Synthesis and rotation barriers in 2, 6-Di-(o-anisyl) anisole

Article abstract of DOI:10.1002/poc.2939

Variable temperature ¹H NMR spectroscopic studies of 2, 6-di(o-anisyl) anisole show syn and anti atropisomers at low temperature. The barrier for interconverting these isomers by rotation about the aryl-aryl bond, found by fitting the experimen

Full text of DOI:10.1002/poc.2939

Research Article
Received: 5 January 2012,
Revised: 1 March 2012,
Accepted: 6 March 2012,
Published online in Wiley Online Library: 2012
(wileyonlinelibrary.com) DOI: 10.1002/poc.2939
Synthesis and rotation barriers in 2,
6-Di-(o-anisyl) anisole
Takuhei Yamamotoa, Pi-Yu Chena, Guangxin Lina, Anna Błoch-Mechkourb,
Neil E. Jacobsena, Thomas Ballyb* and Richard S. Glassa*
Variable temperature 1H NMR spectroscopic studies of 2, 6-di(o-anisyl) anisole show syn and anti atropisomers at low
temperature. The barrier for interconverting these isomers by rotation about the aryl-aryl bond, found by fitting the
experimental data, is 41.2 kJ/mol. Copyright © 2012 John Wiley & Sons, Ltd.
Supporting information may be found in the online version of this paper
Keywords: conformational analysis; dynamic NMR; quantum chemical calculations
The 2D NMR spectra are found in the Supporting Information.
The two singlets at d 3.143 and 3.816 were assigned to hydrogens
INTRODUCTION
We recently reported[1] atropisomerism in certain m-terphenyl
thio- seleno- and telluro- ethers, such as 1a-c. The barriers for
interconversion of the two syn and the anti atropisomers,
respectively, were determined by variable temperature NMR
spectroscopic studies.
of the methoxy groups on the central ring and the outer ring, re-
spectively, based on their chemical shifts and integrations. The
methoxy protons showed correlations in the heteronuclear single
quantum coherence (HSQC) spectrum, which allowed assignment
of the signals at d 56.00 and d 60.75 in the 13C NMR spectrum to
the outer ring and the center ring, respectively. The COSY correla-
tions indicated the protons at d 7.147 and 7.231 are in the same
ring and are assigned to H4 and H3,5 on the basis of the AB2 split-
ting pattern and relative integrated areas. Crosspeaks in the HSQC
spectrum of H4 with C4 and H3,5 with C3,5 allowed the assign-
ment of the signal at d 123.17 and that at d 131.66 to C4 and
C3,5, respectively. The heteronuclear multiple bond correlation
(HMBC) spectrum showed the methoxy protons on the center ring
and on the outer ring correlates with quaternary carbons bound
to each of the oxygens and hence assigned to C1 (d 156.72)
and C2′ (d 157.45), respectively. The signal at d 132.54 in the 13C
NMR spectrum showed correlation with H4, but because there
was no crosspeak in HSQC, this absorption was assigned to C2,6.
The barriers for this interconversion increase on going from S
to Se to Te that is, 1a! 1b! 1c. Calculations on 1a and 1b
revealed that these barriers are due to rotation about the aryl-aryl
bonds. Furthermore, the magnitude of this barrier is determined by
the increase in the size of the chalcogen (S<Se<Te) and not by
that of the length of the aryl-chalcogen bond (S-C<Se-C<Te-C).
Notably absent from this study was the first-row chalcogen:
oxygen. Since little work has been reported[2–5] in these and re-
lated systems involving oxygen, we have synthesized and studied
anisole derivative 1d and the results are reported in this note.
1
In the H NMR spectrum, on the basis of the chemical shifts due
to the methoxy group and the observed splitting patterns, H3′,
H5′, H6′, and H4′ correlations in the HSQC spectrum, could be
assigned to the signals at d 7.016, 7.018, 7.293, and 7.357, respec-
tively. On the basis of correlations of these signals in HSQC C3′
(d 111.34), C5′ (d 120.79), C6′ (d 131.85), and C4′ (d 129.23), signals
could be assigned as shown. These assignments were further con-
firmed by crosspeaks in the ROESY spectrum of H3′/OCH3 in the
RESULTS AND DISCUSSION
Cram and coworkers[6] first synthesized 1d, from bromodibenzofuran
and Dyker and Kellner[7] made it in modest yields in a mixture of
products formed in an interesting Pd-catalyzed cascade reaction from
2-iodoanisole. We synthesized 1d, using a Suzuki coupling reaction[8]
as shown below:
* Correspondence to: Richard S. Glass, Department of Chemistry and Biochemistry,
The University of Arizona, Tucson, AZ 85721, USA.
