Synthesis and characterization of a functionalized chiral biaryl capable of exhibiting unidirectional bond rotation

Article abstract of DOI:10.1016/j.tetlet.2004.10.147

A class of chiral biaryl molecules that have been designed to undergo unidirectional rotation about the aryl-aryl bond may show promise for future application in the area of synthetic molecular motors. These asymmetric molecules should be capable of chann

Full text of DOI:10.1016/j.tetlet.2004.10.147

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 9599–9602
Synthesis and characterization of a functionalized chiral
biaryl capable of exhibiting unidirectional bond rotation
Bart J. Dahl and Bruce P. Branchaud*
Department of Chemistry, Materials Science Institute, Oregon Nanoscience and Microtechnologies Institute,
University of Oregon, Eugene, OR 97403, USA
Received 3 September 2004; revised 22 September 2004; accepted 26 October 2004
Available online 11 November 2004
Abstract—A class of chiral biaryl molecules that have been designed to undergo unidirectional rotation about the aryl–aryl bond
may show promise for future application in the area of synthetic molecular motors. These asymmetric molecules should be capable
of channeling chemical energy into repeated 360° rotation in one direction. Herein we report the synthesis and characterization of
one such biaryl molecule (1).
Ó 2004 Elsevier Ltd. All rights reserved.
Chemists have been very interested in designing mole-
cules that mimic the mechanical behavior of everyday
large scale machines. Synthetic molecular motors, in
which controlled unidirectional molecular motion can
be converted into useful work, are envisioned to be the
central components of future nanoscale machines1 and
may be useful for materials application. The primary
challenge thus far in synthetic molecular motor research
has been in the design and synthesis of molecular sys-
tems capable of controlled unidirectional motion. Fer-
inga and co-workers2–4 have synthesized a class of
chiral molecules that achieve unidirectional bond rota-
tion via light-driven isomerization about sterically
crowded carbon–carbon double bonds. Kelly and co-
workers5,6 have synthesized a triptycyl/helicene-based
phosgene-driven system that can achieve 120° unidirec-
tional bond rotation around the single bond connecting
the triptycene tothe helicene. We have designed a class
of biaryl molecules that should utilize diastereoselective
reactions to undergo repeated unidirectional 360° rota-
tion about the aryl–aryl bond.
Figure 1.
net rotation in one direction versus the other in order for
the system to be useful as a molecular motor. Exclusive
unidirectional rotation is therefore ideal, but not re-
quired. The three processes contributing to directed
bond rotation are (1) ring opening of 1, (2) thermal
isomerization of diastereomers 2aS and 2aR, and (3)
Lactone 1 could be used as the starting material for a
series of three processes that would lead to directed
bond rotation about the aryl–aryl bond (Fig. 1).7 Direc-
ted bond rotation is defined here as energy-driven rota-
tion about a bond in a single direction. There must be a
cyclization of diastereomers 2aS and 2aR toform
1
again. First, addition at lactone 1 can occur at either
diastereoface of the carbonyl yielding either diastereo-
mer 2aS (with S axial chirality) or 2aR (with R axial chi-
rality). Any reaction that generates two diastereomers
will generate one in excess, therefore, ring opening of
Keywords: Molecular motors; Nanomachines; Directed bond rotation.
*
Corresponding author. Tel.: +1 541 346 4627; fax: +1 541 346
0487; e-mail: bbranch@uoregon.edu
0040-4039/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.10.147

9600
B. J. Dahl, B. P. Branchaud / Tetrahedron Letters 45 (2004) 9599–9602
1
will generate either 2aS or 2aR in excess
(k1!2aS 5 k1!2aR). Second, thermal isomerization of
2aS and 2aR should occur by a back-and-forth Ôwind-
shield wiperÕ type of rotation about the aryl–aryl bond
with the C(@O)Nu group passing over aryl hydrogen in-
stead of the sterically bulky cyclohexenol ring. Third,
intramolecular lactonization of either diastereomer re-
sults in the reformation of lactone 1. Diastereomers
inherently cannot react at exactly the same rate
(k2aS!1 5 k2aR!1), soeither 2aS or 2aR will exhibit fas-
ter cyclization. Three examples are given analyzing the
effects these contributing factors would have on the effi-
ciency of the directed bond rotation.
