A rotary molecular motor gated by electrical energy

Article abstract of DOI:10.1021/ol502946v

DFT calculations predict that the chiral pentiptycene derivative E-1 possesses distinct rotational potential energy surfaces in the neutral vs the radical anionic (E-1^?-) form such that continued electrochemical switching between E-1 and E-1

Full text of DOI:10.1021/ol502946v

Letter
pubs.acs.org/OrgLett
A Rotary Molecular Motor Gated by Electrical Energy
†
‡
,‡
,†
Chen-Yi Kao, Hsiu-Feng Lu, Ito Chao,* and Jye-Shane Yang*
†Department of Chemistry, National Taiwan University, Taipei, Taiwan 10617
‡Institute of Chemistry, Academia Sinica, Taipei, Taiwan 11529
*
S Supporting Information
ABSTRACT: DFT calculations predict that the chiral pentiptycene
derivative E-1 possesses distinct rotational potential energy surfaces in
•−
the neutral vs the radical anionic (E-1 ) form such that continued
•−
electrochemical switching between E-1 and E-1 could lead to a
directional rotation of the pentiptycene rotor about the exocyclic C−C
6
bond. The time scale of random Brownian rotation is ∼10 s for E-1 and
•
−
∼
170 s for E-1 at 150 K, and thus a switching time scale of 0.2 s could
readily bias the rotation direction to >99% at 150 K. The synthetic
feasibility, line-shape analysis on the VT H NMR spectra, and
1
electrochemical redox switching of E-1 are demonstrated.
olecular machines perform designated work at the
molecular level. For a (supra-)molecular system to do
the radical anion intermediates that thermodynamically favor the
•−
•−
M
E isomer (i.e., Z-2 → Z-2 →E-2 → E-2). The moderate
4
−1
real work (i.e., a molecular motor), the motional subunit must be
able to break the detailed balance of Brownian motions, namely,
to undergo biased or directional motion in response to the input
rotation rate at room temperature (∼2 × 10 s ), high
electrochemical stability, and tunable structure for E-2 provide
1,12
a unique opportunity toward a Brownian molecular motor.
1
,2
of energies. While naturally occurring molecular motors such3
as ATP synthase, kinesin, and myosin adopt chemical fuels,
artificial counterparts could be driven by other types of energy
We reasoned that enantioselective substitution at the indanone
methylene group could have two effects: (1) creating a chirality
center and thus a ratchet rotational potential energy surface
1
,2
such as light and electrical energy. To date, only a few4
(
RPES); (2) increasing the torsion barrier and thus minimizing
prototypes of rotary molecular motors have been reported,
the random Brownian rotation at ambient temperatures.
Provided that the RPES for the neutral and the anion radical
states of E-1 differ in a way that biased rotation can be achieved
5
including Feringa’s light-driven tetrasubstituted alkenes and
6
chemical-driven lactone system, Leigh’s macrocycle system
7
powered by both light and chemicals, Lehn’s light-driven diaryl-
•−
through electrochemical switching between E-1 and E-1 ,
8
N-alkyl imines, and Rapenne−Joachim−Hla’s ruthenium
directional rotation of E-1 gated by electrical energy would be
9
complex powered by electrical energy. For those fueled by
1
achieved via the “Pulsating Ratchet” mechanism.
9
,10
electrical energy, the operation is on a single molecule with an
STM tip positioned over the molecule or at a specific site of the
molecule for the injection of electrons at low temperatures (<10
K). The rotation is based on energy transfer from the tunneling
electron to the molecule that leads to an electronic excited state,
in analogy to the effect of photoexcitation. We report herein a
new rotary molecular motor (E-1, Figure 1a) that uses electrical
energy to trigger redox reactions and thus bias the Brownian
rotation. Such an operation mechanism differs from that in the
STM-driven systems and allows one to carry on an ensemble of
molecules at relatively high temperatures (∼180 K).
Scheme 1 shows the synthesis of E-1. The Horner−
Wadsworth−Emmons reaction between propiophenone and
triethyl phosphonoacetate led to the cinnamic acid ester 3.
