Probing intramolecular CH-π interactions in o -quinodimethane adducts of [60]fullerene using variable temperature NMR

Article abstract of DOI:10.1021/jo201921h

Several o-quinodimethane adducts of [60]fullerene were synthesized and their intramolecular aryl CH-fullerene π interactions were studied using variable temperature-NMR (VT-NMR). Evaluation of the rate constants associated with the first-order transition

Full text of DOI:10.1021/jo201921h

Article
pubs.acs.org/joc
Probing Intramolecular CH−π Interactions in o-Quinodimethane
Adducts of [60]Fullerene Using Variable Temperature NMR
Ryan P. Kopreski, Jonathan B. Briggs, Weimin Lin, Mikael Jazdzyk, and Glen P. Miller*
̈
Department of Chemistry and Materials Science Program, University of New Hampshire, 23 Academic Way, Durham, New
Hampshire 03824, United States
*
S Supporting Information
ABSTRACT: Several o-quinodimethane adducts of [60]-
fullerene were synthesized and their intramolecular aryl
CH−fullerene π interactions were studied using variable
temperature-NMR (VT-NMR). Evaluation of the rate
constants associated with the first-order transition states for
cyclohexene boat-to-boat inversions enables quantification of
⧧
ΔG values for each inversion. A comparison between two
constitutional isomers, only one of which is capable of intramolecular CH−π interactions, provides a lower limit of 0.95 kcal/mol
for each aryl CH−fullerene π interaction.
INTRODUCTION
crystal structure such that they are within van der Waals
distance of the [60]fullerene cage. In this study, we report
intramolecular CH−π interactions in o-quinodimethane
■
8
Understanding intermolecular and intramolecular interactions
is critical to the rational design of solid-state organic devices
including organic photovoltaics and organic light-emitting
diodes. These interactions control intramolecular conforma-
tions and the relative orientations of molecules in crystalline
and thin-film devices and ultimately impact device character-
istics. Among the set of interactions at play for polycyclic
aromatic hydrocarbons (PAHs), CH−π interactions are
increasingly recognized as critical to our understanding and
(
QDM) adducts of [60]fullerene.
9
10
Shortly after the discovery and preparative scale-up of
60]fullerene, a number of derivatives were synthesized
[
11
including Diels−Alder [4 + 2] cycloadducts. Among the
Diels−Alder derivatives, one of the earliest prepared was QDM
12
1
13
adduct 1 (Scheme 1). The H and C NMR spectra of this
1
Scheme 1. Synthesis of [60]Fullerene o-Quinodimethane
prediction of crystal structures. First suggested in 1952 to
(
QDM) Diels−Alder Adduct
explain the interaction between benzene and chloroform,
CH−π interactions are regarded as the weakest of all hydrogen
bonds with an estimated enthalpy of approximately 1 kcal/
2
,3
mol. Gas-phase studies report the interaction energy of
4
benzene−methane to be 1.03−1.13 kcal/mol. Another
example is the benzene dimer which adopts a T-shaped
conformation in the solid state. NMR studies of the methyl
5
hydrogens in a number of substituted toluene derivatives in
compound indicate a time-averaged C2v symmetric structure
consistent with a [6:6] addend undergoing dynamic ring-
inversion (cyclohexene boat-to-boat inversion) about the
methylenes. The H NMR spectrum shows a broadened AB
spin system centered at 5.03 ppm indicating that “the ring
both CCl and benzene solvent reveal chemical shift differences
4
(
i.e., greater shielding in benzene solvent than in CCl ) as a
4
6
function of the acidity of the methyl groups. Nakagawa and
Fujiwara suggest that this is evidence for CH−π interactions
between the methyl hydrogens and benzene. In these and other
traditional CH−π interactions, the interacting orbitals are
presumably a σ* orbital of the H-donor and a π orbital of the
H-acceptor.
1
inversion of the cyclohexene ring is slow on the NMR time
12
scale at room temperature.”
Numerous QDM analogs have been reacted with [60]-
13−18
fullerene in a similar fashion to give a variety of derivatives.
Recently, Nishio and co-workers made a compelling case that
7
In most cases, a dynamically broadened AB multiplet is
fullerenes routinely participate in CH−π interactions. From a
3
reported for the methylenes attached to the sp hybridized
Cambridge Crystallographic Database survey, they found
numerous examples in which the fullerene π system is located
within 3 Å of a CH donor. Often, a CH−fullerene distance was
observed at or below 2.9 Å, as low as 2.5 Å. For example,
Olmstead and co-workers found that ethyl substituents in
porphyrin−[60]fullerene complexes preferentially orient in the
carbons on the fullerene at the site of addition. Several reports
1
include variable temperature H NMR experiments to further
Received: September 16, 2011
Published: January 4, 2012
©
2012 American Chemical Society
1308
dx.doi.org/10.1021/jo201921h | J. Org. Chem. 2012, 77, 1308−1315

The Journal of Organic Chemistry
Article
Figure 1. [60]Fullerene−QDM adducts showing dynamically broadened AB multiplets due to exchange of methylene hydrogens. 3−16
1
Figure 2. Correlation between PM3-calculated C62−C63 bond lengths in derivatives 1, 3, 5, 6, and 7 and their corresponding barriers to boat-to-
18
boat cyclohexene ring inversion.
