1H NMR
spectroscopy. Table 1 reports the ∆G‡ for 1 in solvents of
different polarity, along with the thermodynamic data
for ∆H‡ and ∆S‡ calculated using a linear least-squares
fit of the Eyring plots (see Figure 1 in Supporting
Information).
A significant barrier to rotation is observed for 1 (14.8
kcal/mol in CDCl3). This value is slightly lower than the
barrier to rotation reported for its amide counterpart,
(11) Reichart, C. Solvent and Solvent Effects in Organic Chemistry;
VCH: New York, 1990.
(12) The rotational barrier was determined in trifluoroethanol using
a D2O inset for an external lock.
(13) (a) Fong, C. W.; Grant, H. G. Org. Magn. Reson. 1980, 14 (2),
147. (b) Buhleier, E.; Wehner, W.; Vogtle, F.; Winfried, W. Chem. Ber.
1979, 112, 559.
(9) Rablen, P. R.; Miller, D. A.; Bullock, V. R.; Hutchinson, P. H.;
Gorman, J . A. J . Am. Chem. Soc. 1999, 121, 218.
(10) (a) Wiberg, K. B.; Laidig, K. E. J . Am. Chem. Soc. 1987, 109,
5935. (b) Wiberg, K. B. Acc. Chem. Res. 1999, 32, 922 and references
therein.
(14) Lectka, T. L.; Cox, C. J . Org. Chem. 1998, 63, 2426.

Notes
J . Org. Chem., Vol. 67, No. 11, 2002 3939
behavior of the enaminonitrile group when polarity of the
solvent is increased.
When 1 was studied in TFE, the frequency of the cyano
vibrational mode increased by 15 cm-1 relative to CH3-
CN. This shift corresponds to an increase in the bond
order of the cyano band in the presence of a protic solvent,
which is in contrast to the trend observed in polar aprotic
solvents. TFE is both a polar solvent and a strong
hydrogen bond donor. Since the trend is not consistent
with polarity effects, where a shift to lower wavenumbers
would be anticipated, hydrogen bonding interactions are
likely to be the dominant effect observed here where the
solvent is the proton donor to 1. The increase in the
barrier to rotation of 1 in trifluoroethanol implies that
resonance structures b and c are favored. The ac-
companying increase in the bond order of the cyano band
further suggests that resonance structure b is the more
important of the two resonance structures when hydro-
gen bonding interactions are present. If the contribution
from structure c was significant, a shift to lower wave-
numbers in the cyano band would have been the result.
Therefore, these results suggest that the hydrogen ac-
ceptor site in enaminonitriles is the vinyl carbon atom
labeled Cv in Scheme 2. The fact that the hydrogen
bonding apparently occurs at the Cv atom suggests that
the vinyl Cv is more basic than the nitrogen atom of the
cyano group.
F igu r e 1. FTIR spectra of compound 1 in (a) CH3CN, (b) CH2-
Cl2, (c) CHCl3, and (d) TFE.
Ta ble 2. Solven t Effects of th e Cya n o Vibr a tion a l Mod es
of 1
Our findings suggest that the enaminonitrile group
may serve as an amide mimic. The amide-like properties
include resonance that leads to a barrier to rotation about
the C-N bond and the capacity to participate in hydrogen
bonding. In particular, 1 shows a barrier to rotation that
follows the same trends in protic polar solvents as do
acyclic tertiary amides. While these properties demon-
strate the amide-like character of enaminonitriles, there
are some subtle differences. First, the barrier to rotation
of 1 is 0.6 kcal/mol lower than that of the amide
equivalent. Second, the hydrogen acceptor site in enami-
nonitrile is proposed to be a vinyl carbon atom (Cv) in
contrast to an oxygen atom in amides, and third, enami-
nonitriles are believed to exhibit weaker intermolecular
interactions than amides.3 In our future work, we plan
to study how these differences play a role in the structure
and properties of peptides containing the enaminonitrile
linkage, with the long-term goal of applying this knowl-
edge toward the synthesis of a novel class of synthetic
proteins.
CN vibrational
CN vibrational
solvent
ꢀr
mode A (cm-1
)
mode B (cm-1
)
CHCl3
CH2Cl2
CH3CN
TFE
4.81
8.93
35.94
25.41
2216
2213
2209
2224
2202
2200
2196
2209
FTIR spectroscopy was used to follow changes in the
frequency of the cyano vibrational mode for 1 in several
solvents of different polarity at concentrations of 2, 5, and
10 mM. No aggregation behavior was observed over these
concentrations. Figure 1 shows the FTIR spectrum of the
cyano region for compound 1 in solvents of different
polarity. Two cyano stretches were observed, and the
results are summarized in Table 2. Changes in the cyano
bond order provide information about the electronic
redistribution that occurs as the enaminonitrile group
interacts with its dielectric environment. A gradual
decrease in the wavenumber of the CN band is observed
as the polarity of the solvent changes from the relatively
nonpolar CHCl3 solvent to the aprotic polar solvent, CH3-
CN. These results are consistent with a reduction in the
cyano bond order as the polarity of the solvent increases
and can be explained by the resonance picture depicted
in Scheme 2. Increasing the polarity of the solvent is
expected to favor resonance structures b and c that have
dipolar character; this interpretation is consistent with
the subtle increase in the barrier to rotation observed
with increasing solvent polarity. A further distinction
between resonance structures b and c can be made on
the basis of changes in the cyano bond order. The slight
decrease in the cyano bond order with increasing solvent
polarity is best explained by contribution from resonance
structure c. If only structure b was involved, no changes
at the cyano band would have been observed. Overall,
the resonance description of enaminonitriles shown in
Scheme 2 provides an adequate explanation of the
Exp er im en ta l Section
Ma ter ia ls. 1-Chloro-2,2-dicyanovinylbenzene was synthesized
using reported procedures3; diazabicylo[2.2.2]octane (DABCO)
was sublimed under vacuum (5 mmHg) between 60 and 65 °C.
Dimethylamine HCl, acetonitrile (CH3CN), methylene chloride
(CH2Cl2), chloroform (CHCl3), dimethylformamide (DMF), di-
methylacetamide (DMAC), and trifluoroethanol (TFE) were
purchased commercially and used as received (HPLC grade).
Syn th esis of 2,2-Dicya n o-1-(N,N′-d im eth yla m in o)vin yl-
ben zen e (1). Dimethylamine HCl (0.65 g, 8.0 mmol), DABCO
(1.82 g, 16.0 mmol), and 6 mL of DMAC were placed in a dried
50 mL three-necked flask, equipped with a nitrogen inlet and a
magnetic stirrer. 1-Chloro-2,2-dicyanovinylbenzene (1.5 g, 8
mmol) was added slowly as a solid, and a light yellow solution
was obtained. After 24 h at room temperature, water was added
to the reaction mixture and a light yellow solid was obtained
upon filtration. The product was purified by recrystallization
from chloroform (0.81 gm, 50% yield), followed by sublimation
at 100 °C under vacumn (1 Torr) to give yellow crystals. Mp:
120-122 °C. FTIR (KBr pellet): 2920, 2216, 2202 (CN), 1540,

3940 J . Org. Chem., Vol. 67, No. 11, 2002
1440, 1365, 1285, 1268, 1010, 885 cm-1 1H NMR (CDCl3): δ
7.44-7.57 (m, 3H), 7.333 (d, 2H, J ) 9 Hz), 2.87, 3.52 (broad
singlets, CH3). 13C NMR (CDCl3): δ 171.8 (N-CdC), [133.0,
131.8, 129.4, 128.6], 116.6, 117.1 (CN), 53.3 (N-CdC), 43.1, 44.6
(CH3). GC/MS (EI): m/e 197 (M+), 156, 131, 91,77. Anal. Calcd
for C12H11N3: C, 73.10; H, 5.58; N, 21.32. Found: C, 73.17; H,
5.66; N, 21.42.
Notes
.
concentrations of 1 in the highest quality solvents available.
Doubling or halving the concentration of 1 in the sample did
not alter the rates. Experiments were performed using temper-
atures of 270-299 K. Temperature settings of the spectrometer
were calibrated to within 1 °C by reference to a glycol standard.
Further experimental and calculation details are available in
Supporting Information.
F TIR Sp ectr oscop y. FTIR spectra were recorded on
a
spectrophotometer using CaF2 cells with a 1 mm path length.
Each spectrum was collected for 32 or 64 scans with a resolution
of within 2 cm-1. Concentrations of 2, 5, and 10 mM of 1 in CH2-
Cl2, CHCl3, CH3CN, and TFE were studied. No concentration
effects were observed.
In ver sion Tr a n sfer NMR Sp ectr oscop y. The isomerization
rates of 1 were determined using the method described by Led
and Gessmar.15 These 1H NMR inversion transfer experiments
were conducted on a spectrometer using a 5 mm broadband
probe. Each enaminonitrile sample was prepared at 10 mM
Ack n ow led gm en t. S.K. and E.E. were supported
from Vassar College start-up funds and the Vassar
College Undergraduate Summer Research Institute.
E.E. was supported by the Research Corporation College
Cottrell Award (CC5221). A.D. was supported by a
generous gift from the Bloch family.
Su p p or tin g In for m a tion Ava ila ble: Eyring plots for
model compound 1 (Figure 1), and further details on treatment
of experimental data for inversion transfer 1H NMR spectro-
scopy experiments. This material is available free of charge
via the Internet at http://pubs.acs.org.
(15) (a) Forsen, S.; Hoffman, R. A. J . Chem. Phys. 1963, 39, 2892.
(b) Led, J . J .; Gessmar, H. J . Magn. Reson. 1982, 49, 444. (c) Eberhardt,
E. S.; Panasik, N., J r.; Raines, R. T. J . Am. Chem. Soc. 1996, 118,
12261.
J O025516A