Asymmetrical conjugated reduction of 3 with CuCl and PMHS
(
polymethylhydrosiloxane) in the presence of the (S)-p-tol-
13
BINAP ligand followed by base-catalyzed hydrolysis of the
ester group produced the chiral acid 4 with an S configuration for
the chirality center. Cyclization of 4 promoted by PPA
14
(
polyphosphoric acid) led to the optically active indanone 5
with a 90% ee value. The target system E-1 was obtained through
a base-catalyzed aldol condensation reaction between 5 and the
The chiral molecular motor E-1 was designed on the basis of
the achiral precursor E-2 (Figure 1b), which was investigated as a
molecular brake on the basis of interconvertable E and Z isomers;
the photochemical switching of E-2 → Z-2 results in a decrease of
the rotation rate by 500-fold for the pentiptycene rotor about the
15
previously reported pentiptycene building block 6 under
microwave irradiation. The basic conditions for aldol con-
densation do not racemize the chirality center in 5 but somewhat
decrease the ee value from 90% to 81% in a “blank” 6-free
11
external C−C bond, mimicking brake performance. The
reverse switching of Z-2 → E-2 that recovers the brake-off
state was effectively achieved by electrochemical redox-gating via
Received: October 6, 2014
©
XXXX American Chemical Society
A
dx.doi.org/10.1021/ol502946v | Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Figure 1. (a) Design of a chiral and electrical energy-gated molecular motor E-1 from (b) an achiral molecular brake precursor E-2 that is operated by
light and electrical energy and (c) schematic drawing of the operation concept for the directional rotation of E-1, in which switching from the rotation-
inhibited neutral form to the radical anionic form allows a biased clockwise rotation of the pentiptycene rotor. For the clarity of the rotational motion,
one of the pentiptycene blades is labeled with a different color. The red CC−CC bond in (a) defines the twist angle α for discussion in the text.
Scheme 1. Synthesis of E-1
1
Figure 2. Variable-temperature H NMR (left) and simulated (right)
spectra of E-1 in DMF-d at the region of bridgehead protons. Values of
7
−1
temperature (T in K) and the interconversion rate (k in s ) are also
given for each trace.
−
1
barrier (E = 11.9 vs 8.3 kcal mol ), preexponential value (log A
a
‡
=
11.3 vs 10.6), and enthalpy of activation (ΔH = 11.4 vs 7.8 kcal
−1
‡
mol ) but less negative entropy of activation (ΔS = −8.7 vs
1
−1
−
12.0 cal mol− K ). Consequently, the replacement of a
hydrogen in the indanone methylene moiety with an ethyl group
raises the free energy of activation (ΔG ) by ∼2.5 kcal mol at
298 K. On the basis of the activation parameters, the rotation rate
‡
−1
experiment. Assuming that the same ee value of 81% is retained
in the product E-1, the specific rotation of E-1 in CHCl at 23 °C
3
(
[α]23D) could be estimated to be +10.9°. The corresponding
for E-1 would be lower than 0.02 s at 200 K.
−1
CD spectrum is shown in Figure S1. Chiral resolution of E-1 was
not performed.