Figure 3. [60]Fullerene−pyrimidine adduct derivatives showing a correlation between aryl substitution and the barrier to inversion of the
19
cyclohexene ring.
study this dynamic phenomenon. Rubin and co-workers carried
out the synthesis of 3,6-dimethyl-4,5-diphenyl QDM derivative
derivatives that all demonstrated dynamically broadened AB
16
multiplets. Two such compounds, 3 and 4, are shown in
Figure 1. VT-NMR experiments reveal activation barriers for 3
and 4 of 13.1 and 13.2 kcal/mol respectively. In a follow-up
study, Nishimura and co-workers additionally examined
2
(Figure 1) which shows coalescence of the AB multiplet at 35
1
3
°C. Variable temperature NMR (VT-NMR) experiments
1
were conducted to obtain a H NMR spectrum of the slow-
17
exchange limit which showed a clear AB quartet below −20 °C
compounds 5 and 6 (Figure 1). VT-NMR experiments gave
free energies of activation of 13.6 and 16.6 kcal/mol,
respectively, for these compounds.
(
ΔνAB = 94.8 Hz, J = 13.8 Hz) indicating a free energy of
AB
⧧
14
activation, ΔG , of 14.6 kcal/mol using eq 1 where T is the
c
Benzoquinone derivative 7 (Figure 2) shows a significantly
reduced barrier to boat-to-boat inversion (11.3 ± 0.1 kcal/mol)
coalescence temperature in Kelvin, ΔνAB is the chemical shift
difference for the A and B nuclei in Hz, JAB is the coupling
constant in Hz, and c is a constant dependent on the nature of
the exchange, here equal to 9.97.
18
of the cyclohexene ring. An interesting correlation was found
⧧
between the ΔG of compounds 1, 3, 5, 6, and 7 and the bond
2
distance between the sp hybridized carbons of the cyclohexene
⎡
⎛
⎜
⎜
⎝
⎞⎤
⎟⎥
⎟⎥
ring (C62−C63) as shown in Figure 2. According to the
authors, a smaller bond distance between C62 and C63 allows
for slightly greater bond angles for C61−C62−C63 and C62−
C63−C64, which in turn relaxes the boat-to-boat inversion
⎢
Tc
‡
ΔG = 4.57T c + log
c
c⎢
Δν + 6J2
2
⎢
⎣
AB
⎠⎥
AB ⎦
(1)
1
3
15
constraints at the C -symmetric transition state.
Rubin and others suggest that the relatively high barrier
to inversion of the cyclohexene ring is due to the rigidity of the
fullerene cage which forces a planar transition state. Nishimura
and co-workers synthesized several benzocyclobutene homo-
logues which were reacted with [60]fullerene to generate
2v
The interest in pyrimidine rings for inclusion in target
pharmaceuticals prompted Martin and co-workers to design
substituted pyrimidine−fullerene derivatives 8 and 9 as shown
19
in Figure 3. VT-NMR experiments reveal a substituent effect
1
309
dx.doi.org/10.1021/jo201921h | J. Org. Chem. 2012, 77, 1308−1315

The Journal of Organic Chemistry
Article
on the barrier to cyclohexene boat-to-boat inversion. Notably,
the presence of aryl groups on the pyrimidine ring adds
approximately 2 kcal/mol to the barrier (ΔG : 8, 15.5, 15.0
kcal/mol; 9a, 17.0, 16.8 kcal/mol; 9b, 17.0, 17.2 kcal/mol; 9c,
(coalescence method) as described above. Structurally similar
fullerene adduct 12 (Scheme 3) also shows an unexpectedly
high T of 87.3 °C in o-dichlorobenzene-d indicating a
⧧
c
4
⧧
ΔG equal to 17.2 ± 0.3 kcal/mol. The error ranges associated
c
⧧
1
7.2, 16.7 kcal/mol). Martin suggests the effect “could be due
with ΔG values for 10 and 12 reflect uncertainties in
determining T values using the coalescence method. A side-by-
c
to high torsional and angular constraints of these structures
because of the rigidity of the [fullerene] cage, which is
reinforced by the presence of the aryl groups on the pyrimidine
ring.” It is unclear, however, how the presence of aryl groups
would reinforce the rigidity of the fullerene cage.
c
side set of stacked VT-NMR spectra for compounds 10 and 12
are shown in Figure 4.
1
9
Here, we synthesized four new [60]fullerene−QDM
derivatives and studied their intramolecular cyclohexene boat-
to-boat inversions using VT-NMR. Evaluation of the rate-
⧧
constants enabled quantification of ΔG values for each
inversion. Upon the basis of these studies, we conclude that
[
60]fullerene−QDM derivatives with properly positioned
phenyl substituents are prone to engage in favorable aryl
CH−fullerene π interactions and we estimate a lower limit of
0.95 kcal/mol for each CH−π interaction.
