Conformational reactions of D2-symmetric twisted acenes

Article abstract of DOI:10.1016/j.tet.2008.05.030

Computational analyses of the longitudinally twisted polycyclic aromatic hydrocarbons 9,18-diphenyltetrabenz[a,c,h,j]anthracene (1), 9,10,19,20-tetraphenyltetrabenzo[a,c,j,l]naphthacene (2), and 9,10,11,20,21,22-hexaphenyltetrabenzo[a,c,l,n]pentacene (3) were performed at the B3LYP/6-31G(d) level of theory. Ground state and transition state structures were located for two classes of conformational reactions in these molecules: twist inversions (enantiomerization or racemization reactions) and phenyl rotations, and the free energies of activation for these processes were calculated. Where possible, the computational results were compared with both new and existing experimental data for the rates of these conformational reactions. Multiple pathways were identified for some processes, but the low-energy transition states were often quite different from those that common chemical intuition might suggest.

Full text of DOI:10.1016/j.tet.2008.05.030

Tetrahedron 64 (2008) 8630–8637
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Conformational reactions of D2-symmetric twisted acenes
*
Robert A. Pascal, Jr. , Qian Qin
Department of Chemistry, Princeton University, Princeton, NJ 08544, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Computational analyses of the longitudinally twisted polycyclic aromatic hydrocarbons 9,18-diphe-
nyltetrabenz[a,c,h,j]anthracene (1), 9,10,19,20-tetraphenyltetrabenzo[a,c,j,l]naphthacene (2), and
9,10,11,20,21,22-hexaphenyltetrabenzo[a,c,l,n]pentacene (3) were performed at the B3LYP/6-31G(d) level
of theory. Ground state and transition state structures were located for two classes of conformational
reactions in these molecules: twist inversions (enantiomerization or racemization reactions) and phenyl
rotations, and the free energies of activation for these processes were calculated. Where possible, the
computational results were compared with both new and existing experimental data for the rates of
these conformational reactions. Multiple pathways were identified for some processes, but the low-
energy transition states were often quite different from those that common chemical intuition might
suggest.
Received 29 March 2008
Received in revised form 7 May 2008
Accepted 8 May 2008
Available online 11 May 2008
Keywords:
Polycyclic aromatic hydrocarbons
Acenes
Conformational analysis
Transition state analysis
Dynamic NMR spectroscopy
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction
are twisted ribbons. In the solid state, they adopt conformations of
approximate D2 symmetry with end-to-end twists of 66ꢀ and 144ꢀ,
The typical acene is a flat, linear chain of fused benzene rings, but
one need onlyattach a few bulkysubstituents tosuch a molecule and
substantial deformationsfromplanarity may result, mostcommonly
in the form of a twisting distortion.1 Thus, while anthracene and
pentacene are flat, 9,18-diphenyltetrabenz[a,c,h,j]anthracene (1)
and 9,10,11,20,21,22-hexaphenyltetrabenzo[a,c,l,n]pentacene (3)
respectively, and various computational studies indicate that D2-
symmetric, twisted conformations are the ground state structures in
the gas phase as well.2–6 These two polyphenyl tetrabenzoacenes,
along with 9,10,19,20-tetraphenyltetrabenzo[a,c,j,l]naphthacene
(2),7 are the paradigmatic twisted acenes.
Compounds 1–3 are obviously chiral, but they have disap-
pointingly low barriers to racemization. For 1, the free energy of
activation for this process (D
Gzrac) is estimated to be 16.7 kcal/mol
based on dynamic NMR studies of a diisopropyl derivative,3 and
such a low barrier precludes the resolution of enantiomers above
ꢁ60 ꢀC. Compound 3 has been resolved by chiral chromatography,
but it racemizes with a half-life of 9.3 h at 25 ꢀC (
D
Gz ¼23.8 kcal/
rac
mol).4,5 For comparison, the helicenes typically have high barriers
to racemization (
D
Gz ¼35–44 kcal/mol8). It has always been clear
rac
that the twisted acenes do not racemize via a single, planar tran-
sition state; such a process would have high activation energy.
Instead, the twists must invert via lower energy, nonplanar
intermediates and transition states. In this paper we explore the
pathways for twist inversion (enantiomerization) of compounds 1–
3, as well as phenyl rotation reactions in these molecules, by means
of computational studies at the B3LYP/6-31G(d) level of theory. The
study of these highly symmetric molecules might be easier, both
computationally and experimentally, than that of less symmetric
twisted compounds, because the former have fewer conformations
to evaluate and simpler spectra. Nonetheless, compounds 1–3,
though quite similar in appearance, proved to be remarkably
different in their conformational dynamics.
1
2
3
* Corresponding author. Tel.: þ1 609 258 5417; fax: þ1 609 258 6746.