The RPES for E-1 was constructed by DFT calculations at the
B3LYP/6-31G(d) level (Figure 3a). To simplify the calculation,
the octyl group was replaced with a methyl group. The fully
optimized local minimum shows a twist angle α of −65.8° that is
defined by the pentiptycene−methinylindanone CC−CC
dihedral angle (Figure 1a). By looking straight down the external
C−C bond with the pentiptycene rotor in the front, clockwise
The rotational kinetics of E-1 have been investigated by
1
variable temperature (VT) H NMR spectroscopy. Because of
the presence of a chirality center in the indanone moiety, the four
bridgehead proton signals of the pentiptycene rotor have
different chemical shifts when the rotation is slower than the
NMR time scale. This is indeed observed for E-1 in DMF-d at
2
rotation of the rotor by ∼96° reaches the transition state TS (α
7
C
43 K (Figure 2). The peak assignment was accomplished with
D NMR spectra, including NOESY, ROESY, COSY, HMBC,
= −161.8°), but the transition state TS (α = 16.6°) for
CC
2
counterclockwise rotation is reached at a rotation of ∼82°,
revealing an asymmetric RPES. Gibbs free energy correction was
applied to the fully optimized local maximum and minimum of
the RPES at the BMK/6-311+G(d,p)//B3LYP/6-31G(d) level
(see Supporting Information). At 298 K the transition states are
∼13.5 kcal mol− higher in free energy than the optimized
and HSQC (Figures S2−S6). Line-shape analysis on the
bridgehead proton signals provided the information on rotation
rates at each temperature, and the activation parameters derived
from Arrhenius and Eyring plots (Figures S7 and S8) are listed in
Table S1. Compared to E-2, E-1 displays a larger activation
1
B
dx.doi.org/10.1021/ol502946v | Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
The differences in RPES for E-1 vs E-1• provide an
opportunity toward directional rotary operation gated by
electrical energy. When the temperature is lowered so that
−
•
−
Brownian rotation is very slow, a switching from E-1 to E-1
•−
could lead to a position near the TSc of E-1 (black vertical
arrow in Figure 3a), which would relax to the ground state of E-
•−
1
either through clockwise (red curved arrow in Figure 3b) or
counterclockwise (blue curved arrow in Figure 3b) rotation. A
•
−
reverse switching of E-1 back to E-1 finishes a clockwise
rotation of 180° for the clockwise relaxed portion (red vertical
arrow in Figure 3b), but it recovers the original status of E-1 for
the counterclockwise relaxed portion (blue vertical arrow in
Figure 3b). Figure 3c depicts such a pulsating ratchet mechanism,
in which the solid and dashed curves represent the RPES of E-1
•
−
and E-1 , respectively, and the red dots represent the position of
the rotors along the RPES. At low temperatures, a Brownian
rocking motion between the transition states TS and TS is
C
CC
expected to dominate the torsions (RPES I). When electrical
•
−
energy is applied and E-1 is formed, some molecules would end
•
−
up near the TS of E-1 (RPES II) and relax clockwise (RPES
C
•−
III). Oxidation of E-1 back to E-1 (RPES IV) followed by
relaxation led to a net clockwise rotation of 180° (RPES V).
Continued performance of double potential steps that repeat the
Figure 3. DFT-derived rotational potential energy surface (RPES) for
the pentiptycene rotor in (a) E-1 and (b) E-1•− about the external C−C
•
−
switching cycle of E-1 → E-1 → E-1 would continuously drive
the 180° clockwise rotation for the pentiptycene rotor as long as
the electrochemical switching is faster than the random
Brownian rotation, leading to directional or biased clockwise
rotation of the system. A schematic drawing of the electrical
energy-biased rotation is depicted in Figure 1c. It should be
bond in both clockwise and counterclockwise directions (see Figure 1c
for definition), and (c) schematic representation of the concept of
directional rotation triggered by electrochemical switching between E-1
•
−
and E-1 .
•
−
noted that at the moment of electron transfer, either E-1 → E-1
ground state, which is in excellent agreement with the
•
−
−1
−1
or E-1 → E-1, the unrelaxed product is 3−6 kcal mol higher
than the relaxed form (Figure S10). Nevertheless, this will not
affect the pulsating ratchet gating of the rotation.
experimental value of 13.9 kcal mol for the activation energy
of rotation. The transition states TS and TS encounter not
C
CC
only the steric repulsion between the protons on the
pentiptycene bridgehead and on the methinyl carbons but also
a distorted C−CC bond angle for the methinyl carbon from
In principle, the electrochemical switching time scale must be
shorter than the time scale for random Brownian rotation to bias
the rotation direction, and as the time scale difference increases,
so does the rotation directionality. At 298 K, each Brownian
128° in the ground state to ∼143° in the transition states (Table
S2).