Both ring-opening and ring-closing reactions are highly
diastereoselective: A high efficiency of directed bond
rotation would result if ring opening of 1 is highly dia-
stereoselective to form mainly 2aS, thermal isomeriza-
tion of 2aS and 2aR is fast relative tothe toher
processes, and cyclization of 2aR is faster than the cycli-
zation of 2aS. Similarly, a high efficiency of directed
bond rotation (in the opposite direction) would result
if the opposite diastereoselectivity in ring-opening and
closing was exhibited. Result: excellent motors with very
efficient directed bond rotation.
Figure 2.
nucleophiles should enhance the selectivity of the ring
opening. Thermal isomerization of 2aS and 2aR should
be fast relative to the other processes. Cyclization of 2aR
should be faster than cyclization of 2aS due toa proxim-
ity effect: in 2aR the hydroxyl and the carbonyl are close
to each other and they can easily get within bonding dis-
tance, with the proper Burgi–Dunitz angle for nucleo-
philic attack on the carbonyl. In contrast, 2aS cannot
easily cyclize because the hydroxyl and carbonyl are
not close to each other and the proper geometry for cycli-
zation is difficult to attain. Thus, this three step system
of 1!2aS!2aR!1 would result in a net 360° unidirec-
tional rotation about the aryl–aryl bond.
Only ring opening is highly diastereoselective: A moder-
ate efficiency of directed bond rotation would result if
ring opening of 1 is highly diastereoselective to form
mainly 2aS, thermal isomerization of 2aS and 2aR is fast
relative to the other processes, and cyclization of 2aR
and 2aS proceed at comparable rates. Similarly, a mod-
erate efficiency of directed bond rotation (in the opposite
direction) would result if the opposite diastereoselectivity
in the ring opening was exhibited. Result: good motors
with moderately efficient directed bond rotation.
The synthesis of racemic 1 was achieved in 11 steps
(Scheme 1).9 Aryl iodide 4, prepared via Fischer esterifi-
cation of 3, was successfully coupled in 10min under
high temperature Suzuki conditions10 toafford 5. An in-
dium mediated allylation reaction11 was employed,
resulting in high yields of 6. Triethyl silane reduction
of 6, hydroboration–oxidation of 7, and Jones reagent
oxidation of alcohol 8 were used in succession to afford
9. The acid chloride of 9 was formed with oxalyl chlo-
ride followed by a standard AlCl3 induced Friedel–
Crafts acylation to form biaryl tetralone 10 followed
by ester hydrolysis to 11. Sodium borohydride reduction
of 11 resulted in high yields of 1 after a 5% HCl quench.
The carboxylic acid ring opened 12 was not detected or
isolated under these conditions as determined by TLC,
1H NMR and mass spectrometry.
Only ring closing is highly diastereoselective: A moderate
efficiency of directed bond rotation would result if ring
opening of 1 is not significantly diastereoselective, ther-
mal isomerization of 2aS and 2aR is fast relative tothe
other processes, and cyclization of 2aR is much faster
than 2aS. Similarly, a moderate efficiency of directed
bond rotation (in the opposite direction) would result
if the opposite diastereoselectivity in the ring closing
was exhibited. Result: good motors with moderately effi-
cient directed bond rotation.