E-mail address: snake@chemvax.princeton.edu (R.A. Pascal Jr.).
0040-4020/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2008.05.030

R.A. Pascal, Jr., Q. Qin / Tetrahedron 64 (2008) 8630–8637
8631
Table 1
2. Results and discussion
Calculated energies for conformations and conformational transition states of
compounds 1–3 at the B3LYP/6-31G(d) level of theory
2.1. Twist inversion in 9,18-diphenyltetra-
benz[a,c,h,j]anthracene (1)
Conformation
Symm.
DE
D
Ecorr
D
G298 nI nmin
D
Gzexpt (T)
Compound 1
Ground state
Highest possible symmetry D2h
D2
0.0
0.0
0.0
0
5
1
0
1
1
1
25.2
The crystal structure of compound 1 as well as those of two
simple derivatives show molecules with approximate D2 symmetry
and end-to-end twists of 61–70ꢀ,2,3 and these findings are well
reproduced by molecular mechanics (Sybyl,9 MMFF10), semi-
empirical (AM111), ab initio (HF/STO-3G, HF/3-21G12), and density
functional [B3LYP/6-31G(d)13,14] calculations.
A frequency calculation for the B3LYP/6-31G(d) structure shows
it to be a potential energy minimum (0 imaginary frequencies; see
Table 1), and extensive computational searches have failed to find
a lower energy conformation. Given these data, the D2 conforma-
tion may safely be taken to be the ground state of 1.
A naı¨ve analysis might suggest that the transition state for the
twist inversion would be a D2h-symmetric structure with a flat
tetrabenzanthracene core. Such a structure possesses the highest
symmetry that compound 1 might attain. However, geometry op-
timization of this D2h structure followed by a frequency calculation
shows it to be a saddle point of order 5 lying 42.1 kcal/mol above
the ground state (Table 1).15 Thus D2h-1 is chemically irrelevant,
and at least some reduction of the symmetry is necessary to locate
the true transition state structure.
37.3 37.3 42.1
17.8 17.5 18.2
24.1 24.0 22.2
28.5 27.8 27.7
37.4 36.7 36.7
22.8 22.4 21.4
ꢁ114.1
Twist inversion TS
‘Slice’ conformation
‘Butterfly’ TS
Phenyl rotation TS
Inversion-rotation TS
Compound 2
C2h
C2h
C2v
C2
ꢁ13.4 16.7 (300 K)b
18.0
ꢁ17.4
ꢁ92.3
Cs
ꢁ30.0 19.2 (348 K)c
Ground state
Highest possible symmetry D2h
D2
0.0
0.0
0.0
0
7
1
0
1
0
1
0
1
18.3
ꢁ135.4
ꢁ27.3
14.9
78.5 78.9 87.2
21.2 21.3 21.2
20.7 20.9 19.9
21.9 22.0 22.0
21.6 21.9 20.5
25.5 25.8 26.6
22.4 22.5 21.4
37.4 37.1 36.5
Twist inversion TS1
Twist inversion INT1
Twist inversion TS2
Twist inversion INT2
Twist inversion TS3
Twist inversion INT3
Phenyl rotation TS
Compound 3
C1
C2
C1
C1
C1
Ci
ꢁ13.6
10.9
ꢁ25.9
12.2
ꢁ104.9
C1
Ground state
Highest possible symmetry D2h
D2
0.0
0.0
0.0
0
12.0
135.8 136.3 153.0 11 ꢁ148.5
Twist inversion TS1
Twist inversion INT
Twist inversion TS2
Central phenyl
rotation TS
C1
C2
C2h
C1
29.9 30.6 33.2
19.3 19.9 20.5
19.5 19.9 20.9
36.9 36.9 37.0
1
0
1
1
ꢁ13.6 23.8 (298 K)d
14.2
ꢁ13.2
ꢁ75.9
The correct potential energy profile (Fig. 1) for the twist
inversion of compound 1 proved to be quite simple, at least at the
B3LYP/6-31G(d) level of theory. Geometry optimization of a wave-
like, C2h-symmetric structure for 1 yields a genuine transition state
(TS, Fig. 1) for the twist inversion that lies 18.2 kcal/mol above the
ground states (GS and eGS, for ‘ground state’ and ‘enantiomeric
ground state’). This TS structure displays a single imaginary
frequency, and when small distortions are applied to reduce the
symmetry to C2, it rapidly evolves to the GS upon geometry opti-
mization. Further, the calculated barrier for this structure is in good
agreement with the experimental value for the enantiomerization
of 1 (16.7 kcal/mol3).