E-mail: rglass@email.arizona.edu
Thomas Bally, Department of Chemistry, University of Fribourg, CH-1700,
Fribourg, Switzerland.
E-mail: thomas.bally@unifr.ch
a T. Yamamoto, P.-Y. Chen, G. Lin, N.E. Jacobsen, R.S. Glass
Department of Chemistry and Biochemistry, The University of Arizona, Tucson
AZ 85721, USA
The 1H and 13C NMR spectra for 2,6-di-(o-anisyl)anisole were
fully assigned on the basis of analysis of 1D and 2D NMR spectra.
b A. Błoch-Mechkour, T. Bally
Department of Chemistry, University of Fribourg, CH-1700, Fribourg, Switzerland
J. Phys. Org. Chem. (2012)
Copyright © 2012 John Wiley & Sons, Ltd.

T. YAMAMOTO ET AL.
Interestingly, only one peak is observed for the methoxy group
appended to the central ring in 1d at the lowest temperature
recorded, whereas 1a-c show two signals for the SMe, SeMe
and TeMe moieties, respectively, below their coalescence tem-
perature. Apparently, the difference in chemical shifts for the
OMe groups of the central ring for the syn and anti atropisomers
is not resolved in our experiments. Note that, as before[1] there
are three possible atropisomers: one anti and two syn isomers[10]
but evidence is obtained for only two. This presumably results
from rapid rotation about the aryl-O bond of the central ring.
Molecular Modeling
The structures of all three atropisomers and the transition states
between them were optimized by the B3LYP/6-31G* and the
B2PLYP-D/cc-pVDZ methods. The calculated structures are depicted
in the Figure 3 (and in Table S2 of the Supporting Information, along
with the three important dihedral angles). It is worth noting that
there are two transition states for the automerization of the anti
conformers which are attained when the central-OMe group rotates
towards one or the other lateral-OMe group. However, the
energy difference between these two transition states is less than
3 kJ/mol, and we will only consider the lower of the two, which
appears in Figure 3. (In Figure S8 and Table S1 of the Supporting
Information both transition states are shown). There is only one
transition state between the two syn conformers because both of
them have a plane of symmetry.
Figure 1. Variable temperature 300 MHz1H NMR spectra of anisole1d in
CD2Cl2 (OMe region)
The relative enthalpies and free energies of the minima and
transition states shown in Figure 3 are listed in Table 1. They were
calculated both in the gas phase and in CH2Cl2 as a dielectric con-
tinuum solvent, but the results obtained from the two sets of calcu-
lations do not differ significantly. They show that the barriers for
rotation around the central Ph-OMe bond, i.e. for the syn-a ! syn-s
interconversion and for the automerization of the anti conformer,
are too low for these processes to be detected in the temperature
range that is accessible in our NMR experiments, i.e. rapid equili-
bration between the syn-a and syn-s conformers is to be expected.
We have previously reported[1] that these barriers decrease on
going from X = S (1a) to X = Se (1b) and we had attributed this to
the increase of the C-X bond length, which leads to a concomitant
increase of the distance between the Me that is attached to chalco-
gen and the lateral phenyl group. By this logic the barrier for rota-
tion of the lateral phenyl group should be the highest for X = O
(1d), but our present results show that this is not the case (the
gas phase activation enthalpy for the the syn-a ! syn-s intercon-
version, calculated by the same method, is 19.4 kJ/mol for 1a
and 16.5 kJ/mol for 1b. The corresponding numbers for the anti
automerization are 19.0 and 15.3 kJ/mol, respectively).
We attempted to single out a reason for these peculiar trends by
model calculations on the fragments from which compounds 1 are
composed. These calculations showed that the steric repulsion
between the Me group and the lateral phenyl group changes by
less than 1 kJ/mol along the series. The distortion of the central
Ph-X-Me moiety in the transition state, relative to optimized struc-
tures, does show an increase on going from X = O to X = S and a
decrease on going to X = Se, but this effect accounts only for about
a third of the total change. Thus we conclude that the observed
trends are due to a combination of causes which cannot be easily
disentangled.
Figure 2. Experimental and simulated 1H NMR OMe peaks of Ar moieties
in 1d in CH2Cl2 at ꢀ79ꢁC (C, see text for parameters)
outer ring and correlations in the HMBC spectrum of H3′ and
H5′/C1′ and H6′/C2,6. Finally, H1′/C1′ was assigned on the basis
of no crosspeaks in the HSQC and correlation with H3,5 in the
HMBC spectrum.