RESULTS AND DISCUSSION
Synthesis of [60]fullerene-QDM derivative 10 was carried out
according to Scheme 2 via 2,3-bis(bromomethyl)-1,4-diphe-
■
Scheme 2. Synthesis of [60]Fullerene−QDM Derivative 10
1
Figure 4. Stacked VT- H NMR plots for compounds 10 and 12 with
Tc values (shown in parentheses) determined using an ethylene glycol
chemical shift thermometer.
To understand the high T values for boat-to-boat inversion
c
of compounds 10 and 12, the ground-state “boat” conformer
for compound 10 was optimized using an appropriate force-
field (Ghemical, following steepest descent algorithm, ΔE 10−
kJ/mol tolerance). The corresponding ground state model
indicates that none of the methylene hydrogens of the
cyclohexene ring (i.e., C −H and C −H ) are properly
4
2
0
1
nylbenzene. Unexpectedly, the room temperature H NMR
spectrum for 10 shows a sharp AB multiplet for the methylene
hydrogens of the cyclohexene ring indicating that boat-to-boat
inversion is already in the slow-exchange regime at 25 °C. This
feature distinguishes 10 from most [60]fullerene−QDM
derivatives like, for example, 1−6.
61
2
64
2
oriented to engage in meaningful C−H π interactions.
However, an ortho hydrogen from each phenyl substituent is
2
placed 2.8 Å from the nearest sp hybridized fullerene carbon
suggesting an aryl CH−fullerene π interaction as shown in
Figure 5. We propose that this electrostatic interaction is the
major contributing factor to the increased activation energies
for the boat-to-boat inversions of 10 and 12 relative to 1, the
unsubstituted QDM−fullerene adduct. To test this proposal,
two additional phenyl substituted QDM−fullerene adducts, 13
and 14 (Figure 6), were synthesized. A comparison of
diastereomers 10 and 13 allows quantification of the influence
VT-NMR studies reveal a strong solvent effect for the
coalescence temperature, T , of the exchangeable hydrogens.
The T is over 100 °C in toluene-d but only 80.5 °C in o-
dichlorobenzene-d as measured by an ethylene glycol chemical
shift thermometer. The Gibbs free energy of activation at the
coalescence temperature, ΔG , is calculated to be 16.6 ± 0.3
kcal/mol in o-dichlorobenzene-d at 80.5 °C using eq 1
c
c
8
4
⧧
c
4
Scheme 3. Synthesis of [60]Fullerene−QDM Derivative 12
1
310
dx.doi.org/10.1021/jo201921h | J. Org. Chem. 2012, 77, 1308−1315

The Journal of Organic Chemistry
Article
Figure 5. Several views of compound 10 depicting the proximity (2.814 Å) of the ortho phenyl hydrogens (green) to the nearest fullerene carbons
blue).
(
Figure 6. Target QDM−fullerene adducts 10, 13, and 14.
of CH−π interactions on the energetics of boat-to-boat
cyclohexene inversions (vide infra).
conformation as illustrated in Figure 7. This molecular gearing
necessitates that only one ortho hydrogen atom can engage in a
The syntheses of compounds 13 and 14 were carried out
according to Schemes 4 and 5, respectively. It was hypothesized
CH−π interaction with the fullerene cage at any one time.
⧧
Thus, we expected a slightly diminished T
and ΔG value for
c
1
4 with respect to 10.
To obtain accurate Gibbs energies of activation (ΔG ) for
⧧
Scheme 4. Synthesis of [60]Fullerene−QDM Derivative 13
boat-to-boat inversions of compounds 10, 13 and 14, NMR line
shapes were simulated at several different temperatures
including those where significant line broadening occurred.
Thus, VT-NMR studies were conducted for compounds 10, 13,
and 14 at the approximate temperatures of 25, 40, 55, and 70
°
C in o-dichlorobenzene-d . Because a single solvent was
4
utilized for all three fullerene derivatives, differences in
measured T values should be minimally impacted by solvent
c
effects. A low temperature spectrum was collected for each
compound to obtain chemical shift values at the slow-exchange
limit. For compound 13, no significant changes in line-shape
were observed below −15 °C. Compounds 10 and 14 are
already at the slow exchange limit at room temperature.
Temperatures of the probe were measured using an ethylene
glycol chemical shift thermometer and are reported in Table 1.
All dynamically broadened spectra were simulated using Reich’s
⧧
that any difference in ΔG values for boat-to-boat inversions of
constitutional isomers 10 and 13 would be due to differences in
intramolecular CH−π interactions. In particular, the phenyl
substituents in 13 are too far removed from the fullerene cage
to engage in intramolecular CH−π interactions. Compound 14
with four successive phenyl substituents adopts a “geared”
Scheme 5. Synthesis of [60]Fullerene−QDM Derivative 14
1
311
dx.doi.org/10.1021/jo201921h | J. Org. Chem. 2012, 77, 1308−1315

The Journal of Organic Chemistry
Article
Figure 7. Several views of compound 14 (calculated ground state equilibrium geometry, Ghemical force-field) depicting the proximity (2.78 Å) of an
ortho phenyl hydrogen (green) to the nearest fullerene carbon (blue).