The discussion above describes how systems such as the
one shown in Figures 1 and 2 could be designed to create
molecular motors. There is good reason to believe that
specific features that have been designed intothe struc-
ture of lactone 1 should result in an excellent molecular
motor system, with highly diastereoselective ring open-
ing and ring closing. The AM1 minimized models in Fig-
ure 2 illustrate these key features. Nucleophilic additions
to carbonyl groups are known to proceed at the Burgi–
Dunitz angle:8 105° 5° relative tothe plane of the carb-
onyl. Therefore, attack by a nucleophile on lactone 1
should proceed via the unhindered exo face. After going
through an sp3 hybridized intermediate, the subsequent
bond breaking should result in a 90° bond rotation
direction to afford the ring opened 2aS in excess. Larger
In order to study the aryl–aryl bond rotation, the lac-
tone 1 must be effectively opened and closed (Figs. 1
and 2) and the resulting kinetic products (2aS and
2aR) must be isolated before equilibration. Ring-opened
12 has prvoen tobe difficult toisloate after LiOH
hydrolysis of 1 because of its propensity to re-lactonize
back to 1 under acidic work-up conditions. However,
1H NMR of 129 (isolated after a careful mild oxalic acid
quench) has indicated that axial diastereomers are not
observed at room temperature. This is the result of fast

B. J. Dahl, B. P. Branchaud / Tetrahedron Letters 45 (2004) 9599–9602
9601
Scheme 1. Reagents and conditions: (a) MeOH, H2SO4, reflux; (b) 3-formyl-4-methoxyphenyl-boronic acid, tetrabutylammonium bromide, Na2CO3,
Pd(OAc)2, H2O, 150°C; (c) allyl bromide, In, DMF, then 5% HCl; (d) triethylsilane, TFA, CH2Cl2; (e) 1M BH3–THF, then H2O2, NaOH, THF; (f)
CrO3, H2SO4, acetone; (g) oxalyl chloride, cat. DMF, CH2Cl2, 0°C; (h) AlCl3, CH2Cl2, reflux; (i) LiOH, THF/H2O (2:1), reflux, then 5% HCl; (j)
NaBH4, 5% NaOH in H2O; (k) 5% HCl; (l) LiOH, THF/H2O (2:1), reflux, then oxalic acid; (m) 5% HCl; (n) N-O-dimethylhydroxylamine
hydrochloride, 2.0M isopropylmagnesium chloride in THF, then 5% HCl; (o) trifluoroacetic acid, neat.
isomerization between 12aS and 12aR on the NMR time
scale, with undetermined equilibrium. In order to ob-
serve a stable ring-opened compound, lactone 1 was
completely converted to the Weinreb amide 139 in
30min with Me(MeO)NMgCl. The 1H NMR studies
of 13 showed two axial diastereomers at room tempera-
ture due to the bulkier amide group. These diastereo-
mers are present in unequal amounts (Keq = 1.6 in
DMSO at room temperature). Coalescence of the dia-
stereomers was observed via 1H NMR at approximately
85°C in DMSO and the barrier to Ôwindshield wiperÕ
rotation was estimated to be approximately 17kcal/
mol. The Weinreb amide 13 was readily lactonized back
to 1 in the presence of neat trifluoroacetic acid, with a
complete reaction occurring in 45min.
Me(MeO)NMgCl and closed with 5% HCl or TFA acti-
vation (instead of a more typical activation with DCC
and DMAP) greatly simplifies the system. One pot itera-
tive bond rotations may be possible by cycling of basic
(nucleophilic) conditions to ring open and acidic condi-
tions to lactonize. Current efforts toward obtaining the
best possible conditions for selective opening and clos-
ing of the chiral lactone 1 are being explored.
Acknowledgements
We are grateful toNSF-CHE-9986926, PRF
#
35372AC-1, the US Department of Education GAANN
P200A010222-03, and the NSF-IGERT DGE-0114419
for financial support. We thank Jonathan Bingham for
some preliminary contributions to this work.