Side phenyl
rotation TS
C1
35.8 35.6 35.5
1
ꢁ103.3
For each entry, the energy relative to the D2 ground state (
point-corrected relative energy [ Ecorr
D
E, kcal/mol), the zero-
D
¼D
(EþZPE)], the relative free energy (
DG298),
the number of imaginary frequencies (nI), and the lowest calculated frequency (nmin
,
cmꢁ1) are given.a The experimental free energies of activation for the conforma-
tional reactions corresponding to the calculated transition states, where known, are
given in the last column.
a
The calculated absolute energies (E, Ecorr, and G298) are found in Supplementary
data.
b
Ref. 3.
This work.
Ref. 5.
c
d
20
15
10
5
TS (C2h
)
expt
GS (D2)
eGS (D2)
0
Figure 1. Free energy profile for the twist inversion of 9,18-diphenyltetrabenz[a,c,h,j]anthracene (1) calculated at the B3LYP/6-31G(d) level of theory. The ball-and-stick drawings
illustrate the potential minima and transition state structures for this process.

8632
R.A. Pascal, Jr., Q. Qin / Tetrahedron 64 (2008) 8630–8637
This simple picture is pleasing, but it is important to note that
2.2. Twist inversion in 9,10,19,20-tetraphenyl-
the structures illustrated in Figure 1 are not the only minima and
transition states potentially available to compound 1, even if one
limits the possibilities to structures that might be involved in
twist inversion. For example, the next lowest energy minimum
located for 1 is the C2h-symmetric ‘slice’ conformation (below
left). As an achiral structure, it is a candidate intermediate for
the twist inversion, but its relative energy is 22.2 kcal/mol above
the GS and thus this conformation is probably unimportant.
There are also alternative TS structures for twist inversion. The
C2v-symmetric ‘butterfly’ structure (below right) is a genuine
enantiomerization TS. It has one imaginary frequency, and slight
tetrabenzo[a,c,j,l]naphthacene (2)
A previous molecular mechanics study indicated that compound
2 has a D2-symmetric ground state, but unlike compounds 1 and 3,
there is no crystal structure of 2.7 At both the AM1 and B3LYP/6-
31G(d) levels of theory, D2-2 is the lowest energy conformation,
and the end-to-end twist is calculated to be 109.4ꢀ at the latter
level. All attempts to resolve compound 2 by chiral chromatography
have been unsuccessful, but this does not necessarily indicate that
the barrier to racemization is low. Computational studies have the
potential to clarify this issue.
distortions give
a
C2 ‘butterfly’ structure that is smoothly
The enantiomerization of compound 2 was examined first at the
AM1 level, and it was soon obvious that the reaction path was
unusually complex and populated by numerous shallow minima
and low transition states. The structures resulting from that study
(not shown) were then used as starting points for the B3LYP/6-
31G(d) calculations that ultimately led to the reaction profile il-
lustrated in Figure 2. Two symmetric structures were initially
identified as likely candidates for intermediates on the racemiza-
converted to the GS upon geometry optimization. Unfortunately,
this C2v-TS is 27.7 kcal/mol above the GS, and cannot compete
with the wavelike C2h TS that is almost 10 kcal/mol lower in
energy. Surprisingly, however, this C2v geometry is chemically
significant: exactly this conformation has been observed in
the X-ray crystal structure of the dilithium dianion of
compound 1.16
tion pathway: the C2-symmetric INT1 (
D
G¼19.9 kcal/mol) and the
Ci-symmetric INT3 (DG¼21.4 kcal/mol). The difficult task was
linking these structures by well-defined transition states and in-
termediates to the D2 GS.
The first barrier to racemization is provided by TS1
(
D
G¼21.2 kcal/mol). In this structure, benzo group A (see Fig. 2) has
already moved past the adjacent phenyl ring, and benzo group B is
in mid-pass relative to its neighboring phenyl; completion of this
movement gives the C2 INT1. So far the reaction path is clear, but no
single transition state has been found to lead directly from INT1 to
INT3. Instead, it is necessary to rotate one of the phenyl groups in
INT1 by roughly 90ꢀ to give the closely related INT2 (
D
G¼20.5 kcal/
D
C
B
B
A
30
eTS3 (C1)
TS3 (C1)
25
TS2 (C1)
INT3 (Ci)
TS1 (C1)
INT2 (C1)
INT1 (C2)
20
5
GS (D2)
0
Figure 2. Free energy profile for the twist inversion of 9,10,19,20-tetraphenyltetrabenzo[a,c,j,l]naphthacene (2) calculated at the B3LYP/6-31G(d) level of theory. The ball-and-stick
drawings below the profile illustrate the potential minima, and the drawings above illustrate the transition state structures for this process. Only half of the free energy profile is
shown; the potential minima and transition states following INT3 are the enantiomers of those illustrated.