Table 1. Summary of VT-NMR Simulation Data for
a
Compounds 10, 13, and 14
−1
⧧
compound
T (K)
ΔνAB (Hz) JAB (Hz) k (s ) ΔG (kcal/mol)
10
10
10
10
13
13
13
13
14
14
14
14
299.25
313.21
326.87
342.10
299.25
313.21
326.87
342.10
299.25
313.21
326.87
342.10
104.1
99.2
13.9
13.9
13.9
13.9
14.0
14.0
14.0
14.0
14.0
14.0
14.0
14.0
5.2
12.8
16.6
16.8
16.9
17.0
15.0
15.0
15.1
15.1
16.2
16.5
16.6
16.7
94.2
34.9
91.5
178.7
178.8
174.1
181.6
49.8
94.0
66.7
213.8
560.9
1640.9
9.8
50.6
19.1
50.9
54.3
51.5
146.4
a
Entries in bold represent data collected nearest the coalescence
temperature.
2
1
WinDNMR software with all input parameters except
temperature taken from the slow-exchange or near slow-
exchange limit spectra. Adjustments of spectral amplitudes and
rate constants (k) were carried out iteratively until the line-
shapes matched. Following this, Δυ was adjusted until the
simulated line-shapes overlaid each spectrum. Subsequent
iterations on k and spectral amplitude were carried out if
needed until a best fit was achieved.
Figure 8. Line-shape analyses (simulations above actual spectra in
each case) for dynamically broadened methylenes on fullerene adducts
10, 13, and 14 recorded at 70, 55 and 70 °C, respectively.
Simulations for 10, 13, and 14 were performed on spectra
recorded at approximately 70, 55 and 70 °C, respectively
⧧
(
Figure 8). The Gibbs energies of activation for boat-to-boat
confined to reflect the ΔG at the temperature simulated.
⧧
inversions (ΔG ) were obtained from the Eyring equation (eq
2
However, compounds 10 and 14 coalesce at temperatures
approximately 15 °C higher than compound 13. We assumed
that this discrepancy would only minimally impact the
) where R is the gas constant, T is the temperature in Kelvin, k
is the rate-constant obtained from the simulation, K is the
transmission coefficient (equal to 1 in a first order transition),
⧧
comparison of ΔG values.
⧧
k is Boltzmann’s constant, and ℏ is Planck’s constant. The
B
With appropriate assumptions in place, the best ΔG values
⧧
calculated ΔG values for 10, 13 and 14 are summarized in
Table 1.
for boat-to-boat inversions of 10, 13 and 14 in o-
dichlorobenzene-d are 17.0 ± 0.1, 15.1 ± 0.1, and 16.7 ±
4
⧧
0
.1 kcal/mol, respectively. The ΔG values are largest for 10
⎡
⎛
⎞⎤
⎟⎥
⎛
T ⎞
KkB
ℏ
⧧
and 14, consistent with the inclusion of stabilizing CH−π
ΔG = RT⎢ln⎜ ⎟ + ln⎜
⎝
⎠
⎝
⎠
interactions in the ground states of these derivatives. To a first
⎣
k
⎦
(2)
⧧
approximation, the difference in ΔG values determined for 10
and 13, 1.9 kcal/mol, can be taken as the total strength of two
aryl CH−fullerene π interactions in 10 suggesting that each
CH−π interaction accounts for approximately 0.95 kcal/mol of
stabilization. This value is in agreement with previously
measured values for the CH−π interaction in the benzene−
To use the information collected from Table 1 to
approximate the strength of the aryl CH−fullerene π
interactions in compounds 10 and 14, two assumptions are
made. First, it is assumed that the steric bulk of the QDM
phenyl substituents is approximately the same in 10, 13 and 14
such that the effects of solvent reorganization can be ignored in
the analysis. Second, activation energies are best approximated
4
methane system (1−1.1 kcal/mol). However, our 0.95 kcal/
mol estimate must be considered a lower limit. Thus, while the
from line-shape analyses performed at or near T and are
conformation for compound 10 shown in Figure 5 is accessible
c
1
312
dx.doi.org/10.1021/jo201921h | J. Org. Chem. 2012, 77, 1308−1315

The Journal of Organic Chemistry
Article
at all temperatures, measured T values are for time and
conformation averaged species. Compounds 10, 13 and 14
and 7 in the original plot. This validates Martin’s claim that
there is a correlation between the PM3 optimized C62−C63
bond lengths for these fullerene derivatives and their
c
adopt multiple conformations in o-dichlorobenzene-d over the
4
1
⧧
time scale of the NMR experiment. For example, H NMR
corresponding ΔG values for cyclohexene boat-to-boat
spectra for compounds 10 and 13 in o-dichlorobenzene-d4
show three sharp, resolved signals for the protons of the phenyl
substituents (at all temperatures, both above and below Tc)
indicating rapid, unrestricted rotation of the phenyl rings
relative to the NMR time scale (see Supporting Information). If
the phenyl substituents were locked into a single conformation
inversions. However, fullerene derivatives 9, 10, 12 and 14
deviate from this trend line by approximately 2 kcal/mol. These
four derivatives are the only ones that bear properly positioned
phenyl rings such that they may engage in favorable CH−π
interactions with the fullerene cage. This additional CH−π
interaction disrupts the trend described by Martin.
