Attractive and repulsive effects in the interactions between electron-rich and electron-deficient groups in peri-substituted naphthalenes

Article abstract of DOI:10.1039/b008226k

The 1,5 Me₂N...C(sp²) separations in a series of peri-substituted naphthalene derivatives are sensitive to the through-space electron-attracting power of the carbon-containing functional group, while the corresponding MeO...C(sp²) separations are relatively insensitive in accord with the different nucleophilicities of the two groups. These separations are a balance between a steric effect and a covalent interaction, and from a comparison of the Me₂N...C and MeO...C distances for a given functional group it is proposed that the Me₂N...C interaction has an attractive component if the Me₂N...C separation is less than ～(MeO...C + 0.15) A, the latter factor arising from the greater size of the bonded nitrogen atom.

Full text of DOI:10.1039/b008226k

View Article Online / Journal Homepage / Table of Contents for this issue
Attractive and repulsive effects in the interactions between
electron-rich and electron-deficient groups in peri-substituted
naphthalenes
2
Jane O’Leary,a Paul C. Bell,a John D. Wallis*a and W. Bernd Schweizerb
a Department of Chemistry and Physics, The Nottingham Trent University, Clifton Lane,
Nottingham, UK NG11 8NS. E-mail: john.wallis@ntu.ac.uk
b Laboratorium für Organische Chemie, ETH Zentrum, Universitatstrasse 16, CH-8092 Zürich,
Switzerland
Received (in Cambridge, UK) 12th October 2000, Accepted 7th December 2000
First published as an Advance Article on the web 12th January 2001
The 1,5 Me2N ؒ ؒ ؒ C(sp2) separations in a series of peri-substituted naphthalene derivatives are sensitive to the
through-space electron-attracting power of the carbon-containing functional group, while the corresponding
MeO ؒ ؒ ؒ C(sp2) separations are relatively insensitive in accord with the different nucleophilicities of the two groups.
These separations are a balance between a steric effect and a covalent interaction, and from a comparison of the
Me2N ؒ ؒ ؒ C and MeO ؒ ؒ ؒ C distances for a given functional group it is proposed that the Me2N ؒ ؒ ؒ C interaction has
an attractive component if the Me2N ؒ ؒ ؒ C separation is less than ~(MeO ؒ ؒ ؒ C ϩ 0.15) Å, the latter factor arising
from the greater size of the bonded nitrogen atom.
Through-space interactions between functional groups are
involved in a range of chemical and biochemical processes.
Thus, van der Waals forces, hydrogen bonding, and dipole–
dipole interactions all play important roles as attractive inter-
actions in supramolecular processes such as crystal packing,
host–guest chemistry and the mode of binding of a substrate to
a biological receptor protein. The balance between attractive
and repulsive forces determines the separation between a pair
of interacting groups. Van der Waals radii were assigned origin-
ally from the closest measured intermolecular contacts between
like atoms to provide estimates of the effective sizes of bonded
atoms,1,2 and have been modified subsequently by Nyburg to
actions between a dialkylamino or a methoxy group and a
account for the asymmetric shapes of bonded atoms from the
variety of electrophilic groups.6,9,10 Since the original studies6 of
peri interactions of dimethylamino or methoxy groups with
carbonyl-containing groups further data have become available,
and selected geometric parameters for 16 compounds are given
second and later rows of the Periodic Table.3 However, to use
these parameters for predicting the closest interaction distance
between unlike atoms, by just adding the appropriate ‘radii’, is
not likely to be very successful since there may be additional
aspects to the interaction which are not present when either
atom interacts with one of its own kind. Bent has made a wide
survey of the types of compounds containing such inter-
actions.4 This is especially relevant for interactions between
electron-rich and electron-poor atoms which may interact in a
way akin to the early stage of a chemical reaction5 as has been
observed between tertiary amino groups and carbonyl groups.6,7
Thus, Bürgi, Dunitz and Schefter pointed out that the inter-
actions between a tertiary amino group and a carbonyl group in
a series of alkaloids including molecules 1–3 showed a corre-
lation between a decreasing N ؒ ؒ ؒ C distance and an increasing
C᎐O bond length, which parallels the process of the corre-
᎐
sponding addition reaction.7 Nevertheless, it is not necessarily
straightforward to deduce whether there is an attractive com-
ponent to an intramolecular interaction from the contact dis-
tance alone. Greater insight may be obtained in the future by
analysis of the experimental charge density, determined from
accurate low-temperature X-ray diffraction data, using Bader’s
‘Theory of Atoms in Molecules’8 in which bonding interactions
are characterised by saddle points in the total electron density.
The peri-disubstituted naphthalene framework has played a
prominent role in this area, most notably for investigating inter-
in Tables 1 and 3. In nearly all cases the carbonyl group is
oriented so that one face of the group is presented to the peri
substituent, and there are in-plane displacements of the two
DOI: 10.1039/b008226k
J. Chem. Soc., Perkin Trans. 2, 2001, 133–139
This journal is © The Royal Society of Chemistry 2001
133

View Article Online
Fig. 1 Patterns of displacements in peri-substituted naphthalenes: (a)
with nucleophilic group (N) and electrophilic group (E) displaced in the
same direction, and (b) with these groups attracted towards each other.
functional groups in the same sense such that the dimethyl-
amino or methoxy group is displaced towards the carbonyl-
containing group which itself is displaced outward as shown in
Fig. 1a. The orientations of the dimethylamino and alkoxy
groups are such that a lone pair of electrons is directed towards
the adjacent carbonyl group. This mode of displacement
produces an angle of attack of the ‘nucleophilic’ nitrogen
or oxygen atom on the carbonyl group of ca. 110Њ. These
arrangements have been interpreted as attractive interactions
involving overlap of lone pair electron density from the nitro-
gen or oxygen atom with the LUMO of the carbonyl group.