R.A. Pascal, Jr., Q. Qin / Tetrahedron 64 (2008) 8630–8637
8633
mol) via the modest TS2 (
the bond denoted by solid atoms in Figure 2. After many trials using
the quadratic synchronous transit search algorithms (QST2 and
D
G¼22.0 kcal/mol); rotation occurs about
eINT). TS2 is merely the low barrier for counter-rotation of the two
central phenyl groups in INT (about the bonds designated by solid
atoms in Fig. 2). The chemically significant barrier to racemization
is again provided by a C1-symmetric TS1 that is 33.2 kcal/mol above
the D2 GS. In terms of conformational changes, TS1 is rather ‘late’: in
order to reach the transition state structure, benzo group A (see
Fig. 2) has largely moved past the face of the adjacent phenyl ring,
and benzo B is in the process of slipping by its neighboring phenyl.
QST3) in GAUSSIAN 03, a transition state (TS3,
DG¼26.6 kcal/mol)
was found to link INT2 and INT3. Animation of the imaginary mode
in TS3 reveals a complex motion in the transition state: benzo
group B is moving past phenyl C while phenyls C and D are moving
in the opposite direction. When small distortions are applied to the
TS3 structure, it evolves smoothly into INT2 or INT3 upon geometry
optimization, as expected. Once INT3 is attained, racemization is
completed via enantiomeric transition states and intermediates
that lead to the enantiomeric ground state.
TS3 is the rate limiting transition state for racemization of 2. It is
26.6 kcal/mol above the ground state, suggesting that the enan-
tiomers of 2 should be configurationally stable at room tempera-
ture. However, in a prior study of a complex polyphenylene, we
found that B3LYP/6-31G(d) overestimated conformational barriers
by an average of 3 kcal/mol.17 We suspect, then, that the true barrier
for the enantiomerization of 2 is no greater than 24 kcal/mol,
making it potentially resolvable, but not configurationally stable at
room temperature.
The calculated relative free energy (
significantly higher than the experimental
D
G¼33.2 kcal/mol) for TS1 is
D
Gz for 3. As noted
rac
before, we have observed that B3LYP/6-31G(d) tends to over-
estimate the barriers in conformational reactions of complex aro-
matic molecules,17 but this discrepancy (9.4 kcal/mol) is unusually
large. GAUSSIAN 0318 frequency calculations provide several mea-
sures for comparing the relative energies of minima and transition
states, but the two most useful are (a) the zero-point-corrected
energy and (b) the free energy (that contains both zero-point and
thermal corrections). When these are used to compile tables of
relative energies and relative free energies (
1), the two values rarely differ by more than 1 kcal/mol for con-
formational reactions. Interestingly, in the case of TS1 for 3, Ecorr
(30.6 kcal/mol) is rather lower than G298 (33.2 kcal/mol), although
still higher than the experimental value.
DEcorr and DG298, Table
D
D
2.3. Twist inversion in 9,10,11,20,21,22-
The TS1 structure is very similar to the AM1 transition state
hexaphenyltetrabenzo[a,c,l,n]pentacene (3)
structure5 for this process (that is in close agreement with the
experimental D
Gzrac), but there remains the possibility of an alter-
As observed for compound 1, the X-ray crystal structures of 3
and a dimethyl derivative show that the molecules adopt confor-
mations with approximate D2 symmetry and end-to-end twists of
138–144ꢀ,4,5 and a D2 GS is also indicated by calculations at several
levels of theory.4–6 We have previously reported AM1 calculations
indicating that the enantiomeric D2 ground states of 3 interconvert
via a C2h-symmetric intermediate, by means of enantiomeric C1-
symmetric transition states lying 23.0 kcal/mol above the GS, in
nate pathway for the racemization of 3. For example, there exists
a C2-symmetric structure for compound 3 that is analogous to INT1
for compound 2; this minimum is 20.7 kcal/mol above the GS at the
B3LYP/6-31G(d) level of theory. However, we have so far been
unable to locate any path for the racemization of 3 that proceeds
through this C2 intermediate. If one exists, it may be even more
complex than the enantiomerization pathway for compound 2, but
not necessarily lower in energy than the path illustrated in Figure 3.
excellent agreement with the experimental
D
Gz
of 23.8 kcal/
rac
mol.5
The racemization path for 3 at the B3LYP/6-31G(d) level is a bit
more complex and is illustrated in Figure 3. In this case, the C2h
structure is not an intermediate but a slight transition state (TS2)
connecting enantiomeric C2-symmetric intermediates (INT and
2.4. Phenyl rotations in compounds 1–3
Besides twist inversion the only other conformational reactions
of compounds 1–3 are the rotations of phenyl groups. Transition
B
A
35
TS1 (C1)
eTS1 (C1)
30
25
expt
TS2 (C2h
)
eINT (C2)
INT (C2)
20
5
eGS (D2)
GS (D2)
0
Figure 3. Free energy profile for the twist inversion of 9,10,11,20,21,22-hexaphenyltetrabenzo[a,c,l,n]pentacene (3) calculated at the B3LYP/6-31G(d) level of theory. The ball-and-
stick drawings illustrate the potential minima and transition state structures for this process.