(
i.e., slow rotation on the NMR time scale) like that depicted in
1
Figure 5, then the corresponding H NMR spectra would show
at least five separate signals for the multiple distinct phenyl
protons on 10 and 13. Since this is not the case in o-
CONCLUSIONS
■
Several o-quinodimethane adducts of [60]fullerene were
synthesized and studied using variable temperature NMR.
Evaluation of the rate constants associated with the first-order
transition states for cyclohexene boat-to-boat inversions enables
dichlorobenzene-d , we must conclude that our 0.95 kcal/mol
4
value for each aryl CH−fullerene π interaction in 10 is a lower
limit.
⧧
quantification of ΔG values for each inversion. The results are
Interestingly, we observe more than six but far fewer than 20
distinct phenyl proton signals in the room temperature H
interpreted to indicate that ortho phenyl hydrogens on properly
positioned phenyl substituents engage in stabilizing aryl CH−
fullerene π interactions. These interactions contribute to the
ground state stabilities of [60]fullerene−QDM adducts like
derivatives 9, 10, 12 and 14, thereby leading to unusually high
barriers to cyclohexene boat-to-boat inversions compared to
derivatives like 1−8 and 13 that do not enjoy similar CH−π
1
NMR spectrum of 14 in o-dichlorobenzene-d (see Supporting
4
Information), consistent with restricted rotation of the four
geared” rings, but inconsistent with a conformationally locked
“
ground state. The gearing in 14 prohibits more than one aryl
CH−fullerene π interaction at any one time, similar to Martin’s
1
9
⧧
derivative 9. We conclude that CH−π interactions contribute
to the overall stabilities of conformation and time averaged
species 9, 10, 12 and 14. While none of these molecules is
locked into a single conformation, much less one that includes a
CH−π interaction, the impact of the CH−π interactions is
interactions. The difference in activation energies (ΔG ) for
boat-to-boat inversions of constitutional isomers 10 and 13
suggests a lower limit for an aryl CH−fullerene π interaction of
approximately 0.95 kcal/mol.
⧧
sufficient to raise their corresponding ΔG values for
EXPERIMENTAL SECTION
■
cyclohexene boat-to-boat inversions relative to [60]fullerene−
QDM derivatives like 1−8 and 13 that do not enjoy favorable
CH−π interactions.
VT- H NMR Spectroscopy. Variable temperature 1H NMR
spectra were collected on an FT-NMR operating at 499.763 MHz.
Chemical shift values were reported in parts per million (ppm) relative
to (CH ) Si (TMS). The solvent used in all cases was o-
1
An updated graph similar to the one prepared by Martin and
3
4
1
8
co-workers (Figure 2) supports these conclusions, as
illustrated in Figure 9. Newly included compounds 2, 4, 8
and 13 all fall along the same trend line as compounds 1, 3, 5, 6
dichlorobenzene-d . Low temperature spectra were obtained with
4
the use of a FTS cooling system.
NMR Line-shape Analyses. Dynamically broadened NMR
2
1
spectra were simulated using WinDNMR. All parameters except
the rate-constant (k) were taken from the slow-exchange spectra or
near the slow-exchange limit and adjusted until the line-shapes
matched. Subsequently, Δυ was iterated until the simulated line-shape
overlaid the spectrum. Iteration on both k and spectral amplitude was
carried out sequentially until a best fit was achieved.
Syntheses. Fullerene Adduct of 2,3-Bis(bromomethyl)-1,4-
diphenylbenzene (10). A 250 mL round bottomed flask was charged
with [60]fullerene (0.300 g, 0.42 mmol) and toluene (125 mL). The
mixture was stirred and purged with nitrogen for 10 min. A QDM
precursor, 2,3-bis(bromomethyl)-1,4-diphenylbenzene (0.052 g, 0.12
mmol), potassium iodide (0.240 g, 1.44 mmol) and 18-crown-6 (0.420
g, 1.59 mmol) were added to the fullerene solution. The resulting
mixture was boiled for 18 h in the dark under nitrogen. The cooled
mixture was concentrated under reduced pressure. The crude solid was
washed with chloroform until the filtrate ran clear. The chloroform
solution was washed with water, concentrated under reduced pressure
and purified via two column chromatography steps (silica, carbon
disulfide eluent). The fullerene adduct was isolated as a brown powder
1
with Rf = 0.65 (0.072 g, 59%). H NMR (500 MHz, o-
dichlorobenzene-d ) δ 4.52 (2H, J = 14 Hz), 4.72 (2H, J = 14 Hz),
4
7
4
1
1
1
1
.29−7.33 (m, 2H), 7.41−7.46 (m, 4H), 7.52 (s, 2H), 7.53−7.57 (m,
H); 13C NMR (125 MHz, CS /CDCl ) δ 41.8, 65.8, 127.4, 128.5,
2
3
28.8, 129.5, 135.3, 135.9, 136.0, 140.1, 140.2, 140.5, 140.7, 141.6,
41.7, 142.0, 142.09, 142.11, 142.15, 142.5, 142.6, 143.06, 143.09,
44.7, 145.0, 145.32 145.37, 145.44, 145.54, 145.6, 145.66, 145.71,
46.19, 146.23, 146.4, 147.6, 156.4, 156.5. LDI-MS m/z 976, 977, 978.