The variable temperature 1H NMR spectra of 1d are shown in
Figure 1. At room temperature the o-anisyl methoxy signals give
rise to a singlet at 3.8 ppm and their equivalence connotes rapid
rotation about the aryl-aryl bond at this temperature. However,
as the temperature is lowered the signal broadens and
ultimately splits into two peaks. Analysis of the coalescence data
provided an excellent fit of the observed and calculated spectra
as shown in Figure 2. The fit around the critical coalescence
temperature (ꢀ79 ꢁC) gave the following fitting parameters
and calculated barrier: line-width 1.5 Hz, nA - nB 11.2 Hz, kAB + kBA
¼
32.3 s-1, %A 54.7, ΔG (A ! B) = 41.2 kJ/mol. As expected the
barrier for 1d is substantially less than that reported for 1a-c.
As reported previously for these latter compounds[1], a detailed
analysis of the barrier was required for 1d, because the variable
temperature NMR spectra show that the chemical shifts vary
significantly with temperature. Consequently, the WinDNMR
program,[9] which was used to obtain the parameters by fitting
the ꢀ79 C data and, thus, to determine the activation barrier
for interconversion of the atropisomers.
On the other hand, the calculated barriers for converting the
syn-a or syn-s to the anti conformers, i.e., for rotation of a lateral
phenyl group, are much higher than the barriers for rotating
wileyonlinelibrary.com/journal/poc
Copyright © 2012 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. (2012)

2,6-DI-(O-ANISYL)ANISOLE ATROPISOMERS
Figure 3. Interconversion of atropisomers of 1d (structures from B2PLYPD/cc-pVDZ calculations; note that the two anti structures are identical).
¼
the central O-Me moiety, and the calculated ΔG are in excel-
lent accord with the experimental free energy of activation of
41.2 kJ/mol. For 1d these barriers are lower than for 1a and 1b as
it was expected according to interpretation given previously.[1]
Based on the geometries from B2PLYPD/cc-pVDZ calcula-
groups are within 0.1 ppm of the experimental values, while
for the central -OMe groups they are off by about 0.2 ppm.
If we assume that rotation around the central Ph-OMe bond is
rapid on the NMR timescale, i.e. the signals of the two syn confor-
mers are averaged, even at ꢀ84ꢁC, then the signals for the
central OMe group should almost coincide (2.89 ppm for the
averaged syn-, and 2.91 ppm for the anti conformer), which
may explain why only one signal is seen for these protons. With
1
tions, the H NMR chemical shifts for the O-Me moieties were
calculated following the protocol outlined in ref 21 (see
Table 2). The predicted chemical shifts for the lateral -OMe
Table 1. Relative Energies in kJ/mola of Different Atropisomers of 1d and Barriers for Their Interconversion (See Figure 3 and Table
S1 in the Supporting Information for Structures).
gas phase
SCRF @CH2Cl2
conformer
syn-s
syn-a
Hrel
Grel
Hrel
1.39
(0)
0.10
12.16
32.46
33.61
13.01
Grel
1.13
0.96
(0)
13.90
32.70
33.82
15.09
0.44
1.16
(0)
20.90
42.57
43.88
21.84
0.60
0.10
(0)
19.16
42.33
43.67
19.75
anti
TS(syn-a ! syn-s)
TS(syn-s ! anti)
TS(anti ! syn-a)
TS(anti ! anti)
aCalculated by the B2PLYP-D/cc-pVDZ method, using B3LYP/6-31G* thermal corrections and entropies.
Table 2. Calculated1H NMR Spectroscopic Chemical Shifts for O-Me groups in 1d a
isomer
d O-Me (central) (ppm)
d O-Me (1) (ppm)
d O-Me (2) (ppm)
syn-a
syn-s
anti
2.85
2.94
2.91
3.90
3.81
3.81
3.90
3.81
3.85
ausing the GIAO/ WP04/cc-pVDZ method with CH2Cl2 as a continuum solvent, and the scaling factors given in ref 21.
J. Phys. Org. Chem. (2012)
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

T. YAMAMOTO ET AL.
regard to the OMe groups on the lateral phenyl rings, averaging
is predicted to lead to a signal at 3.83 ppm for the for the anti
conformer and one at 3.86 ppm for the two syn-conformers,
under conditions where the syn- and anti-conformers do not
interconvert rapidly on the NMR time scale.Although the abso-
lute chemical shifts are off by ca. 0.08 ppm (which is well within
expectations from these calculations), their difference is in good
accord with the observed Δd = 0.026 ppm. The fact that the anti-
conformer is predicted to predominate in the equilibrium, may
explain why at ꢀ84ꢁC the signal at lower field is more intense
than the other one (see Fig. 2).