Because of the fast equilibrium occurring between 12aS
and 12aR and 13aS and 13aR, the kinetic products of
ring opening cannot be isolated and the initial direction
of rotation upon ring opening is currently undetermin-
able. The rate of diastereomer isomerization could be
slowed down by introducing more steric hindrance to
bond rotation by using bulkier nucleophiles or by
replacing the ortho C–H on the lower ring with a larger
group. The discovery that 1 can be opened with LiOH or
References and notes
1. Balzani, V.; Credi, A.; Raymo, F. M.; Stoddart, J. F.
Angew. Chem., Int. Ed. 2000, 39, 3348–3391.
2. Koumura, N.; Geertsema, E. M.; van Gelder, M. B.;
Meetsma, A.; Feringa, B. L. J. Am. Chem. Soc. 2002, 124,
5037–5051.

9602
B. J. Dahl, B. P. Branchaud / Tetrahedron Letters 45 (2004) 9599–9602
3. van Delden, R. A.; ter Wiel, M. K. J.; de Jong, H.;
Meetsma, A.; Feringa, B. L. Org. Biomol. Chem. 2004, 2,
1531–1541.
4. Koumura, N.; Zijistra, R. W. J.; van Delden, R. A.;
Harada, N.; Feringa, B. L. Nature (London) 1999, 401,
152–155.
5. Kelly, T. R.; De Silva, H.; Silva, R. A. Nature 1999, 401,
150–152.
6. Kelly, T. R.; Silva, R. A.; De Silva, H.; Jasmin, S.; Zhao,
Y. J. Am. Chem. Soc. 2000, 122, 6935–6949.
7. One enantiomer of 1 is shown in Figure 1, but the analysis
of diastereoselective processes would work the same for
either enantiomer or a racemic mixture.
8. Burgi, H. B.; Dunitz, J. D.; Lehn, J. M.; Wipff, G.
Tetrahedron 1974, 30, 1563–1572.
6.88 (d, J = 8Hz, 1H), 5.96–5.87 (m, 1H), 5.09–4.97 (m,
2H), 3.88 (s, 3H), 3.69 (s, 3H), 2.75 (t, J = 7Hz, 2H), 2.37
(t, J = 7Hz, 2H). Spectroscopic data for 8: 1H NMR
(300MHz, CDCl3): 7.79 (d, J = 8Hz, 1H), 7.51 (t,
J = 8Hz, 1H), 7.39–7.35 (m, 2H), 7.17–7.10 (m, 2H),
6.89 (d, J = 8Hz, 1H), 3.87 (s, 3H), 3.69 (s, 3H), 2.68 (t,
J = 7Hz, 2H), 1.70–1.55 (m, 4H). Spectroscopic data for
9: Rf (1:1 EtOAc/hexanes) 0.44; IR (KBr, cmꢀ1): 3430,
1724, 1706; 1H NMR (300MHz, CDCl3): 7.79 (d,
J = 8Hz, 1H), 7.52 (t, J = 7Hz, 1H), 7.39–7.36 (m, 2H),
7.16 (d, J = 8Hz, 1H), 7.09 (s, 1H), 6.88 (d, J = 8Hz, 1H),
3.87 (s, 3H), 3.71 (s, 3H), 2.71 (t, J = 8Hz, 2H), 2.40
(t, J = 7Hz, 2H), 1.97 (q, J = 7Hz, 2H). Spectroscopic
data for 10: 1H NMR (300MHz, CDCl3): 8.00 (d,
J = 9Hz, 1H), 7.53 (t, J = 7Hz, 1H), 7.39 (t, J = 8Hz,
1H), 7.14 (d, J = 7Hz, 1H), 7.04–6.96 (m, 2H), 3.91 (s,
3H), 3.65 (s, 3H), 2.98 (t, 2H), 2.54 (t, 2H), 2.13 (t, 2H).