Figure 9. Correlation between the C62−C63 bond-lengths in
[
60]fullerene−QDM derivatives and their corresponding barriers to
boat-to-boat cyclohexene ring inversion.
HRMS (ESI) calcd for C H (M + H) 977.1325; found, 977.1303.
80
17
1
313
dx.doi.org/10.1021/jo201921h | J. Org. Chem. 2012, 77, 1308−1315

The Journal of Organic Chemistry
Article
5
,6-Bis(bromomethyl)-4,7-diphenylisobenzofuran-1,3-dione (11).
Hz), 6.70−6.75 (m, 2H), 6.78−6.85 (m, 2H), 6.88−6.95 (m, 2H),
6.98−7.04 (m, 2H), 7.06−7.13 (m, 4H), 7.16−7.22 (m, 2H), 7.23−
A round bottomed flask was charged with 5,6-dimethyl-4,7-
22
13
diphenylisobenzofuran-1,3-dione (0.200 g, 0.61 mmol), N-bromo-
succinimide (305 mg, 1.7 mmol), carbon tetrachloride (25 mL) and
catalytic benzoyl peroxide. The mixture was boiled for 54 h. The
cooled mixture was diluted with chloroform and washed several times
with water. The organic layer was dried over calcium chloride and
concentrated under reduced pressure to yield a sticky solid. The solids
were triturated with hexanes affording the dibromide as a light yellow
powder (260 mg, 88%). H NMR (400 MHz, CDCl ) δ 4.58 (s, 4H),
.39−7.46 (m, 4H), 7.54−7.59 (m, 6H); C NMR (100 MHz,
CDCl ) δ 26.0, 128.7, 128.9, 129.1, 129.6, 133.0, 142.9, 144.9, 160.8.
7.30 (m,4H), 7.44−7.50 (m, 2H); C NMR (125 MHz, CS /CDCl )
2
3
δ 42.6, 65.9, 125.5, 126.6, 126.7, 126.8, 127.8, 130.2, 130.7, 131.2,
135.3, 135.6, 136.1, 139.4, 140.0, 140.2, 140.3, 141.7, 141.8, 142.0,
142.14, 142.19, 142.23, 142.6, 143.13, 143.18, 144.7, 144.8, 145.1,
145.4, 145.5, 145.6, 145.8, 146.3, 146.5, 147.6, 156.5, 157.0. LDI-MS
m/z 1129, 1130, 1131. HRMS (ESI) calcd for C92H24Na (M + Na)
1151.1770; found, 1151.1809.
1
4,5-Bis(bromomethyl)-1,2-diphenylbenzene (15). Following a
3
13
23
7
procedure adapted from Amijs and co-workers, a 10 mL closed
vessel microwave tube was charged with 4,5-dimethylbenzene (0.1 g,
0.39 mmol) and methyl acetate (3 mL) and a magnetic stirrer. N-
Bromosuccinimide (0.15 g, 0.89 mmol) and a catalytic amount of
AIBN were added to the solution. The mixture was heated in a CEM
Discover microwave to 112 °C over 3 min. The mixture was held at
this temperature for an additional 10 min. The cooled mixture was
concentrated under reduced pressure to yield an orange oil. The crude
oil was purified via column chromatography (silica, dichloromethane
3
79
14
HRMS (ESI) calcd for C H Br NaO (M + Na) 506.9202; found,
06.9216.
Fullerene Adduct of 5,6-Bis(bromomethyl)-4,7-diphenylisobenzo-
22
2
3
5
furan-1,3-dione (12). A nitrogen flushed round bottomed flask was
charged with 5,6-bis-bromomethyl-4,7-diphenylisobenzofuran-1,3-
dione (80 mg, 0.16 mmol), [60]fullerene (370 mg, 0.51 mmol), 18-
crown-6 (725 mg, 2.7 mmol), potassium iodide (300 mg, 1.8 mmol)
and toluene (150 mL). The resulting mixture was boiled under
nitrogen in the dark for 14 h. The cooled mixture was washed
successively with saturated aqueous sodium bisulfite and water. The
organic layer was concentrated under reduced pressure. The crude
brown solids were purified by flash column chromatography (silica)
using carbon disulfide eluent to remove unreacted [60]fullerene
eluent). A sticky solid with R
= 0.8 was triturated with hexanes to yield
f
4,5-bis(bromomethyl)-1,2-diphenylbenzene as a white powder (0.11 g,
1
66%). H NMR (400 MHz, CDCl
4H), 7.21−7.24 (m, 6H), 7.43 (s, 2H); C NMR (100 MHz, CDCl
δ 30.0, 127.2, 128.2, 129.8, 133.5, 135.7, 140.2, 141.8. HRMS (EI)
) δ 4.74 (s, 4H), 7.10−7.14 (m,
3
13
)
3
79
calcd for C20H17 Br (M + H − Br) 336.0514; found, 336.0515.