The RPES for E-1•− (Figure 3b) was also constructed by DFT
−3
−5
•−
rotation takes 3.0 × 10 and 1.4 × 10 s for E-1 and E-1 ,
respectively. This indicates that the potential pulse time for the
double potential step switching must be shorter than 10 μs to bias
the rotation direction at 298 K. However, the Brownian rotation
rate could be largely reduced at lower temperatures. For example,
according to the activation parameters, the Brownian rotation
calculations at the same level of theory as that for E-1. The
ground state possesses a smaller twist angle α of −48.7° as
compared to the neutral form (−65.8°). The TS (α = −116.3°)
C
is reached with a clockwise rotation of the pentiptycene rotor by
∼
67°, and the TSCC (α = 70.7°) is formed with a counter-
•
−
clockwise rotation by ∼119°, again corresponding to an
time scale becomes ∼900 s for E-1 and ∼0.73 s for E-1 at 180 K
‡
−1
6
•−
asymmetric RPES. The calculated ΔG is 10.8 kcal mol for
E-1 at 298 K, which is 2.7 kcal mol lower than that for E-1.
and is ∼10 s for E-1 and ∼170 s for E-1 at 150 K. Therefore,
•
−
−1
the use of potential pulse times of 0.1 s would bias the rotation
direction to ∼70% at 180 K and to >99% at 150 K. Such an
operation temperature is considerably higher than that (<10 K)
•
−
17
The difference in RPES for E-1 vs E-1 could be understood by
the fact that adding an electron to the π* orbital of a conjugated
system reduces the bond order of a double bond and increases
that of a single bond, as indicated by the lengthened exocyclic
CC bond (1.38 vs 1.34 Å) and shortened exocyclic C−C bond
9
,10
for STM-based single molecule systems.
The robustness of electrochemical switching between E-1 and
•−
E-1 is evidenced by the persistent cyclic voltammograms
(Figure 4a) and by the stable chronocoulometric charge−time
curves (Figure 4b). The initial and final potentials were 1.40 and
−1.70 V relative to the ferrocene/ferrocenium redox couple (Fc/
(
1.45 vs 1.48 Å). The increased bond order for the exocyclic C−
C bond accounts for the decrease of α in the ground state. An
increase of structural planarity for π-conjugated systems upon
oxidation or reduction has also been experimentally demon-
+
Fc ), respectively. The pulse time for chronocoulometry was 0.2
1
6
strated. The decreased bond order for the exocyclic CC
bond makes the methinylindanone moiety less rigid and thus
shifts the positions of the transition states and reduces the
rotation barrier. It was also predicted that isomerization from E-
s, but a pulse time as short as 100 μs was also achievable (Figure
•
−
•−
S11). The stability of E-1 relative to the isomer Z-1 was also
investigated by HPLC analysis on the solution after 100 and 200
cycles of switching (Figure 4c). An authentic sample of Z-1 was
prepared by the irradiation of E-1 in DCM at 306 nm (see
Supporting Information). The results showed that the E/Z ratio
is 99:1 and 97:3 after 100 and 200 cycles of switching,
respectively. The latter is similar to that observed for the achiral
•
−
•−
1
to Z-1 is energetically unfavorable, because the latter is less
−1
stable than the former by 0.84 kcal mol (Figure S9). This value
is likely underestimated according to the results of HPLC
analysis on E-1 after electrochemical treatment (vide infra).
C
dx.doi.org/10.1021/ol502946v | Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
(2) (a) Kottas, G. S.; Clarke, L. I.; Horinek, D.; Michl, J. Chem. Rev.
2
005, 105, 1281−1376. (b) Coskun, A.; Banaszak, M.; Astumian, R. D.;
Stoddart, J. F.; Grzybowski, B. A. Chem. Soc. Rev. 2012, 41, 19−30.
(
(
3) Kinbara, K.; Aida, T. Chem. Rev. 2005, 105, 1377−1400.
4) Kelly’s triptycene-helicene system provides an early example of
directional rotation, but it is not a full 360° rotation; see: Kelly, T. R.; De
Silva, H.; Silva, R. A. Nature 1999, 401, 150−152.
(5) (a) Koumura, N.; Zijlstra, R. W. J.; van Delden, R. A.; Harada, N.;
Feringa, B. L. Nature 1999, 401, 152−155. (b) Feringa, B. L. J. Org.
Chem. 2007, 72, 6635−6652.
(6) Fletcher, S. P.; Dumur, F.; Pollard, M. M.; Feringa, B. L. Science
2
(
005, 310, 80−82.
7) Leigh, D. A.; Wong, J. K. Y.; Dehez, F.; Zerbetto, F. Nature 2003,
4
(
(
24, 174−179.