Supportive of this are the small displacements (0.02–0.08 Å) of
each carbonyl carbon atom from the plane of its bonded neigh-
bours towards the peri substituent, which parallel the sort of
distortion which would be expected as a nucleophile begins to
add to a carbonyl group.
Fig. 2 Molecular structures of 20 (top left), 21 (top right), 27 (bottom
left), 28 (bottom right), with anisotropic displacement parameters
However, despite the expected difference in nucleophilicities
between a dimethylamino and a methoxy group there is no clear
drawn at the 50% level.38
systematic difference between the N ؒ ؒ ؒ C᎐O and O ؒ ؒ ؒ C᎐O
᎐
᎐
distances for interactions with the same carbonyl-containing
functional group. Thus, in the peri-dimethylamino methyl
ketone 5 the N ؒ ؒ ؒ C(᎐O) distance (2.557(3) Å)6 is slightly
᎐
shorter than the O ؒ ؒ ؒ C(᎐O) distances in peri-methoxy methyl
᎐
ketones 12 and 13 (2.606(9) and 2.579(3) Å)6,11 but for the pairs
of N,N-dimethyl- and N,N-diisopropylcarboxamides, 8, 18 and
9, 19, the N ؒ ؒ ؒ C(᎐O) distances (2.698(3) and 2.764(3) Å)6,12
᎐
are distinctly longer than the O ؒ ؒ ؒ C(᎐O) distances (2.597(5)
᎐
and 2.623(2) Å). This latter result has been interpreted recently
as indicating greater ‘peri bonding’ between a methoxy and
a carboxamide group than between a dimethylamino and a
carboxamide group, and was used to explain the slower rate of
racemisation of the peri-methoxy carboxamide 19 compared to
the peri-dimethylamino carboxamide 9 which is effected by
rotation about the aryl–carbonyl bonds.12 We now examine
more closely these interactions in the light of further results
from peri-substituted electron-deficient alkenes.
If two peri groups were to react to form a bond then the two
groups would be displaced towards each other as shown in Fig.
1b rather than in the same direction as shown in Fig. 1a. Indeed
this trend has been found in our recent studies9,10 on naph-
thalenes containing peri interactions between a dimethylamino
group and an electron-deficient alkene. Structural data for these
compounds along with some further data13 from the Kirby
group are summarised in Table 2. Along the series 25 to 20 there
the lengths of the N ؒ ؒ ؒ C separations in 20 and 21 is mirrored
in the O ؒ ؒ ؒ C distances in 27 and 28. However, in 27 and 28
the in-plane displacements of the two peri substituents are in
the same sense, and there is no indication of the alkene sub-
stituent being attracted back towards the methoxy group.
Nevertheless, both alkene groups show small pyramidalisations
at the carbon attached to the naphthalene ring. The longer
interatomic separations and the lack of inward displacement of
both groups are consistent with an aryl methoxy group having a
lower nucleophilicity than the aryl dimethylamino group, and
so showing a reduced interaction with the electrophilic alkene.
Tables 1–4 contain geometric information on a wide range of
8-(dimethylamino)- and 8-methoxy-naphthalene derivatives
which contain an sp2 C atom in the 1-position and only hydro-
gen atoms ortho to these substituents. For the dimethylamino
derivatives, the N ؒ ؒ ؒ C distances decrease with peri group along
is a progressive decrease in the N ؒ ؒ ؒ C᎐C separation culminat-
᎐
ing in the formation of a bond in 20 whose structure is better
represented as the zwitterion 26 with a N–C bond length of
1.651 Å. Thus, the greatest degree of peri N ؒ ؒ ؒ C interaction
occurs when the alkene’s terminal substituents are two lactone
carbonyl groups both of which are coplanar with the alkene.
The next largest degree of interaction is in 21 where the alkene
substituents are both nitriles. We have now measured the struc-
tures of 27 and 28, the peri-methoxy analogues of these two
compounds, to investigate the degree of interaction of the
electron-deficient alkenes with methoxy groups.
the series: –CH᎐CHBr ~ –CONR , –CH᎐C(COPh) , –CO H ~
᎐
᎐
2
2
2
The molecular structures of 27 and 28 are shown in Fig. 2,
along with the structures of the corresponding dimethylamino
compounds. Selected geometric data are given in Table 4. In 27
–CO2R, –COCH3, –CH᎐C(CN)CO CH , –CHO, –CH᎐
᎐ ᎐
2 3
C(CN) , –CH᎐C(C(᎐O)O) CMe from 2.764 to 1.651 Å. This
᎐
᎐
2
2
2
forms a relative order of through-space electron-accepting
power of these groups. The decrease in the N ؒ ؒ ؒ C distance is
accompanied by a decrease in the outward bending of the
alkenyl or carbonyl group, so that for the dicyanoethenyl
and 28 the MeO ؒ ؒ ؒ C᎐C separations are longer than the
᎐
corresponding Me N ؒ ؒ ؒ C᎐C distances (27: 2.550(2) Å cf. 20:
᎐
2
1.651(3) Å and 28: 2.611(1) Å cf. 21: 2.413(2) Å). The trend in
134
J. Chem. Soc., Perkin Trans. 2, 2001, 133–139

View Article Online
Table 1 Selected geometric data for 8-(dimethylamino)naphthalene-1-carbonyl derivatives
∆C: deviation of C11 from plane of its three neighbours towards the peri substituent.
T1 and T2: torsion angles of N–Me bonds with the C7–C8 aryl bond.