followed by carbon disulfide/dichloromethane (1:1, v/v) to afford the
fullerene adduct as a brown solid with R = 0.2 (55 mg, 32%). H
NMR (400 MHz, CS /CDCl ) δ 4.67 (2H, J = 14.2 Hz), 4.55 (2H, J =
1
f
ASSOCIATED CONTENT
■
2
3
*
S
Supporting Information
1
5
6
1
1
4.2 Hz), 7.19−7.23 (m, 2H), 7.37−7.43 (m, 2H), 7.47−7.54 (m,
1
13
Select variable temperature H NMR spectra for 10, 13 and 14;
H), 7.57−7.62 (m, 2H); C NMR (100 MHz, CS /CDCl ) δ 41.9,
2
3
1
13
5.2, 127.9, 128.7, 128.8, 129.1, 129.4, 133.5, 135.5, 140.29, 140.30,
40.5, 141.8, 142.0, 142.2, 142.7, 142.8, 143.2, 144.5, 144.7, 145.3,
45.6, 145.7, 146.2, 146.4, 146.6, 147.8, 155.0, 155.3, 161.4. LDI-MS
room temperature H NMR, C NMR and LDI-MS spectra for
compounds 10−15. This material is available free of charge via
the Internet at http://pubs.acs.org.
m/z 1046, 1047, 1048. HRMS (ESI) calcd for C H NaO (M + Na)
82
14
3
1
069.0835; found, 1069.0836.
Fullerene Adduct of 4,5-Bis(bromomethyl)-1,2-diphenylbeznene
13). A 200 mL round bottomed flask was charged with [60]fullerene
AUTHOR INFORMATION
■
(
(
(
Corresponding Author
*glen.miller@unh.edu
0.300 g, 0.42 mmol), 4,5-bis(bromomethyl)-1,2-diphenylbenzene, 15
see below), (0.050 g, 0.12 mmol) and toluene (125 mL). Nitrogen
was bubbled through the stirred mixture for 10 min. Potassium iodide
0.240 g, 1.44 mmol) and 18-crown-6 (0.570 g, 2.2 mmol) were
ACKNOWLEDGMENTS
■
(
We thank the National Science Foundation for generous
support through the Nanoscale Science & Engineering Center
for High-rate Nanomanufacturing (NSF EEC 0832785) and
the New Hampshire EPSCoR Research Infrastructure Improve-
ment award (NSF EPS 0701730).
added. The reaction mixture was boiled for 16 h in the dark under
nitrogen. The cooled mixture was concentrated under reduced
pressure. The crude solid was washed with chloroform until the
filtrate ran clear. The combined filtrate was concentrated and purified
via two column chromatography steps (silica, carbon disulfide eluent).
The fullerene adduct was isolated as a brown powder with R = 0.6
f
1
(
0.075 g, 64%). H NMR (500 MHz, o-dichlorobenzene-d ) δ 4.42−
4
REFERENCES
4
4
1
1
1
.54 (br, 2H), 4.77−4.90 (br, 2H), 7.14−7.22 (m, 6H), 7.30−7.34 (m,
■
H), 7.76 (s, 2H); 13C NMR (125 MHz, CS /CDCl ) δ 45.1, 65.9,
(1) Nishio, M. CrystEngComm 2004, 6, 130−158.
(2) (a) Tamres, M. J. Am. Chem. Soc. 1952, 74, 3375. (b) Reeves, L.
W.; Schneider, W. G. Can. J. Chem. 1957, 35, 251.
(3) Nishio, M.; Hirota, M. Umezawa, Y. The CH/π Interaction:
Evidence, Nature, and Consequences; Wiley-VCH: New York, 1998.
(4) (a) Shibasaki, K.; Fujii, A.; Mikami, N.; Tsuzuki, S. J. Phys. Chem.
A. 2006, 110, 4397. (b) Morita, S.; Fujii, A.; Mikami, N.; Tsuzuki, S. J.
Phys. Chem. A 2006, 110, 10583.
2
3
26.8, 127.2, 127.6, 128.2, 129.1, 130.2, 135.4 (br), 136.6 (br), 137.3,
40.4 (br), 141.3, 141.7 (br), 141.9 (br), 142.2, 142.3, 142.7, 143.19,
43.27 (br), 144.8, 145.3 (br), 145.6 (br), 145.7 (br), 145.9, 146.4
(
br), 146.6, 147.8, 156.6 (br), 156.9 (br). LDI-MS m/z 976, 977, 978.