120.79, 123.17, 128.61, 129.23, 131.66, 131.85, 132.54, 156.72,
157.45; IR (KBr) 2934, 2833, 1581, 1491, 1465, 1272, 1237, 1025,
756 cm-1; HRMS (GC MS EI+, m/z): Calcd for C21H20O3, 320.1412;
Found: 320.1412.
Variable temperature NMR studies
CD2Cl2 was used as the solvent. The 1H NMR exchange line fitting
results were obtained using WinDNMR version 7.1.14 (Reich, H.J.,
University of Wisconsin, Madison, WI). The spectra used for fitting
were acquired just below the coalescence temperature for the
resonance being fitted in order to maximize the information
content. Line widths were determined by fitting peaks well
below the coalescence temperature where exchange broaden-
ing is negligible. Chemical shifts and relative peak areas of the
two conformers, as well as the exchange rate constants were
obtained by visual fit to the experimental data. The conformer
present in greater concentration just below coalescence is
referred to as the “A” conformer, and the other conformer is
described as “B”. The free energy of activation was calculated
using the equation: ΔG{ = [23.76 - ln(k/T)]RT.[15]
METHODS
General
All reactions were performed using standard Schlenk techniques
under an atmosphere of argon. Dioxane was purified by distilla-
tion under N2 from potassium-benzophenone ketyl. Column chro-
matography was done using 32–63 m flash silica gel following the
1
method of Still et al.[11] All mp are uncorrected. All H variable-
temperature and 13C NMR spectra were obtained using an
NMR spectrometer operating at a 1H frequency of 299.956
MHz, using a 5 mm Four-Nucleus probe. Theꢁambient temper-
ature without heating or cooling was 22–23 C. NMR chemical
shifts and coupling patterns in the aromatic rings were
elucidated by simulation and curve fitting using WinDNMR
version 7.1.12 (Reich, H. J., University of Wisconsin, Madison, WI).
Low-temperature experiments used dry nitrogen gas cooled to
77 K in a heat exchanger, and temperatures were calibrated using
the 1H shift separation of a methanol sample: T(ꢁK) = 409.0 -
36.54 x - 21.85 x2 where x is the chemical shift difference in ppm
Quantum chemical calculations
Geometry optimizations and frequency calculations were first
done with the B3LYP/6-31G* method, where all stationary
points (minima, transition states) were properly characterized
and thermodynamic functions were calculated. Geometries were
then reoptimized with the recently introduced “double hybrid”
B2PLYP functional,[16] corrected for contributions of dispersion
energy,[17,18] which accounts much better for steric interactions
than pure DFT functionals do. As analytic second derivatives were
not yet available for the B2PLYP method, we used the thermal cor-
rections and entropies from B3LYP calculations in the evaluation of
the relative enthalpies and free energies listed in Table 1. Finally,
single-point SCRF calculations in CH2Cl2 as a continuum solvent
were done, with the B2PLYP-D functional at the corresponding
geometries, to evaluate relative enthalpies and free energies in
solution. NMR chemical shifts were calculated using the GIAO
method and the WP04 functional that was developed with that
purpose in mind,[19,20] using the cc-pVDZ basis set. SCRF calcula-
tions were again carried out in CH2Cl2 as solvent and the raw
isotropic magnetic shieldings (IMS) were converted to chemical
shifts relative to TMS d by scaling them with the parameters
elaborated recently by Jain et al.[21] (d = 31.844 – (IMS/1.0205)).
All calculations were done with the Gaussian program package.[22]
1
between CH3 and OH proton resonances.[23] All H NMR spectra
are referenced to residual solvent at d 5.32 ppm (CHDCl2), and all
13C NMR spectra are referenced to deuterated at 54.00 (CD2Cl2).
High resolution MS were obtained by direct insertion.
Synthesis of 2,6-di(o-anisyl)anisole, 1d
The method reported by Azzena and co-workers[12] for Suzuki
coupling was used with slight modification. To a solution of
1,3-dibromo-2-methoxybenzene[13,14] (300 mg, 1.13 mmol) in
1,4-dioxane (1.5 mL) under Ar, were added a 2M aqueous solu-
tion of Na2CO3 (1.13 mL, 2.26 mmol), LiCl (153 mg, 3.61mmol),
2-methoxyphenylboronic acid (686 mg, 4.5 mmol), and Pd
(PPh3)4 (130 mg, 0.113 mmol). This mixture was stirred and
heated at reflux for 24 h. After cooling to room temperature,
H2O (20 mL) was added to the suspension. The aqueous layer
was extracted with EtOAc (3 X 20 mL). The combined organic
layers were washed successively with 1M aqueous NaOH
(50 mL), brine (50 mL), and H2O (50 mL). The organic layer was
dried (MgSO4) and then concentrated under reduced pressure.