8634
R.A. Pascal, Jr., Q. Qin / Tetrahedron 64 (2008) 8630–8637
experimental values for phenyl rotations in simple hexaar-
ylbenzenes (17–18 kcal/mol19–21) and comparable to the barriers
observed for the rotation of ortho-tolyl groups in crowded hex-
aarylbenzenes and octaarylnaphthalenes (33–38 kcal/mol19,20,22).
To be sure, the resistance provided by the spring-like twisted acene
cores of 1–3 might give rise to such high barriers, but there is no
experimental measurement for any of these processes. In addition,
there is always the possibility that a stepwise pathway for the ro-
tation may operate. For example, the phenyl rotation barrier in the
‘slice’ conformation of 1 (Table 1; illustrated in Section 2.1) cannot
be very high, but one or more high-energy steps may be required to
reach this conformation, itself 22 kcal/mol above the GS. For this
reason, some experimental data bearing on the phenyl rotations is
desirable.
The best candidate for such a measurement, in terms of both
synthesis and analysis, is compound 6 (Scheme 1) in which the
phenyl groups of
3
have been replaced by m-(tri-
fluoromethyl)phenyls. The CF3s are located so that they do not in-
terfere with aryl rotation, but they are large enough that
chromatographic separation of the resulting cis and trans isomers
of 6 might be possible, and CF3s are excellent reporter groups for 19
F
NMR spectroscopy.
CO2Me
(1) LDA
CF3
CF3
(2) H3O+/Δ
2
CF3
O
phenanthrene-
quinone
KOH
CF3
4
O
CF3
5
345 °C 9-bromophenanthrene
CF3
CF3
Figure 4. B3LYP/6-31G(d)-calculated high-energy transition state structures for the
phenyl rotations in compounds 1–3. The solid atoms highlight the bonds about which
the rotations occur.
cis-6
trans-6
Scheme 1.
F3C
CF3
states for these processes are well defined and easily located, and
they are illustrated in Figure 4. In each case the phenyl rotation
causes the acene core of the molecule to increase substantially in
twist. Thus, the B3LYP/6-31G(d)-calculated end-to-end twist of the
GS conformation of 1 is 71.5ꢀ, but in the C2-symmetric phenyl ro-
tation TS structure, the twist is 97.1ꢀ, an increase of 25.6ꢀ. Similarly,
in the phenyl rotation TS for 2, the twist increases by 25.4ꢀ, in the
‘side’ phenyl rotation TS for 3, by 26.8ꢀ, and in the central phenyl
rotation TS for 3, by 11.4ꢀ (along with substantial bending of the
pentacene core).
Compound 6 was synthesized in the usual manner for de-
rivatives of 1.3 The CF3-substituted diphenylacetone 4 was prepared
from the commercially available ester, and 4 was condensed with
phenanthrenequinone to give cyclopentadienone 5. Pyrolysis of 5,
mixed with a bit of 9-bromophenanthrene for ‘solvent’, at 345 ꢀC
gave 6 as a golden crystalline solid in 54% yield.
These phenyl rotation TS structures ‘look just as one would
expect’, in contrast to the rather counterintuitive racemization TS
structures discussed previously. Instead, it is the calculated relative
free energies of the rotation transition states (Table 1) that are
unexpected: 36.7 kcal/mol for 1, 36.5 kcal/mol for 2, and 37.0 and
35.5 kcal/mol for 3. These values are much higher than the
The 19F NMR spectrum of 6 shows two closely spaced resonances
in a 1:1 ratio for the cis and trans isomers, the expected equilibrium
mixture generated by a high temperature synthesis. Unfortunately,
chromatographic trials on a variety of supports failed to resolve the
mixture, so an estimation of the rotation barrier by the thermal
isomerization of one of the isomers was not possible. However,

R.A. Pascal, Jr., Q. Qin / Tetrahedron 64 (2008) 8630–8637
8635
Given the 17 kcal/mol discrepancy between the calculated and
experimental
Gzrots for 1, it is clear that phenyl rotation in 1
D
cannot be the reaction defined by the TS illustrated in Figure 4.
After an additional round of computational searches, a remarkable
TS was discovered for a concerted phenyl rotation and twist in-
version, and this new reaction profile is illustrated in Figure 6. In
this process, as the phenyl rotation begins, the phenyl group moves
‘above’ the polycyclic core, which simultaneously begins to flatten.
In the TS, which possesses Cs symmetry, the phenyl group is exactly
half rotated, and the two benzo groups on the opposite side of the
molecule are in the midst of slipping past the non-rotating phenyl
group in opposite directions. This new TS is only 21.4 kcal/mol
above the GS, in good agreement with the experimental D
Gz for
rot
compound 6. It is important to note that, although this reaction also
results in twist inversion, this has no experimental consequence for
the racemization barrier, because a lower energy pathway for
racemization exists.