8) Greb, L.; Lehn, J.-M. J. Am. Chem. Soc. 2014, 136, 13114−13117.
9) Perera, U. G. E.; Ample, F.; Kersell, H.; Zhang, Y.; Vives, G.;
Echeverria, J.; Grisolia, M.; Rapenne, G.; Joachim, C.; Hla, S. W. Nat.
Nanotechnol. 2013, 8, 46−51.
(10) Kudernac, T.; Ruangsupapichat, N.; Parschau, M.; Macia, B.;
Katsonis, N.; Harutyunyan, S. R.; Ernst, K.-H.; Feringa, B. L. Nature
011, 479, 208−211.
11) Chen, Y.-C.; Sun, W.-T.; Lu, H.-F.; Chao, I.; Huang, G.-J.; Lin, Y.-
C.; Huang, S.-L.; Huang, H.-H.; Lin, Y.-D.; Yang, J.-S. Chem.Eur. J.
011, 17, 1193−1200.
(12) Astumian, R. D. Science 1997, 276, 917−922.
13) Appella, D. H.; Moritani, Y.; Shintani, R.; Ferreira, E. M.;
Buchwald, S. L. J. Am. Chem. Soc. 1999, 121, 9473−9474.
14) Stephan, E.; Rocher, R.; Aubouet, J.; Pourcelot, G.; Cresson, P.
2
(
Figure 4. (a) Differential pulse voltammogram (red) and overlay plot of
5
cyclic voltammograms (black) of E-1 in DCM (0.55 mM) with 0.1 M
−1
Bu NPF at a scan rate of 100 mV s and (b) chronocoulometric
4
6
2
diagram of E-1 with 100 cycles of alternating switching of applied
+
electric potential at −1.7 and 1.4 V (vs Fc/Fc ) for 0.2 s, and the first 5
(
cycles are shown in the inset figure; (c) HPLC trace of (i) E-1 and Z-1
mixture, (ii) E-1 only, and the solution of E-1 after (iii) 100 and (iv) 200
cycles of chronocoulometric measurements (pulse time is 0.2 s).
(
Tetrahedron: Asymmetry 1994, 5, 41−44.
(
(
15) Yang, J.-S.; Ko, C.-W. J. Org. Chem. 2006, 71, 844−847.
16) Fujitsuka, M.; Tojo, S.; Yang, J.-S.; Majima, T. Chem. Phys. 2013,
4
(
19, 118−123.
17) The rotation directionality (r.d.) was calculated according to the
equation: r.d. = (t − 2t )/t , where t is the time for a 180° Brownian
1
1
precursor E-2 (96:4). All these data secure the electrical
energy-gated operation of molecular motor E-1.
B
ps
B
B
•−
In summary, the distinct rotational potential energy surfaces
for E-1 in the neutral vs the radical anionic forms and the reliable
electrochemical switching between the two forms pave a path
toward a new Brownian motor that is gated by electrical energy.
The pulsating ratchet mechanism renders this system to be
operable at much higher temperatures than the previously
rotation in E-1 and 2t denotes the time for each redox switching cycle
E-1 → E-1 → E-1 that consists of double potential switching (tps).
ps
•−
9
,10
reported electrical energy-gated molecular motors.
ASSOCIATED CONTENT
Supporting Information
■
*
S
Experimental methods, synthetic details, 1D and 2D NMR
spectra, Arrhenius and Eyring plots, activation parameters for the
rotation, and DFT-calculated structural parameters. The material
is available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Authors
■
*
E-mail: jsyang@ntu.edu.tw.
*E-mail: ichao@chem.sinica.edu.tw.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank the Ministry of Science and Technology of Taiwan and
National Taiwan University for financial support. The authors
thank Prof. Ying-Chih Lin and Ms. Shou-Ling Huang (NTU) for
the support of VT NMR measurements.
REFERENCES
■
(
4
1) Kay, E. R.; Leigh, D. A.; Zerbetto, F. Angew. Chem., Int. Ed. 2007,
6, 72−191.
D
dx.doi.org/10.1021/ol502946v | Org. Lett. XXXX, XXX, XXX−XXX