9. Spectroscopic data for 1: Rf (1:1 EtOAc/hexanes) 0.76; 1H
NMR (300MHz, CDCl3): 7.97 (d, J = 8Hz, 1H), 7.67–
7.56 (m, 2H), 7.50–7.7.44 (m, 2H), 6.98 (d, J = 8Hz, 1H),
5.15 (t, J = 4Hz, 1H), 3.92 (s, 3H), 3.11–3.03 (m, 1H),
2.49–2.37 (m, 2H), 2.08–1.82 (m, 3H); ESI MS m/z (M+1)
calcd 281.11, obsd 281.0. Spectroscopic data for 4: Rf
(CH2Cl2) 0.52; 1H NMR (300MHz, CDCl3): 8.00 (d,
J = 8Hz, 1H), 7.81 (d, J = 8Hz, 1H), 7.40 (t, J = 8Hz,
1H), 7.17 (t, J = 8Hz, 1H), 3.94 (s, 3H). Spectroscopic
data for 5: Rf (1:1 hexanes/CH2Cl2) 0.19; 1H NMR
(300MHz, CDCl3): 10.52 (s, 1H), 7.89 (d, J = 7Hz, 1H),
7.82 (s, 1H), 7.57–7.51 (m, 2H), 7.43 (t, J = 7Hz, 1H), 7.36
(d, J = 8Hz, 1H), 7.04 (d, J = 9Hz, 1H), 4.00 (s, 3H), 3.72
(s, 3H). Spectroscopic data for 6: Rf (CH2Cl2) 0.44; 1H
NMR (300MHz, CDCl3): 7.80 (d, J = 7Hz, 1H) 7.52 (t,
J = 7Hz, 1 H),7.41–7.33 (m, 3H), 7.24 (d, J = 6Hz, 1H),
5.96–5.82 (m, 1H), 5.19–5.14 (m, 1H), 5.12–5.04 (m, 1H),
4.00 (s, 3H), 3.72 (s, 3H), 2.66–2.55 (m, 2H), 1.77 (br s,
1H). Spectroscopic data for 7: Rf (CH2Cl2) 0.67; 1H NMR
(300MHz, CDCl3): 7.77 (d, J = 8Hz, 1H), 7.52 (t,
J = 7Hz, 1H), 7.40–7.36 (m, 2H), 7.18–7.11 (m, 2H),
1
Spectroscopic data for 11: H NMR (300MHz, CDCl3):
9.5 (br s, 1H), 8.02 (d, J = 8Hz, 1H), 7.54 (t, J = 7Hz, 1H),
7.39 (t, J = 7Hz, 1H), 7.13 (d, J = 7Hz, 1H), 7.02–6.93 (m,
2H), 3.91 (s, 3H), 2.95 (t, J = 8Hz, 2H), 2.51 (t, J = 7Hz,
2H), 2.10 (q, J = 7Hz, 2H). Spectroscopic data for 12: 1H
NMR (300MHz, CDCl3): 7.82 (d, J = 7Hz, 1H), 7.53–
7.42 (m, 2H), 7.21 (d, J = 7Hz, 1H), 6.95 (d, J = 8Hz, 1H),
6.81 (d, J = 8Hz, 1H), 4.92 (s, 1H), 3.87 (s, 3H), 3.07–3.01
(m, 1H), 2.50–2.40 (m, 2H), 1.99–1.82 (m, 3H); APCI MS
m/z (Mꢀ1) calcd 297.11, obsd 297.1. Spectroscopic data
for 13: Rf (1:1 EtOAc/hexanes) 0.33; VT 1H NMR
(300MHz, DMSO-d6, 150°C): 7.41 (s, 4H), 6.92 (d,
J = 8, 1H), 6.83 (d, J = 8, 1H), 4.53 (s, 1H), 3.83 (s, 3H),
3.48 (s, 3H), 3.01 (s, 3H), 2.95–2.81 (m, 2H), 2.14–1.56 (m,
4H); ESI MS m/z (Mꢀ17) calcd 324.2, obsd 324.2.
10. Leadbeater, N. E.; Marco, M. J. Org. Chem. 2003, 68,
888–892.
11. Nair, V.; Ros, S.; Jayan, C. N.; Pillai, B. S. Tetrahedron
2004, 60, 1959–1982.