In light of these experimental and computational results, it is
fair to question the chemical relevance of the high-energy phenyl
rotation transition states for compounds 2 and 3. Although we have
not conducted searches for alternate phenyl rotation transition
states in 2 and 3, it is obvious that a phenyl rotation in INT2 for the
inversion of 2 (Fig. 2) might be relatively easy, and similar struc-
tures for compound 3, if they exist, might provide lower energy
pathways for phenyl rotations in that molecule. On the other hand,
2 and 3 are even more tightly twisted molecules than 1, and a low-
energy pathway for concerted twist inversion and one or more ring
inversions, as was found for 1, is extremely unlikely. From an
experimental perspective, both the synthesis and NMR analysis of
labeled versions of 2 and 3 would be more complicated than for 1,
because of the proliferation of cis and trans diastereomers, to
a greater or lesser degree, depending on the labeling scheme. We
happily leave such a task to the interested reader.
Figure 5. Variable temperature 19F NMR spectrum (376 MHz, toluene-d8) of com-
pound 6. The small ripples in the lower temperature spectra are due to long-range
1H–19F spin–spin coupling.
NMR spectroscopy of 6 can provide a lower limit for the barrier to
aryl rotation. The fact that two close 19F resonances (Dn¼3 Hz at
376 MHz) are observed for 6 in toluene-d8 at room temperature is
already an indication that the barrier is at least 16 kcal/mol. It was
most surprising to find, however, that coalescence of these reso-
nances occurred at a mere 75 ꢀC (see Fig. 5), indicating that the
barrier to aryl rotation is only 19.2 kcal/mol (from the Gutowsky–
Holm approximation, kc¼6.7 sꢁ1 is at Tc¼348 K, and assuming
a transmission coefficient of 1 for the Eyring equation).
3. Conclusion
The exceptional chiroptical properties of resolved twisted
25
acenes4,5,23 (e.g., for 3, [
a
]
D
¼7400ꢀ) make them interesting can-
didates for some materials applications, but only if the
25
20
15
TS (Cs)
expt
eGS (D2)
GS (D2)
0
Figure 6. Free energy profile for concerted phenyl rotation and twist inversion in 9,18-diphenyltetrabenz[a,c,h,j]anthracene (1) calculated at the B3LYP/6-31G(d) level of theory. The
ball-and-stick drawings illustrate the potential minima and transition state structure for this process, as well as a transitional structure that is not a stationary point, but which helps
to illustrate the process. The solid atoms highlight the trans–cis isomerization that was observed experimentally.

8636
R.A. Pascal, Jr., Q. Qin / Tetrahedron 64 (2008) 8630–8637
configurational stability can be improved. Unfortunately, the con-
formational complexity of these molecules, as revealed by the
present study, makes it difficult to say that any particular structural
modification will ensure configurational stability. Even for the
simplest compound (1), three separate racemization transition
states have been identified, and the situation is not necessarily less
complex for 2 and 3. The fact that compound 3 has a higher ex-
perimental barrier to racemization than 1 makes it tempting to say
that an even longer twisted acene should be configurationally
stable. However, since racemization is a multistep process in the
larger twisted acenes, at some point increasing the length may
cease to have a significant effect on the barriers for the individual
steps, even though they may be more numerous.
collected by filtration (0.18 g, 0.35 mmol, 25%): mp 207–215 ꢀC
(dec); 1H NMR (CDCl3)
6.99 (td, J¼8, 1 Hz, 2H), 7.34 (td, J¼8, 1 Hz,
d
2H), 7.46 (dd, J¼8, 1 Hz, 2H), 7.59 (m, 4H), 7.65 (d, J¼8 Hz, 2H), 7.69
(s, 2H), 7.85 (d, J¼8 Hz, 2H); 13C NMR (CDCl3)
d 121.7, 123.3, 124.8,
125.0 (q, JCF¼4 Hz), 126.1 (q, JCF¼4 Hz), 126.9 (q, JCF¼4 Hz), 127.1,
127.7, 127.8, 128.0, 128.5, 128.6, 128.9, 129.2, 129.4, 130.4 (q,
JCF¼4 Hz),130.8,131.2 (q, JCF¼33 Hz),131.6,132.2,132.8,133.4,133.7,
133.8, 135.1, 138.0, 149.4, 197.1, 199.0 (29 of 32 expected resonances
observed for a mixture of cis and trans isomers; the two CF3s and
one of the attached quaternary C’s were not observed); exact mass
(ESI) 518.1106, calcd for C31H16OF6 518.1105.
4.1.3. 9,18-Bis[3-(trifluoromethyl)phenyl]tetra-
These studies also serve as a reminder (if one was needed) that
computational predictions must be tested by experiment whenever
possible. The high-energy phenyl rotation transition states identi-
fied here are genuine transition states, but they are not necessarily
the only transition states for phenyl rotation, as was revealed by the
experimental rotation barrier determination. Thus the calculations
provide possibilities and insights for structures and mechanisms,
but only by comparison of the computational barriers with exper-
imental barriers can one have confidence that the calculated
transition states and TS structures are chemically relevant. When
they do agree, however, the calculated TS structures provide win-
dows into the reactions that cannot be obtained in any other way.
benz[a,c,h,j]anthracene (6)
Compound 5 (160 mg, 0.309 mmol) and 9-bromophenanthrene
(79 mg, 0.31 mmol) were mixed in a Pyrex screw-capped tube and
heated at 345 ꢀC for 1 h. After cooling, the resulting black solid was
chromatographed on a silica gel column eluted with 25:1 hexanes–
benzene. The desired product (Rf 0.6 by TLC in 2:1 hexanes–ben-
zene) eluted as a yellow band. Concentration of these fractions
yielded pure compound 6 as a yellow crystalline solid (55 mg,
0.083 mmol, 54%): mp 279–283 ꢀC; 1H NMR (CDCl3)
d 7.00 (m, 6H),
7.07 (d, J¼8 Hz, 2H), 7.43 (m, 4H), 7.50 (t, J¼8 Hz, 2H), 7.64 (m, 2H),
7.69 (d, J¼8 Hz, 2H), 7.73 (m, 2H), 8.40 (t, J¼8 Hz, 4H); 19F NMR
(CDCl3)
d
ꢁ63.11, ꢁ63.10; 19F NMR (toluene-d8)
d
ꢁ62.83, ꢁ62.82;
MS (EI) m/z 666 (Mþ, 100), 647 (MꢁF, 2); exact mass (ESI) 666.1772,
4. Experimental
calcd for C44H24F6 666.1782.
4.1. Data for compounds
4.2. General computational methods
All semiempirical (AM111) and density functional [B3LYP/
6-31G(d)13,14] calculations were performed by using GAUSSIAN
03,18 and its default thresholds for wave function and gradient
convergence were employed. Transition states were located by
means of quadratic synchronous transit algorithms (QST2 and QST3
program options) or by optimization under a symmetry constraint.
All potential minima and transition states were verified by ana-
lytical frequency calculations. For a more detailed discussion of the
strategies employed to locate conformational transition states in
a very large, symmetric hydrocarbon, see Ref. 17.
4.1.1. 1,3-Bis[3-(trifluoromethyl)phenyl]acetone (4)
A solution of n-butyllithium (6.9 mL of a 2.5 M solution in
hexanes, 17.2 mmol) was added to a solution of diisopropylamine
(2.43 mL, 17.2 mmol) in ether (40 mL) at 0 ꢀC. Methyl 3-tri-
fluoromethylphenylacetate (2.25 g, 11.5 mmol) was slowly added,
and the enolate solution was stirred for 30 min. A second equiva-
lent of the ester (2.25 g, 11.5 mmol) was added, the ice bath was
removed, and stirring was continued for 16 h at room temperature.
The reaction mixture was poured into 1 M HCl (50 mL) and it was
extracted with ether (3ꢂ50 mL). The combined extracts were
washed with water, dried over Na2SO4, and concentrated to give an
orange oil. This material was taken up in acetic acid (50 mL) and
6 M HCl (10 mL), and the solution was heated at reflux for 8 h.
Heating was discontinued, the solvent was removed, and the resi-
due was dissolved in ether (200 mL). The solution was washed with
water and satd NaHCO3, and dried over Na2SO4. Concentration left
a light orange oil that deposited white needles of pure compound 4
Acknowledgements
This work was supported by National Science Foundation Grant
CHE-0614879, and it is gratefully acknowledged.
Supplementary data
(2.78 g, 8.03 mmol, 70%): mp 48–49 ꢀC; 1H NMR (CDCl3)
d 3.84 (s,
4H), 7.34 (d, J¼7.5 Hz, 2H), 7.39 (s, 2H), 7.44 (t, J¼7.5 Hz, 2H), 7.54 (d,
It consists of a PDF file containing a table of the calculated
absolute energies (E), zero-point-corrected energies (EþZPE), and
free energies (G298) for the structures calculated in this study, and
an ASCII text file containing the atomic coordinates of the mini-
mum and transition state structures calculated at the B3LYP/
6-31G(d) level. Supplementary data associated with this article can
be found in the online version, at doi:10.1016/j.tet.2008.05.030.