X
d/Å
α/Њ
β/Њ
γ/Њ
δ/Њ
ε/Њ
∆C/Å
T1/Њ
T2/Њ
413
56
H
CH3
OH
OCH3
NMe2
NPri2
2.489(5)
2.557(3)
2.606(5)
2.594(4)
2.698(3)
2.764(3)
124.3(4)
123.4(2)
123.0(4)
123.6(3)
122.9(2)
121.2(3)
116.0(3)
116.6(2)
117.4(3)
117.4(3)
117.0(2)
118.4(2)
120.6(3)
122.3(2)
122.5(3)
122.4(3)
123.7(2)
125.1(2)
122.2(4)
123.2(2)
123.9(3)
123.9(3)
124.8(2)
125.4(2)
118.5(4)
117.2(2)
116.3(3)
116.3(3)
115.5(2)
114.1(2)
0.061(4)
0.088(3)
0.061(5)
0.062(3)
0.055(3)
0.051(3)
44.5(5)
47.0(4)
53.1(5)
48.2(5)
34.8(5)
40.4(4)
Ϫ85.2(5)
Ϫ81.3(3)
Ϫ77.1(5)
Ϫ80.1(4)
Ϫ93.2(3)
Ϫ85.8(3)
66
76
812
912
Table 2 Selected geometric data for 8-(dimethylamino)naphthalene-1-alkenes
∆C: deviation of C11 from plane of its three neighbours towards the peri substituent.
T1 and T2: torsion angles of N–Me bonds with the C7–C8 aryl bond.
Y, Z
d/Å
α/Њ
β/Њ
γ/Њ
δ/Њ
ε/Њ
∆Ca/Å
T1/Њ
T2/Њ
209
219
2210
239
2413
24
–(C(᎐O)O)2CMe2
1.651(3)
2.413(2)
2.531(2)
2.679(2)
2.717(5)
2.742(5)
2.758(2)
2.749(5)
128.6(2)
124.3(2)
123.3(2)
122.3(2)
122.0(4)
122.0(4)
123.2(1)
122.3(4)
108.8(2)
115.9(1)
116.8(1)
118.3(2)
118.8(4)
118.0(4)
117.1(1)
118.5(3)
113.3(2)
120.4(1)
122.4(1)
123.5(2)
124.0(4)
125.1(4)
127.0(1)
124.2(3)
109.9(2)
120.2(1)
121.6(1)
122.5(2)
122.0(4)
122.5(4)
121.8(1)
123.3(3)
131.4(2)
120.3(1)
118.6(2)
117.6(2)
117.8(4)
118.1(4)
116.6(1)
117.2(3)
0.362(2)
0.087(2)
0.041(2)
0.015(2)
—
—
—
49.5(3)
49.7(2)
47.9(2)
26.0(2)
23.8(6)
22.7(6)
14.9(2)
26.6(5)
Ϫ71.9(2)
Ϫ81.4(2)
Ϫ82.9(2)
Ϫ105.0(2)
Ϫ106.5(5)
Ϫ105.9(5)
Ϫ110.4(2)
Ϫ103.6(4)
᎐
CN, CN
CN, CO2Et
COPh, COPh
Br, H
Br, H
Br, H
24
2513
H, Br
—
a Hydrogen atom positions for 24 and 25 were calculated, thus forcing ∆C to be zero.
Table 3 Selected geometric data for 8-methoxynaphthalene-1-carbonyl derivatives
∆C: deviation of C11 from plane of its three neighbours towards the peri substituent.
X
d/Å
α/Њ
β/Њ
γ/Њ
δ/Њ
ε/Њ
∆C/Å
1011
1111
126
H
H
Me
Me
OH
OH
OH
OMe
NMe2
NPri2
2.628(4)
2.644(4)
2.606(9)
2.579(3)
2.559(4)
2.566(2)
2.526(3)
2.588(3)
2.597(5)
2.623(2)
123.4(2)
123.5(3)
124.4(9)
124.0(2)
124.7(3)
124.7(2)
124.7(2)
125.6(3)
122.7(4)
124.2(2)
115.1(2)
115.3(2)
115.9(8)
114.1(2)
113.2(3)
114.1(2)
114.1(2)
114.3(2)
114.4(4)
114.8(2)
122.5(2)
123.9(3)
121.7(8)
122.0(2)
124.4(2)
123.3(2)
123.3(2)
123.7(2)
124.3(4)
123.4(2)
126.4(2)
124.7(3)
125.5(8)
125.5(2)
123.4(3)
124.0(2)
122.1(2)
124.2(2)
124.0(3)
124.2(2)
114.9(2)
115.8(3)
115.3(8)
115.3(2)
115.8(3)
115.0(2)
117.4(2)
115.6(2)
115.1(4)
116.0(2)
0.089(4)
0.072(4)
0.044(10)
0.070(3)
0.022(5)
0.042(2)
0.036(2)
0.037(2)
0.039(4)
0.049(1)
1311
146
1514
1615
1714
186
1912
Table 4 Selected geometric data for 8-methoxynaphthalene-1-alkenyl derivatives
∆C: deviation of C11 from plane of its three neighbours towards the peri substituent.
Y, Z
d/Å
α/Њ
β/Њ
γ/Њ
δ/Њ
ε/Њ
∆C/Å
27
28
–(C(᎐O)O)2CMe2
CN, CN
2.550(2)
2.611(1)
123.6(2)
124.3(1)
115.0(2)
114.9(1)
122.9(2)
124.1(1)
123.2(2)
123.5(1)
117.3(2)
117.0(1)
0.077(2)
0.029(1)
᎐
J. Chem. Soc., Perkin Trans. 2, 2001, 133–139
135

View Article Online
derivative 21 the exo angles where the alkene joins the aromatic
ring are almost equal while for the carboxamide and bromo-
ethenyl derivatives, 8–9 and 24–25, the group is splayed out by
3–5.6Њ. In contrast, for the methoxy-substituted series there
is much less variation in the O ؒ ؒ ؒ C distances which only differ
by 0.097 Å (2.526–2.623 Å), consistent with the much lower
nucleophilicity of this group. In related work, Jones and Kirby
have made the connection between structural chemistry and
reactivity by relating the sensitivities of bond length changes in
several series of acetals to their relative rates of hydrolysis.16
Of the carbonyl-containing functional groups in the peri
derivatives discussed here, only the aldehyde group is suf-
ficiently small to have the possibility of lying in, or close to,
the naphthalene plane. Thus, it is indicative of the poorer
nucleophilic character of the methoxy group that in the
8-methoxy-1-naphthaldehydes 10 and 11 the aldehyde groups
lie at 30.8Њ and 38.5Њ to the naphthalene planes, and the
methoxy oxygen atom is not directed at a face of the carbonyl
group. Thus, these two compounds have been excluded from
this survey of peri interactions with sp2 carbon atoms. In
comparison, the aldehyde group in 8-(dimethylamino)-1-
naphthaldehyde 4 lies at 65.8Њ to the naphthalene plane and
compounds of type 3221 is 2.50 Å. This corresponds to a differ-
ence in the effective sizes of bonded nitrogen and oxygen atoms
of ca. 0.15 Å.22 Thus, refining the original proposal, if a
Me2N ؒ ؒ ؒ sp2C interaction distance is less than the sum of
the corresponding MeO ؒ ؒ ؒ sp2C interaction distance and the
difference in size of the bonded nitrogen and oxygen atoms
(~0.15 Å) the Me2N ؒ ؒ ؒ sp2C interaction may have an attractive
component. The larger this difference then the more attractive
the Me2N ؒ ؒ ؒ sp2C interaction. Thus, the interactions of a
dimethylamino group with the electron-deficient alkenes in 20
and 21 have some covalent character (d(MeO ؒ ؒ ؒ X) ϩ 0.15 Ϫ
d(Me2N ؒ ؒ ؒ X) = 1.049 and 0.348 Å), while the interactions of
dimethylamino groups with carboxamides in 8 and 9 are dom-
inatedbystericeffects(d(MeO ؒ ؒ ؒ X) ϩ 0.15 Ϫ d(Me2N ؒ ؒ ؒ X) =
0.049 and 0.009 Å). This parallels chemical reactivity, with
addition of amines to electron-deficient alkenes being much
easier than addition of an amine to a carboxamide group. The
interactions of dimethylamino groups with ketones and esters,
for which d(MeO ؒ ؒ ؒ X) ϩ 0.15 Ϫ d(Me2N ؒ ؒ ؒ X) = 0.185 and
0.144 Å respectively, are interpreted as involving small attract-
ive interactions which lead to the similarities in the Me2N ؒ ؒ ؒ C
and MeO ؒ ؒ ؒ C distances. The contractions in the N ؒ ؒ ؒ C dis-
tances due to HOMO–LUMO overlaps are of the same order as
the difference in the effective sizes of the nitrogen and oxygen
atoms. This proposed interpretation is illustrated further by
the naphtho-1-nitriles 3323 and 3424 which have a peri-dimethyl-
does show a short Me N ؒ ؒ ؒ C᎐O contact.
᎐
2
It may not be realistic to look for clear trends among the
O ؒ ؒ ؒ C distances in these peri-methoxynaphthalene com-
pounds, since the differences in these distances are of the same
order as those produced by adoption of different packing
arrangements or by different measurement temperatures. Thus,
although the order of increasing O ؒ ؒ ؒ C distances mirrors
those of the N ؒ ؒ ؒ C series more or less, the value for the
dicyanoethenyl derivative 28 does appear to be out of step. To
determine the trend with more confidence it would be necessary
to measure a greater number of peri-methoxy compounds,
including polymorphic forms where available, to average out
the effects of crystal packing, and to make these measurements
at very similar temperatures, ideally at ca. 100 K. It is interest-
ing to consider compounds 2917 and 3018 which contain peri
interactions between methoxy groups and aromatic rings. The
O ؒ ؒ ؒ C separations (29: 2.606(5) and 2.644(5) Å; 30: 2.659(5)
amino or a peri-methoxy group respectively: the N ؒ ؒ ؒ C dis-
tance in 33 is 2.704(6) Å and is greater than the O ؒ ؒ ؒ C distance
in 34 of 2.594(4) Å. The small value of 0.040 Å for the par-
ameter [d(MeO ؒ ؒ ؒ X) ϩ 0.15 Ϫ d(Me2N ؒ ؒ ؒ X)] indicates that
the peri dimethylamino–nitrile interaction is determined mainly
by steric effects. Comparison with the series constructed earlier
indicates that the (weak) through-space electron-attracting
power of the nitrile group is similar to that of a carboxamide
group.
A further indication of the attractive nature of certain peri
Me2N ؒ ؒ ؒ C interactions is the orientation of the dimethyl-
amino group’s lone pair towards the peri group. This is indi-
cated by the torsion angles the two Me–N bonds make with
the naphthalene ring (Tables 1 and 2). It is notable that as the
N ؒ ؒ ؒ C distance increases there is a greater tendency for the
axis of the nitrogen atom’s lone pair to move out of the naph-
thalene plane. In this way the lone pair can gain some conju-
gation with the π electrons of the aromatic system when there is
little to gain by interaction with a peri substituent. Thus, for
the ethenyl derivatives 23–25, which have N ؒ ؒ ؒ C distances in
the range 2.679–2.758 Å, the lone pair axes make torsional
angles of 38.5–47.8Њ with their naphthalene rings. Furthermore,
a similar effect is observed for the peri-dimethylamino nitrile 33
in which the axis of the amino nitrogen’s lone pair makes an
angle of 43.5Њ to the aromatic plane, in line with the predicted
minimal attractive interaction between this pair of functional
groups.
Å) do not lie far outside the range for the peri-methoxy com-
pounds discussed above, further indicating the low sensitivity
of the O ؒ ؒ ؒ C separation to the electronic character of the
carbon atom. The low variability in O ؒ ؒ ؒ C separations indi-
cates that these are controlled mainly by steric interactions.
We propose that the Me2N ؒ ؒ ؒ sp2C interactions have a
significant attractive covalent component when the N ؒ ؒ ؒ C dis-
tance is shorter than the corresponding O ؒ ؒ ؒ C distance in
accord with the greater nucleophilicity of a tertiary amine over
an aryl methyl ether. However, when the Me2N ؒ ؒ ؒ sp2C inter-
action is primarily steric in nature, the greater size of the nitro-
gen atom over the oxygen atom leads to the N ؒ ؒ ؒ C distance
between peri substituents being longer than the O ؒ ؒ ؒ C separ-
ation. To develop this further, it would be helpful to have an
estimate of the difference in effective sizes of the bonded nitro-
gen and oxygen atoms in these systems. Indeed, data from the
Cambridge Structural Database19 for peri-naphthalenes con-
taining either two dimethylamino groups or two methoxy
groups provide some useful information. The average N ؒ ؒ ؒ N
separation in 1,8-bis(dimethylamino) compounds of type 3120
is 2.81 Å, and the average O ؒ ؒ ؒ O separation in 1,8-dimethoxy
The interactions of dimethylamino or methoxy groups with
carbonyl or alkenyl groups lead to small pyramidalisations of
136
J. Chem. Soc., Perkin Trans. 2, 2001, 133–139

View Article Online
the sp2 C atoms (0.01–0.09 Å) involved in the short contact,
such that the carbon atom appears drawn towards the nitrogen
or oxygen atom. This feature has been interpreted as indicating
an attractive interaction since it represents a small deviation
along the reaction path for the addition reaction. However, this
feature is not always diagnostic for an attractive interaction.
Thus, a survey of all aryl alkyl ketones in the CSD shows a
mean pyramidality at the carbonyl carbon of 0.014 Å. Indeed,
in a more detailed study Cieplak5 has pointed out that the
molecular environment can induce a small pyramidality at
carbonyl carbon. Two specific examples from the peri-
naphthalenes are the benzoyl derivatives 3525 and 36,26 in which
gen atom’s only lone pair lies close to the aromatic plane, not
perpendicular to it. Thus, loss of conjugation between the
methoxy group and the naphthalene ring would contribute to
the greater activation energy for the racemisation process. The
most stable conformation of anisole in a molecular beam has
been determined experimentally27 to be that with the methoxy
group coplanar with the benzene ring, and this is supported by
numerous observations from the solid state.28 Furthermore,
no other energy minimum was detected corresponding to any
other orientation of the methoxy group.27 The energy difference
between conformations of anisole in which the methoxy group
lies coplanar with, or perpendicular to, the aromatic ring has
been calculated to be 5.8 and 10 kJ molϪ1 for the isolated
molecule.29,30 Measurements from the liquid phase have given
values of 6.3 and 25.2 kJ molϪ1 31,32
.
It is notable that in the peri-dimethoxynaphthalene deriv-
atives, i.e. molecules containing fragment 32, the methoxy
groups have in-plane displacements of ca. 2.5Њ towards one
another. Thus, there cannot be a strong repulsive interaction
between them at this rather short distance, which the van der
Waals radius would have predicted. The stronger influences are
the steric interaction of the O-methyl group with the ortho
hydrogen atom (typical H ؒ ؒ ؒ CH3 distance ~ 2.42 Å), and the
retention of conjugation with the naphthalene which acts to
reduce slightly the electron density at oxygen. Thus, in the many
peri-methoxy derivatives discussed above the conjugation of the
methoxy group with the aromatic system is the root of the con-
straint which displaces the methoxy group towards the other
peri substituent. In contrast, in 1-(dimethylamino)naphthal-
enes, the nitrogen atom has something to gain by not having its
lone pair lying in the plane of the naphthalene ring.
In summary we propose that the character of nucleophile–
electrophile interactions in peri-substituted napthalenes can be
deduced by observing the relative sizes of the contact distances
between the electrophile and two groups of differing nucleo-
philicity, and this approach should be transferable to other sys-
tems. Further details of the interactions can be approached
experimentally by examining the topography of the total elec-
tron density measured by X-ray diffraction measurements.
Recently, Akiba et al. have published the structure of a 1,8-
each carbonyl carbon is slightly pyramidalised (by 0.029 and
0.042 Å) towards an incipient hydrogen atom belonging to the
other peri substituent, but this may simply be a way for each
benzoyl group to increase its distance from the peri substituent.
In summary, it is misleading to interpret some of the small
pyramidalities at the carbonyl or alkenyl carbon atom
in compounds 4–28 as indicators of attractive interactions.
A comparison of Me2N ؒ ؒ ؒ C and MeO ؒ ؒ ؒ C separations
provides a better criterion.
In view of these proposals the slower rate of racemisation
of the N,N-diisopropyl 8-methoxy-1-naphthamide 19 over the
corresponding 8-(dimethylamino)-1-naphthamide
9
needs
consideration. An even larger effect is seen in the rates of
racemisations of compounds 37 and 38 which have an
dimethoxyanthracene which carries
a carbocation group
((MeO)2Cϩ–) in the 9-position.33 This group lies perpendicular
to the anthracene system and has a peri interaction to each side
with a methoxy group, leading to a five-coordinate carbon
atom. Not surprisingly, the MeO ؒ ؒ ؒ Cϩ separations of 2.43(1)
and 2.45(1) Å are ca. 0.1 Å shorter than any of the MeO ؒ ؒ ؒ C
contacts in Tables 3 and 4.
additional 2-substituent. The differences in rates correspond to
differences in the free energy of activation of ca. 5.8 kJ molϪ1
for 9 and 19 and of at least 10 kJ molϪ1 for 37 and 38. These
results were ascribed to ‘stronger peri bonding’ in the methoxy
derivatives on account of the MeO ؒ ؒ ؒ C᎐O contacts being
᎐
Experimental
shorter than the Me N ؒ ؒ ؒ C᎐O contacts, yet ‘stronger peri
᎐
2
bonding’ would not be expected for interaction of the amide
with the less nucleophilic group. We propose that the shorter
O ؒ ؒ ؒ C contacts are simply a reflection of the smaller size of
the methoxy oxygen atom, and that the reason for the difference
in racemisation rates must be sought elsewhere. The transition
states for these racemisations must have the carboxamide group
coplanar with the naphthalene ring, but this will involve signifi-
cant steric pressure from the peri substituent. A peri methoxy
group is limited in the amount to which it can splay away from
the carboxamide group by the steric interaction of the methyl
group with the ortho hydrogen atom. This effect is not so severe
for the dimethylamino group since neither methyl group lies in
the aromatic plane, and the bonding geometry at nitrogen could
distort somewhat from pyramidal towards a planar arrange-
ment in the transition state for racemisation. A further point to
consider is that in the transition state a peri methoxy group may
rotate out of the aromatic plane to avoid repulsion from the
ortho hydrogen atom. This requires loss of conjugation with the
aromatic system and has no counterpart in the racemisations of
the dimethylamino systems where in the ground state the nitro-
NMR spectra were measured on a JEOL GX 270 machine at
270 MHz for 1H and 67.8 MHz for 13C spectra using deuterated
chloroform as solvent and tetramethylsilane (TMS) as stand-
ard, and measured in ppm downfield from TMS unless other-
wise stated. Coupling constants (J ) are given in Hz. IR spectra
were recorded on an ATI Mattson Genesis Series FTIR instru-
ment as Nujol mulls or liquid films. Mass spectra were recorded
at the EPSRC Mass Spectrometry Centre, University of Wales,
Swansea. Flash chromatography was performed on 40–63 silica
gel (Merck). UV spectra were obtained on a PU 8740 UV–Vis
scanning spectrophotometer using methanol as a solvent unless
otherwise stated. Capillary melting points were taken on a
Gallenkamp melting point apparatus and are uncorrected.
8-Methoxy-1-naphthaldehyde
8-Hydroxy-1-naphthaldehyde (3.02 g, 17.6 mmol) was dissolved
in methanol (40 ml) and powdered potassium hydroxide (2.24 g,
40 mmol) and methyl iodide (10 g, 22.8 ml, 70.4 mmol) were
added to the stirred mixture and left overnight at RT. The
J. Chem. Soc., Perkin Trans. 2, 2001, 133–139
137

View Article Online
resulting solution was evaporated under reduced pressure, the
residue extracted with chloroform and subjected to flash
chromatography on silica, eluting with chloroform. This
yielded the desired ether (2.38 g, 73%) as cream crystals, mp
71 ЊC. νmax/cmϪ1: 1729, 1263, 818, 768; λmax/nm: 217, 237, 249,
V = 1162.54(8) Å3, Z = 4, P21/c, Dc = 1.34 g cmϪ3, µ (MoKα) =
0.086 mmϪ1, 2046 unique reflections with I > 2σ(I), R = 0.044,
wR = 0.113. The data set was collected at the EPSRC National
Service at 150 K, solved with SHELXS-8636 and refined as for
27. Hydrogen atoms were located in difference Fourier maps
and included in the refinement with their own isotropic
displacement parameters.
312, 323; δ : 11.08 (1H, s, O᎐C-H), 7.98 (1H, dd, J = 8.2, 1.4
᎐
H
Hz, Ar-H1), 7.94 (1H, dd, J = 7.3, 1.4 Hz, Ar-H1), 7.55–7.40
(3H, m, Ar-H3), 6.97 (1H, dd, J = 7.4, 1.4 Hz, Ar-H1), 3.99 (3H,
s, O-CH ); δ : 195.51 (C᎐O), 156.10, 135.21, 134.93, 133.03,
᎐
3
C
127.15, 126.42, 125.64, 123.24, 121.42, 106.70 (Ar-C10), 55.60
(O-CH3).
Acknowledgements
We are very grateful to Professor M. B. Hursthouse and the
EPSRC X-ray crystallography service for a dataset, to the
EPSRC Mass Spectrometry Service Centre at Swansea, and
to the EPSRC Chemical Database Service.37 We thank EPSRC
for two studentships (J. O. and P. C. B.). We thank a referee for
helpful comments.
5-(8Ј-Methoxy-1Ј-naphthylmethylidene)-2,2-dimethyl-1,3-
dioxane-4,6-dione 27
8-Methoxy-1-naphthaldehyde (200 mg, 1.1 mmol) was dis-
solved in DMSO (3 ml) and Meldrum’s acid (500 mg, 3.5
mmol) was added. The stirred mixture was left for 48 h.
Distilled water (20 ml) was added and the resulting mixture was
extracted with chloroform (5 × 30 ml). The organic extracts
were washed with water (6 × 20 ml) and the organic layer dried
(MgSO4). The solvent was removed under reduced pressure and
the residue subjected to flash chromatography on silica eluting
with 1:1 ether–hexane. This yielded 27 (90 mg, 26.2%) as yellow
crystals, mp 139 ЊC. νmax/cmϪ1: 1733, 1608, 1270, 1267, 1228,
1189, 1187, 1069, 929, 772; λmax/nm: 225, 233; δH: 9.41 (1H, s,
References
1 A. Bondi, J. Phys. Chem., 1964, 68, 441.
2 L. Pauling, The Nature of the Chemical Bond, 3rd edn., Cornell
University Press, New York, p. 260.
3 S. Nyburg and C. H. Faerman, Acta Crystallogr., Sect. B, 1985, 41,
274.
4 H. Bent, Chem. Rev., 1968, 68, 587.
5 A. S. Cieplak, in Structure Correlation, ed. H. B. Bürgi and
J. D. Dunitz, VCH, Weinheim, vol. 1, p. 205.
6 W. B. Schweizer, G. Procter, M. Kaftory and J. D. Dunitz, Helv.
Chim. Acta, 1978, 61, 2783.
7 H. B. Bürgi, J. D. Dunitz and E. Schefter, J. Am. Chem. Soc., 1973,
95, 5065.
8 R. F. W. Bader, Atoms In Molecules, A Quantum Theory, Oxford
University Press, Oxford, 1990.
H-C᎐C), 7.91 (1H, dd, J = 7.6, 1.9 Hz, Ar-H), 7.40–7.51 (4H,
m, Ar-H4), 6.91 (1H, dd, J = 7.3, 1.5 Hz, 7Ј-H), 3.93 (3H, s,
᎐
O-CH ), 1.85 (6H, s, C–CH ); δ : 164.4 (H-C᎐C), 163.1 and
᎐
3
3
C
159.7 (C᎐O), 155.9, 135.2, 130.9, 129.4, 127.2, 126.6, 125.3,
᎐
123.3, 121.4 (Ar-C ), 113.1 (HC᎐C), 106.5 (Ar-C ), 104.6
᎐
9
1
(C(CH3)2), 56.2 (O-CH3), 29.7 (C-(CH3)2); HRMS (EI): Found:
312.0994, C18H16O5 requires: 312.0998.
9 P. C. Bell and J. D. Wallis, Chem. Commun., 1999, 257.
10 P. C. Bell, M. Drameh, N. Hanly and J. D. Wallis, Acta Crystallogr.,
Sect. C, 2000, 56, 670.
1,1-Dicyano-2-(8Ј-methoxy-1Ј-naphthyl)ethene 28
11 J. C. Galluci, D. J. Hart and G. J. Young, Acta Crystallogr., Sect. B,
1998, 54, 73.
12 J. Clayden, C. McCarthy and M. Helliwell, Chem. Commun., 1999,
2059.
13 D. R. W. Hodgson, A. J. Kirby and N. Feeder, J. Chem. Soc., Perkin
Trans. 1, 1999, 949.
14 A. C. Blackburn and R. E. Gerkin, Acta Crystallogr., Sect. C, 1997,
53, 1077.
15 R. E. Gerkin, Acta Crystallogr., Sect. C, 1997, 53, 1987.
16 P. G. Jones and A. J. Kirby, J. Chem. Soc., Chem. Commun., 1986,
444.
17 A. Ford, E. Sinn and S. Woodward, J. Chem. Soc., Perkin Trans. 1,
1997, 927.
18 K. Fuji, M. Sakurai, N. Tohkai, A. Kuroda, T. Kawabata, Y.
Fukazawa, T. Kinoshita and T. Tada, Chem. Commun., 1996,
1609.
19 F. H. Allen and O. Kennard, Chem. Des. Autom. News, 1993, 8, 1;
F. H. Allen and O. Kennard, Chem. Des. Autom. News, 1993, 8, 31.
20 Data for Me2N ؒ ؒ ؒ NMe2: refcodes: DIGZIC, DMANAP10,
HOMFAQ, HOMFEU, NUPJIR, RISBAW, VIZNOH.
21 Data for MeO ؒ ؒ ؒ OMe: refcodes: JIPNOL, KEPKUL, REHGAM,
RIBKOC, RISBAW, RISBEA.
8-Methoxy-1-naphthaldehyde (200 mg, 1.1 mmol) was dis-
solved in DMSO (3 ml) and malononitrile (250 mg, 3.8 mmol)
was added. The stirred mixture was left for 48 h. Distilled water
(20 ml) was added and the resulting mixture was extracted with
chloroform (5 × 30 ml). The organic extracts were washed with
water (6 × 20 ml) and the organic layer dried (MgSO4). The
solvent was removed under reduced pressure and the residue
subjected to flash chromatography on silica (eluting with 1:1
ether–hexane). This yielded 28 (50 mg, 19%) as yellow crystals,
mp 128–127 ЊC. νmax/cmϪ1: 2226, 1580, 1241, 1072, 902, 768,
722; λmax/nm: 222, 226; δ : 9.05 (1H, s, H-C᎐C), 7.99 (1H,
᎐
H
d, J = 8.2 Hz, Ar-H1), 7.70 (1H, d, J = 7.1 Hz, Ar-H1), 7.49–7.57
(3H, m, Ar-H3), 6.98 (1H, dd, J = 3.0, 2.9 Hz, 7Ј-H), 3.99 (3H,
s, O-CH ); δ : 166.5 (H-C᎐C), 155.7, 135.3, 133.1, 128.3, 127.6,
᎐
3
C
᎐
᎐
127.2, 125.6, 122.6, 121.6 (Ar-C ), 113.8 (C᎐N), 112.3 (C᎐N),
᎐
᎐
9
107.4 (Ar-C1), 83.0 (C(CN)2), 56.1 (O-CH3); HRMS (EI):
Found: 234.0793, C15H10N2O requires: 234.0793.
X-Ray crystallography for 27†
22 The value of 0.15 Å for the difference in the effective radii of the
nitrogen and oxygen atoms is intended only as a ‘reasonable
estimate’ based on the nearest possible comparison data. There is
no perfect comparison compound for determining the N ؒ ؒ ؒ N
Crystal data: C18H16O5, Mr = 312.3, monoclinic, a = 10.291(2),
b = 11.607(3), c = 12.988(3) Å, β = 105.421(4)Њ, V = 1495.5(6)
Å3, Z = 4, Cc, Dc = 1.39 g cmϪ3, µ (CuKα) = 0.842 mmϪ1, 1317
unique reflections with F > 4σ(F), R = 0.026, wR = 0.078. The
data were measured at 103 K. The structure was solved with
SIR9734 and refined with SHELXL-93.35 Hydrogen atom posi-
tions were located in difference Fourier maps and refined with
individual isotropic displacement parameters.
separation in
a naphthalene derivative which has two peri
dimethylamino groups oriented so that the lone pairs are facing each
other. The nearest to this situation is in the structure of the 2,7-
dimethoxy derivative,19 in which the N ؒ ؒ ؒ N distance is 2.756 Å, ca.
0.05 Å shorter than in the derivatives without ortho substituents.
However, each dimethylamino group will experience some steric
repulsion from an ortho methoxy group which acts to decrease the
N ؒ ؒ ؒ N separation. In the structures of the bis(dimethylamino)-
naphthalenes which are used to estimate an effective radius for
nitrogen, each functional group is oriented with one methyl group
roughly in the aromatic plane and directed to the ortho position. The
two remaining methyl groups are directed to opposite sides of the
aromatic plane, so that the two lone pair axes lie on average at 54Њ to
the aromatic plane. The fact that the lone pairs are not directed at
each other would be expected to reduce the repulsion between the
X-Ray crystallography for 28†
Crystal data: C15H10N2O, Mr = 234.25, monoclinic, a =
6.5046(3), b = 17.5411(7), c = 10.3306(3) Å, β = 99.499(2)Њ,
† CCDC reference number 188/282. See http://www.rsc.org/suppdata/
p2/b0/b008226k/ for crystallographic files in .cif format.
138
J. Chem. Soc., Perkin Trans. 2, 2001, 133–139

View Article Online
nitrogen atoms, however each in-plane methyl group will experience
a larger steric repulsion from the ortho proton than if it had adopted
a conformation with the lone pair lying in the molecular plane which
would be expected to decrease the N ؒ ؒ ؒ N separation. Thus, half of
this N ؒ ؒ ؒ N separation distance may not be a bad estimate! The
estimation of typical O ؒ ؒ ؒ O separations is not fraught with such
problems since the methoxy groups lie in the aromatic plane.
However, it is notable that for naphthalene-1,8-diols the average
O ؒ ؒ ؒ O separations are longer (by ca. 0.08 Å) than in the dimethoxy
cases, due to the reduced ortho repulsions. Thus, it is important to
recognise that the values for the effective radii of the nitrogen and
oxygen atoms are specific for the particular functional group and its
environment and not necessarily for wider application. Estimates of
the van der Waals radii for nitrogen and oxygen rarely differ by more
than 0.1 Å.
27 P. J. Breen, E. R. Bernstein, H. V. Secor and J. I. Seeman, J. Am.
Chem. Soc., 1989, 111, 1958.
28 W. Hummel, K. Huml and H. B. Bürgi, Helv. Chim. Acta, 1988, 71,
1291.
29 T. M. Krygowski, R. Annulewicz, A. Jarmula, T. Bak, D. Rasala and
S. Howard, Tetrahedron, 1994, 46, 13155.
30 M. A. Vincent and I. H. Hillier, Chem. Phys., 1990, 140, 35.
31 S. K. Garg and C. P. Smyth, J. Chem. Phys., 1967, 46, 373.
32 N. L. Owen and R. E. Hestor, Spectrochim. Acta, Part A, 1969, 25,
343.
33 K. Akiba, M. Yamashita, Y. Yamamoto and S. Nagase, J. Am.
Chem. Soc., 1999, 121, 10644.
34 A. Altomare, M. C. Burla, M. Camalli, G. L. Cascarano, C.
Giacovazzo, A. Guagliardi, A. G. G. Moliterni, G. Polidori and
R. Spagna, J. Appl. Crystallogr., 1999, 32, 115.
35 G. M. Sheldrick, SHELXL-93, Program for Crystal Structure
Refinement, 1993, University of Göttingen, Germany.
36 G. M. Sheldrick, SHELXS-86, Program for the Solution of Crystal
Structures, 1985, University of Göttingen, Germany.
37 D. A. Fletcher, R. F. McMeeking and D. J. Parkin, The United
Kingdom Chemical Database Service, Chem. Inf. Comput. Sci.,
1996, 36, 746.
23 M. Parvez and I. I. Schuster, Acta Crystallogr., Sect. C, 1990, 46,
947.
24 G. Procter, D. Britton and J. D. Dunitz, Helv. Chim. Acta, 1981, 64,
471.
25 K. J. McCullough, A. Masayama, K. M. Morgan, M. Nojima,
Y. Okada, S. Satake and S. Takeda, J. Chem. Soc., Perkin Trans. 1,
1998, 2353.
26 H. Irngartinger, P. W. Fettel and V. Siemund, Eur. J. Org. Chem.,
1998, 2079.
38 C. K. Johnson and M. N. Burnett, ORTEPIII, Report ORNL-6895,
1996, Oak Ridge National Laboratory, Tennessee, USA.
J. Chem. Soc., Perkin Trans. 2, 2001, 133–139
139