HRMS (ESI) calcd for C H (M) 976.1241; found, 976.1257.
80
16
Fullerene Adduct of 5,6-Bis(bromomethyl)-1,2,3,4-tetraphenyl-
benzene (14). A 200 mL round bottomed flask was charged with
[
60]fullerene (0.210 g, 0.29 mmol), 5,6-bis(bromomethyl)-1,2,3,4-
(5) Cox, E. G.; Cruickshank, D. W. J.; Smith, J. A. S. Proc. R. Soc.
1958, A247, 1.
(6) Nakagawa, N.; Fujiwara, S. Bull. Chem. Soc. Jpn. 1961, 34, 143.
(7) Suezawa, H.; Yoshida, T.; Ishihara, S.; Umezawa, Y.; Nishio, M.
CrystEngComm 2003, 5, 514−18.
(8) Olmstead, M. M.; Costa, D. A.; Maitra, K.; Noll, B. C.; Phillips, S.
L.; van Calcar, P. M.; Balch, A. L. J. Am. Chem. Soc. 1999, 121, 7090.
(9) Kroto, H. W.; Heath, J. R.; O’Brien, S. C.; Curl, R. F.; Smalley, R.
E. Nature 1985, 107, 7779−7780.
(10) Kratschmer, W.; Lamb, L. D.; Fostiropoulos, K.; Huffman, D. R.
Nature 1990, 347, 354−358.
(11) Taylor, R. Lecture Notes on Fullerene Chemistry; Imperial College
Press: London, 1999; pp 268.
tetraphenylbenzene (0.054 g, 0.095 mmol) and toluene (100 mL).
The stirred mixture was purged with nitrogen for 10 min. Potassium
iodide (0.162 g, 0.98 mmol) and 18-crown-6 (0.320 g, 1.21 mmol)
were added to the fullerene solution. The resulting mixture was boiled
for 17 h in the dark under nitrogen. The cooled mixture was
concentrated under reduced pressure. The crude solids were washed
with chloroform until the filtrate ran clear. The chloroform filtrate was
washed successively with saturated aqueous sodium bisulfite and water
and concentrated under reduced pressure. These solids were purified
by column chromatography (silica, CS eluent) to yield the fullerene
adduct as a brown solid with R = 0.5 (0.038 g, 36%). H NMR (500
MHz, o-dichlorobenzene-d ) δ 4.51 (2H, J = 14 Hz), 4.61 (2H, J = 14
2
1
f
4
1
314
dx.doi.org/10.1021/jo201921h | J. Org. Chem. 2012, 77, 1308−1315

The Journal of Organic Chemistry
Article
(
12) Belik, P.; Gugel, A.; Spickermann, J.; Mullen, K. Angew. Chem.,
Int. Ed. Engl. 1993, 32 (1), 78−80.
13) Rubin, Y.; Khan, S.; Freedberg, D. I.; Yeretzian, C. J. Am. Chem.
(
Soc. 1993, 115, 344−345.
(14) Kost, D.; Carlson, E. H.; Raban, M. Chem. Commun. 1971, 656.
(15) Zhang, X.; Foote, C. S. J. Org. Chem. 1994, 59, 5235−5238.
(16) Tago, T.; Minowa, T.; Okada, Y.; Nishimura, J. Tetrahedron Lett.
1
993, 34 (52), 8461−8464.
17) Nakamura, Y.; Minowa, T.; Tobita, S.; Shizuka, H.; Nishimura, J.
J. Chem. Soc., Perkin Trans. 2 1995, 12, 2351−2357.
18) Illescas, B. M.; Martin, N.; Seoane, C. J. Org. Chem. 1997, 62,
585−7591.
19) Gonzalez, B.; Herrera, A.; Illescas, B.; Martin, N.; Martinez, R.;
Moreno, F.; Sanchez, L.; Sanchez, A. J. Org. Chem. 1998, 63, 6807−
813. See also: Herrera, A.; Martínez-Alvarez, R.; Martín, N.; Chioua,
M.; Chioua, R.; Molero, D.; Sanchez-Vazquez, A.; Almy, J. Tetrahedron
009, 65, 5817−5823.
20) (a) Bredow, K.; Jaeschke, A.; Schmid, H.; Friebolin, H.; Kabuss,
(
(
7
(
6
́
́
2
(
S. Org. Magn. Reson. 1970, 2, 543−555. (b) Briggs, J. B.; Jazdzyk, M.
D.; Miller, G. P. Acta Cryst. E 2010, 6 (1), o12.
(
21) Reich, H. J. WinDNMR: Dynamic NMR Spectra for Windows
Journal of Chemical Education Software 3D2.
22) Chardonnens, L.; Bitsch, S. Helv. Chim. Acta 1972, 55, 1345−
352.
23) Amijs, C. H. M.; Klink, G. P. M. v.; Koten, G. v. Green Chem.
003, 5, 470−474.
(
1
(
2
1
315
dx.doi.org/10.1021/jo201921h | J. Org. Chem. 2012, 77, 1308−1315