J¼7.5 Hz, 2H); 13C NMR (CDCl3)
d
48.9, 123.9 (q, JCF¼272 Hz), 124.2
(q, JCF¼4 Hz), 126.3 (q, JCF¼4 Hz), 129.2, 131.1 (q, JCF¼32 Hz), 132.9,
134.3, 203.7 (9 of 9 expected resonances observed); MS (EI) m/z 346
(Mþ, 3), 327 (MꢁF, 4), 187 (MꢁCF3C6H4CH2, 19), 159 (CF3C6H4CHþ2 ,
100); exact mass (ESI) 346.0794, calcd for C17H12OF6 346.0792.
4.1.2. 1,3-Bis[3-(trifluoromethyl)phenyl]cyclopenta[l]phenanthren-
2-one (5)
References and notes
Phenanthrenequinone (0.287 g, 1.38 mmol) was added to a
solution of compound 4 (0.478 g, 1.38 mmol) in ethanol (10 mL).
KOH (four drops of a 20% solution in EtOH) was added at room
temperature, the quinone dissolved, and the reaction mixture was
then immediately immersed in a boiling water bath with constant
swirling. An additional four drops of KOH solution were added, and,
after heating for 3 min (at which point the solution was boiling),
the reaction mixture was chilled in an ice bath with constant
swirling for 7 min. Dark green crystals of compound 5 were
1. Review: Pascal, R. A., Jr. Chem. Rev. 2006, 106, 4809–4819.
2. Pascal, R. A., Jr.; McMillan, W. D.; Van Engen, D. J. Am. Chem. Soc. 1986, 108,
5652–5653.
3. Pascal, R. A., Jr.; McMillan, W. D.; Van Engen, D.; Eason, R. G. J. Am. Chem. Soc.
1987, 109, 4660–4665.
4. Lu, J.; Ho, D. M.; Vogelaar, N. J.; Kraml, C. M.; Pascal, R. A., Jr. J. Am. Chem. Soc.
2004, 126, 11168–11169.
5. Lu, J.; Ho, D. M.; Vogelaar, N. J.; Kraml, C. M.; Bernhard, S.; Byrne, N.; Kim, L. R.;
Pascal, R. A., Jr. J. Am. Chem. Soc. 2006, 128, 17043–17050.
6. Norton, J. E.; Houk, K. N. J. Am. Chem. Soc. 2005, 127, 4162–4163.

R.A. Pascal, Jr., Q. Qin / Tetrahedron 64 (2008) 8630–8637
8637
7. Smyth, N.; Van Engen, D.; Pascal, R. A., Jr. J. Org. Chem. 1990, 55, 1937–1940.
8. Meurer, K. P.; Vogtle, F. Top. Curr. Chem. 1985, 127, 1–76.
9. Clark, M.;Cramer, R.D., III;VanOpdenbosch, N.J. Comput. Chem.1989,10, 982–1012.
10. Halgren, T. A. J. Comput. Chem. 1996, 17, 490–519.
11. Dewar, M. J. S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J. J. P. J. Am. Chem. Soc. 1985,
107, 3902–3909.
12. Hehre, W. J.; Radom, L.; Schleyer, P. v. R.; Pople, J. A. Ab Initio Molecular Orbital
Theory; John Wiley & Sons: New York, NY, 1986, pp 133–226.
13. Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652.
Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.;
Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,
X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.;
Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.;
Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg,
J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.;
Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.;
Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.;
Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,
C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.;
Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, Revision C.02; Gaussian:
Wallingford, CT, 2004.
14. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785–789.
15. The energies quoted in the text and plotted in Figures 1–3 and 6 are the
calculated relative free energies (
relative energies ( E) and zero-point-corrected energies (
generally similar, may be found in Table 1 for comparison.
D
G298) of the structures under discussion. The
D
DEcorr), which are
19. Gust, D. J. Am. Chem. Soc. 1977, 99, 6980–6982.
20. Gust, D.; Patton, A. J. Am. Chem. Soc. 1978, 100, 8175–8181.
16. Bock, H.; Havlas, Z.; Gharagozloo-Hubmann, K.; Sievert, M. Angew. Chem., Int.
Ed. 1999, 38, 2240–2243.
17. Pascal, R. A., Jr.; Kraml, C. M.; Byrne, N.; Coughlin, F. J. Tetrahedron 2007, 63,
11902–11910.
18. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman,
J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.;
Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.;
21. Komber, H.; Stumpe, K.; Voit, B. Tetrahedron Lett. 2007, 48, 2655–2659.
22. Qiao, X.; Pelczer, I.; Pascal, R. A., Jr. Chirality 1998, 10, 154–158.
23. At the request of a referee, the specific rotations of the D2 ground states of
compounds 1–3 were calculated at the B3LYP/6-31G(d) level (using the dipole
electric field polarizabilities), yielding [
a
]D¼559ꢀ, 460ꢀ, and 1737ꢀ, respectively.
The last is obviously much lower than the experimentally determined value of
7400ꢀ for 3.