The residue was purified by silica gel chromatography using
hexanes:chloroform (1:1) and then chloroform to elute 1d as
a white solid (160 mg, 44%). The product was further purified
by recrystallization twice from diethyl ether:hexanes (1:1) : mp
Acknowledgements
We gratefully acknowledge financial support of this work by the
U.S. National Science Foundation (CHE-0956581, R.S.G.) and the
Swiss National Science Foundation (Project 2000–121747, T.B.).
REFERENCES
[1] U. I. Zakai, A. Błoch-Mechkour, N. E. Jacobsen, A. Abrell, G. Lin, G. S.
Nichol, T. Bally, R. S. Glass, J. Org. Chem. 2010, 75, 8363–8371.
[2] H. O. House, J. A. Hrabie, D. VanDerveer, J. Org. Chem. 1986, 51, 921–929.
[3] L. Lunazzi, M. Mancinelli, A. Mazzanti, J. Org. Chem. 2007, 72, 5391–5394
[4] L. Lunazzi, M. Mancinelli, A. Mazzanti. J. Org. Chem. 2009, 74, 1345–1348.
[5] D. Casarini, L. Lunazzi, A. Mazzanti, Eur. J. Org. Chem. 2010, 2035–2056.
[6] D. J. Cram, I. B. Dicker, M. Lauer, C. B. Knobler, K. N. Trueblood, J. Am.
Chem. Soc. 1984, 106, 7150–7167.
1
118-119ꢁC (lit.3 117-118ꢁC); H NMR (500 MHz, CD2Cl2) d 3.143
(s, 3H), 3.816 (s, 6H), 7.016 (dd, J = 8.3, 1.2 Hz, 2H), 7.018
(dt, J = 1.1, 7.6 Hz, 2H), 7.147 (A, 1H) and 7.231 (B, 2H), AB2 system
(JAB = 7.6 Hz), 7.293 (dd, J = 7.7, 1.8 Hz, 2H), 7.357 (ddd, J = 8.2, 7.5,
1.8 Hz, 2H). 13C NMR (125 MHz, CD2Cl2) d 56.00, 60.75, 111.34,
[7] G. Dyker, A. Kellner, J. Organomet. Chem. 1998, 555, 141–144.
wileyonlinelibrary.com/journal/poc
Copyright © 2012 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. (2012)

2,6-DI-(O-ANISYL)ANISOLE ATROPISOMERS
[8] N. Miyaura, A. Suzuki, Chem. Rev. 1995, 95, 2457–2483.
[9] H. J. Reich, WinDMR version 7.1.12, University of Wisconsin, Madison, WI.
[10] In one syn isomer the O-Me group of the central ring is facing away
from the MeO groups and in the other the O-Me group is facing
toward the MeO groups.
[11] W. C. Still, M. Kahn, A. Mitra, J. Org. Chem. 1978, 43, 2923–2925.
[12] J. Azzena, M. Cattari, G. Melloni, L. Pisano, Synthesis 2003, 2811–2814.
[13] F. F. Blicke, F. D. Smith, J. L. Powers, J. Am. Chem. Soc. 1932, 54, 1465–1471
[14] J. M. W. Chan, T. M. Swager, Tetrahedron Lett. 2008, 49, 4912–4914.
[15] H. Günther, NMR Spectroscopy, 2nd ed., Wiley, New York, 1992; 343.
[16] S. Grimme, J. Chem. Phys. 2006, 124, 034108/1–034108/16.
[17] T. Schwabe, S. Grimme, Phys. Chem. Chem. Phys. 2007, 9, 3397–3406.
[18] for an overview of these methods, see: T. Schwabe, S. Grimme, Acc.
Chem. Res. 2008, 41, 569–579.
[19] K. W. Wiitala, T. R. Hoye, C. J. Cramer, J. Chem. Theory Comput. 2006, 2,
1092.
[20] In Gaussian, the WP04 functional is invoked by specifying the BLYP
keyword and adding iop(3/76=1000001189,3/77=0961409999,
3/78=0000109999) to the keyword line.
[21] R. Jain, T. Bally, P. R. J. Rablen, Org. Chem. 2009, 74, 4017–4023.
[22] M. J. Frisch et al. Gaussian 03, Revision E.01, and Gaussian 09,
Revision A.02, 2003, Gaussian, Inc., Wallingford, CT
[23] C. Ammann, P. Meier, A. E. Merbach, J. Magn. Reson. 1982. 46,
319–321.
J. Phys. Org. Chem. (2012)
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc