University of Groningen
Autonomy and Chirality in Molecular Motors
Kistemaker, Jozef Cornelis Maria
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2017
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Kistemaker, J. C. M. (2017). Autonomy and Chirality in Molecular Motors. Rijksuniversiteit Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
217 Parts have been published as:
B. S. L. Collins, J. C. M. Kistemaker, E. Otten, B. L. Feringa, Nat. Chem. 2016, 8, 860–866, doi:10.1038/nchem.2543.
Chapter 8: Chemically Driven Rotary Motor with
Autonomous Directionality
Herein is reported: A concise and critical evaluation of advances made towards the chemically driven autonomous unidirectional rotation around a single bond is presented. This is followed by the design of a rotary motor whit the potential to exhibit an increased degree of autonomy. This motor’s properties are evaluated using computational chemistry resulting in a proposed change in design. That design is similarly studied and its calculated properties are presented. The computationally supported design is compared to an adapted experimental realization of the chemically driven rotary motor exhibiting autonomous unidirectionality.
218
8
Introduction
The molecular rotary motors studied in previous chapters are all driven by light, however, as is known from Nature’s motor; ATP synthase, rotation can also be fuelled by other sources.[1] ATP synthase can make use of a proton gradient over a membrane to synthesize ATP from ADP, while it can use the reverse conversion of ATP to ADP as an energy source to drive its unidirectional rotation. Since light has a limit to penetration in materials it would be beneficial to the expansion of the nanoengineer’s toolkit if synthetic rotary motors could run on a chemical fuel as well. For chemically driven rotary motors we uphold the same criteria as for the light driven members; rotation around a bond, in a unidirectional fashion, complete 360° revolution and the rotation should have the ability to be repetitive. The independent operation of a motor would profit from the ability to govern its own rotational direction and a final feature which might be beneficial to any motor is the ability to function continuously when fuelled.
Kelly et al. were the first to claim chemically driven unidirectional rotary motion in a molecular system (Figure 8.1).[2] In this system a [4]helicene bearing an alcohol is connected to a triptycene amine ((P,P)-1a) which is transformed to an isocyanate using phosgene allowing for the formation of the carbamate (P,P)-1b (calculated geometries of 1c depicted in Figure 8.2 as visual aids). This puts the system in a metastable state which purportedly undergoes a helix inversion in which the helicene crosses over the triptycene yielding (P,M)-1b, upon treatment with sodium triethoxyhydroborate the hydroxyl-amine (P,M)-1a is obtained which has then undergone a 120° rotation with respect to its starting geometry (P,P)-1a.
Figure 8.1. First apparent example of a chemically driven molecular motor which rotates 120°
unidirectional around a single bond. i) COCl2, NEt3, CDCl3. ii) 6 h, rt, 80%. iii) NaBH(OEt)3.[2]
Only a single enantiomer is shown although the experiments were conducted on racemic material using NMR.[3] This was justified by the assumption that the helicity of the helicene (first stereo descriptor of 1) is unchanged during the rotational process. This might have been an oversight since calculated barriers on a simplified model suggest the possibility of a second pathway. The barrier for single bond rotation in 1c ((P,P)-1c (P,M)-1c) is calculated (semi-empirical
219
CHEMICALLY DRIVEN ROTARY MOTOR WITH AUTONOMOUS DIRECTIONALITY
8
PM6) to roughly equal the barrier of helicene helix inversion in 1c ((P,P)-1c (M,P)-1c) (Figure 8.2). The two products are enantiomeric and therefore indistinguishable by NMR, warranting the use of an asymmetric analysis technique though such a technique has not been applied to 1. However, this does not exclude the presence of unidirectional motion in a molecular system, the motion would only be of a different nature (instead of rotation over a single bond it would be the helix inversion of a helicene).
Figure 8.2. Geometries of 1c depicted as viewed from the top of the helicene. First stereo
descriptor indicates the helicity of the helicene and the second descriptor the helicity over the helicene-triptycene single bond.
In either case, rotor 1 does not fulfil all of the required criteria set for a chemically driven unidirectional rotary motor, but the use of atropodiastereomerization is similarly applied to a system suggested by Dahl and Branchaud in which a biphenyl would be able to rotate 360° (Figure 8.3).[4] They propose a chiral lactone 2a which would undergo diastereoselective ring-opening to either (P)-2b or (M)-2b followed by another diastereoselective ring-closing after a 180° rotation. The system should work if one sequence possesses different rates than the other (k1-k180°-k2 k-1-k180°-k-2). 2a O OMe OH OMe O Nu OH OMe O Nu Nu (P)-2b (M)-2b k1 k-1 k2 k-2 k180° rotation O
Figure 8.3. Proposed system for unidirectional rotation around a biphenyl bond based on
ring-opening/ring-closing of a chiral lactone. Based on the assumptions that k1k-1 and k2k-2 several
conditions are imaginable under which unidirectional rotation might be achieved. Adapted from [4].
It is not mentioned that a molecule such as 2a is not flat over the biaryl but possesses helical chirality. There will be a preference of one diastereoisomer over the other due to the point chirality of the hydroxyl substituted carbon (i.e. in the
220
8
depicted enantiomer of 2a; (M,S) with the carbonyl in the front will be preferred over (P,S)). An achiral nucleophile will therefore always lead to (M)-2b upon ring-opening (k-1) followed by ring-closing by k2 resulting in zero net rotation. This is completely in accord with the second law of thermodynamics but if the process could be directed asymmetrically to (M)-2b through either k1 or k180° without it being in equilibrium, unidirectionality could be achieved. The first step of such a feat has been achieved by the same authors in an achiral lactone 3a.[5] This lactone can be opened asymmetrically by the use of a chiral nucleophile yielding (P)-3b and (M)-3b with a preference of 3 to 1 for one over the other (unknown which). After this 90° rotation the free phenolate is allowed to cyclize on the opposite side reforming lactone 3c completing a 180° rotation. Although not performed, a selective hydrolysis of the amide would have reproduced 3a and half the cycle of 360° rotation would have been completed.
Figure 8.4. Working system based on a diastereoselective ring-opening, rotation and subsequent
closing resulting in a 180° rotation around a biphenyl bond with a 58% rotational yield and a preference of 3 to 1 for a specific direction. Adapted from [5].
Interestingly, without ever claiming it as such Bringmann and Hartung achieved unidirectional rotation in a very similar system 13 years earlier with a 94% rotational yield and a preference of 98.5 to 1.5 over a 90° angle of rotation (Figure 8.5).[6] With only a minor addition to their experiment it could have easily been extended to a 180° rotation. Theoretically, one might afford 4b using a lactonization of (P)-4a (experimentally 4b was obtained from bromo-naphthoic acid and dimethylphenol) after which the reported stereoselective reduction using (S)-2-methyl-CBS-oxazaborolidine gives (M)-4c. During such a sequence the system would have undergone a 180° rotation.
Figure 8.5. Ninety degrees unidirectional rotation around a single bond with the potential to extend
it to 180°. i) Lactonization of enantiopure starting material. ii) (S)-CBS, BH3·THF, 0–30 °C, 0.7 h.
Adapted from [6].
This method was explored by Fletcher et al. in the design and realization of the first chemically driven motor able to unidirectionally rotate 360° around a single bond
221
CHEMICALLY DRIVEN ROTARY MOTOR WITH AUTONOMOUS DIRECTIONALITY
8
and with the ability to do so in a repetitive fashion.[7] Starting from lactone 5a similar to 4b except with a p-methoxybenzyl (PMB) protected alcohol ortho to the biaryl bond, an asymmetric ring-opening using the same chiral organoboron catalyst followed by an allylation of the free phenol and oxidation of the benzylic alcohol to an acid yields (P)-5b with a unidirectional yield of 94%. Selective deprotection of the PMB protected alcohol directly leads to the lactone 5c finishing the first 180° of the rotary cycle. Another stereoselective reduction using (S)-CBS, protection of the phenol with PMB and oxidation of the benzylic alcohol accomplishes the next 90° in the cycle with a preference of 9 to 1 leading to (P)-5b. The 360° cycle is completed by quantitative removal of the allyl protecting group and lactonization regenerating 5a affording an overall yield of 21% and a unidirectional yield of 75%.
Figure 8.6. Chemically driven molecular motor undergoing full and potentially repetitive 360°
unidirectional rotation around a single aryl-aryl bond. i) 1. (S)-CBS, BH3·THF, 0 °C, 0.5 h. 2.
allylbromide, K2CO3, 20 h. 3. CrO3·H2SO4·H2O, acetone, 2 h. 4. NaClO2, 2-methyl-2-butene, 1 h,
70% yield with a preference of 97 to 3 for the indicated direction. ii) Ce(OTf)3,
1,3-dimethoxybenzene, MeNO2, 60 °C, 0.5 h, 76%. iii) 1. (S)-CBS, BH3·THF, 0 °C, 0.1 h. 2.
p-methoxybenzyl chloride, K2CO3, NaI, , 30 h. 3. MnO2, 48 h. 4. NaClO2, 2-methyl-2-butene, 1 h,
40% yield with a preference of 9 to 1 for the indicated direction. iv) 1. Pd(PPh3)4, HCO2H, , 24 h.
2. DCC, 0.3 h, 99%. Adapted from [7].
While this system has fulfilled the basic requirements for a chemically driven molecular rotary motor, it, however, lacks autonomous directionality. For every rotation chiral reagents have to be fed to the reaction to control the rotational direction. It might be very advantageous to have a system which could undergo a chemically driven 360° rotation around a single bond with autonomous directionality as proposed by Dahl and Branchaud.[4] This could possibly be achieved by combining this system with the orthogonal protection/deprotection steps of the Feringa system. If the carboxylic acid of 5b could be moved to a stereogenic benzylic position such as the alcohol in 2b, then the selective lactonization step could possibly direct the rotation by atropodiastereomerization in which one helicity is favoured over the other. It has been shown by Bringmann et al. that such an atropodiastereomerization works for six membered rings, although the stereogenic centre is lost upon ring-opening of the acetal.[8,9]
Based on the atropodiastereomerization of chiral bridged biaryls we propose a system as depicted in Figure 8.7 that functions as a chemically driven rotary motor
222
8
possessing autonomous directionality with the ability to undergo repetitive 360° rotation. The first step in the proposed system (i in Figure 8.7) is a selective ring-closing of (P)-6a to (P)-6b, this can be achieved by either a deprotection of X1 or a chemoselective preference for X1 in the subsequent ring-closing bond formation. The next step (ii) is the atropodiastereomerization in which an energy difference between the diastereomers (P)-6b and (M)-6b originating from the chiral directing group (DG) rotates the phenyls with respect to each other. A ring-opening of (M)-6b affords (M)-6a in which the system has undergone a 180° rotation. The second 180° to achieve a full rotation is based on similar steps except now the ring-closing is specific for X2.
Figure 8.7. Proposed chemically fuelled molecular motor with autonomous directional 360°
rotation. R=any atom or group which sterically locks the helicity of 6a, DG=chiral directing group, X1 and X2=any although different groups able to undergo reversible cyclization to form a six
membered ring. i / iv) ring formation. ii / v) atropodiastereomerization. iii / vi) ring-opening.
This loss of chirality for acetals as observed by Bringmann is not an issue with chiral sulfoxides used by Colobert and coworkers which are also able to form six membered rings through cyclopalladation.[10–12] The sulfoxide directing group (DG = SOTolyl) can make use of metal catalysed additions / insertions such as oxidative addition of palladium(0) to a halogen (X1 = halogen) and C-H activation of palladium(II) (X2 = H). These reactions have the benefit that additional protection and deprotection steps can be avoided. This proposed chemically driven molecular rotary motor is investigated computationally.
Primary Design of a Chemically Driven Motor
Starting geometries for DFT calculations were created with the semi-empirical PM6 method. Geometry optimizations were performed using the Becke 3-parameter Lee-Yang-Parr (B3LYP)[13][14] functional with a split valence basis set (def2-SVP[15] and dhf-SVP[16] for Pd and I) and effective core potentials (ECP) for Pd[17] and I[18] in dichloromethane (DCM) using the integral equation formalism polarizable continuum model (IEFPCM)[19][20]. Single point energy calculations
223
CHEMICALLY DRIVEN ROTARY MOTOR WITH AUTONOMOUS DIRECTIONALITY
8
were performed on optimized geometries using B3LYP and a triple zeta valence basis set (def2-dhf-TZVP)[15][16] with ECP on Pd and I in dichloromethane (IEFPCM).
Figure 8.8. Scheme for ground state rotation around the biphenyl bond of 7a with the optimized
geometries for minima and transition states (indicated by ‡) displayed above as viewed from the top through the biphenyl C-C bond and the lower phenyl in the z-x plane.
Table 8.1. Gibbs free energies (kJ·mol−1) and dihedral angles of 7a
(B3LYP/def2-dhf-SVP//def2-dhf-TZVP with ECP(I) in DCM (IEFPCM))
Compound Gibbs free energy (kJ·mol−1) I-C-C-S dihedral angle
(P,S)-7a 2.1 110°
(Z,S)-7a‡ 173 −14.4°
(M,S)-7a 0 −76.5°
(E,S)-7a‡ 141 185°
Biphenyl 7a with a single configuration on the tolylsulfoxide (S) exists as two twisted conformations with P and M helicity ((P,S)-7a and (M,S)-7a) which can interconvert through a rotation around the biphenyl carbon-carbon bond (Figure 8.8). During these rotations the iodine on the upper phenyl group has to pass either the fluorine ((E,S)-7a‡) or the tolylsulfinyl group ((Z,S)-7a‡). These transition states
approach a coplanar structure, however, the large size of iodine and sulfoxide force the structure to deviate from coplanarity to an anti-folded geometry ((M)-(Z,S)-7a‡
and (P)-(E,S)-7a‡)). The diastereomeric anti-folded transition state geometries
((P)-(Z,S)-7a‡ and (M)-(E,S)-7a‡), not depicted) were found to be higher in energy. To
increase the reliability of the energy levels the energies of minima and transition states were calculated at a higher level (Table 8.1) which revealed high barriers for racemization (t½rac = 14 200 years) indicating a very high helical stability for the two isomers. These atropodiastereoisomers differ slightly in energy (2.1 kJ·mol−1) and notably in geometry as indicated by the dihedral angle of the iodine and sulfur over the biphenyl carbon-carbon bond in which (P,S)-7a exceeds the perpendicular
224
8
plane while the iodine in (M,S)-7a stays within 90 degrees of the sulfoxide (Table 8.1 and Figure 8.8).
Figure 8.9. Scheme for cis-trans isomerization and atropodiastereomerization by helix inversion of
the complex of 7b with the optimized geometries for minima and transition states (indicated by ‡) displayed to the sides of structures as viewed from the top through the biphenyl C-C bond and the lower phenyl in the z-x plane.
Table 8.2. Gibbs free energies (kJ·mol−1) and dihedral angles of 7b
(B3LYP/def2-dhf-SVP//def2-dhf-TZVP with ECP(Pd,I) in DCM (IEFPCM))
Compound Gibbs free energy (kJ·mol−1) Pd-C-C-S dihedral angle
trans-(P,R)-7b 21.8 38.8° trans-(Z,R)-7b‡ 83.3 1.51° trans-(M,R)-7b 0 −36.8° cis-(P,R)-7b 27.9 37.2° cis-(Z,R)-7b‡ 81.1 3.39° cis-(M,R)-7b 9.5 −35.8°
Reacting (S)-7a with palladium(0) leads to an oxidative addition into the carbon-iodine bond forming (R)-7b directed by the asymmetric sulfur. The relative configuration of the stereogenic sulfur centre does not change upon coordination to palladium even though the absolute stereochemical descriptor changes. Practically an array of ligands can be used on palladium(0) such as phosphines, phosphites, arsines and isonitriles but for theoretical purposes we chose pyridine to reduce calculation costs (e.g. 34 atoms in PPh3 vs 11 in pyridine). In solution, free pyridine can exchange to form the thermodynamically most stable isomer (cis- vs. trans-7b) and the complex can undergo reversible helix inversion of the biphenyl directed by the stereogenic sulfoxide (Figure 8.9). As shown by the dihedral angle (Table 8.2)
225
CHEMICALLY DRIVEN ROTARY MOTOR WITH AUTONOMOUS DIRECTIONALITY
8
the geometrical differences between trans-7b and cis-7b are small, though it does significantly influence their respective energy levels, where the trans-7b minima are stabilized with respect to cis-7b except for the transition state. If the equilibrium between trans-(P,R)-7b and cis-(P,R)-7b is fast, the helix inversion proceeds mainly via cis-(Z,R)-7b‡ and would exhibit a half-life of 4.2 ms, whereas a slow
trans-cis isomerization would primarily direct the pathway for atropodiastereomerization through trans-(Z,R)-7b‡ (t½ = 10 ms). Both pathways
would lead quantitatively to trans-(M,R)-7b and cis-(M,R)-7b in a 98 to 2 ratio. Similar to the oxidative addition, a C-H activation of (S)-7a using palladium(II) forms (R)-7c directed by the sulfoxide with retention of the relative configuration of the sulfur (Figure 8.10). Acetate might seem to be the first ligand of choice on palladium(II) however those complexes prefer to dimerize, trimerize and oligomerize. The palladium(II) acetate adducts of 1a are too bulky to form dimers and trimers and attempts to calculate oligomers would be extremely costly. Therefore acetylacetonate (acac) was used which binds to the Pd(II) in a bidentate fashion which allows for the complex to be investigated as a monomer ((R)-7c).
Figure 8.10. Scheme for atropodiastereomerization by helix inversion of the complex of 7c with
the optimized geometries for minima and transition states (indicated by ‡) displayed above as viewed from the top through the biphenyl C-C bond and the lower phenyl in the z-x plane.
Table 8.3. Gibbs free energies (kJ·mol−1) and dihedral angles of 7c
B3LYP/def2-dhf-SVP//def2-dhf-TZVP with ECP(Pd,I) in DCM (IEFPCM))
Compound Gibbs free energy (kJ·mol−1) Pd-C-C-S dihedral angle
(M,R)-7c 13.1 38.5°
(M,R)-7c‡ 104 −29.1°
(P,R)-7c‡ 116 29.5°
(P,R)-7c 0 −36.7°
C-H activation of Pd(acac)2 to (M,S)-7a and (P,S)-7a would lead to two minima found to be the twisted atropodiastereomers (M,R)-7c and (P,R)-7c, respectively.
226
8
The minima are connected by two transition states ((M,R)-7c‡ and (P,R)-7c‡) which
either have first undergone a ringflip of the palladacycle or a crossing of the halogens on the opposite side. These structures are sterically forced from coplanarity adopting an anti-folded conformation. Although the effect is not as large as for 7b, the atropodiastereochemistry energetically favours one isomer over the other (Table 8.3). Pure (M,R)-7c would exhibit unimolecular decay via (M,R)-7c‡ to (P,S)-7a with a half-life of 0.5 h at room temperature to form an
equilibrium with a 0.5 to 99.5 ratio, respectively.
These computed atropodiastereomerizations are summarized in an energy diagram (Figure 8.11). The relative energy levels of the three potential energy surfaces are chosen arbitrarily and the vertical transitions between different energy surfaces are shown as adiabatic transitions where in practice those would involve chemical reactions for which the activation parameters are not calculated here.
Figure 8.11. Energy diagram of 7. i) oxidative addition, ii) atropodiastereomerization, iii) reductive
elimination, iv) C-H activation, v) atropodiastereomerization, vi) hydrogenation. See Figure 8.8, 8.9, and 8.10 for structures.
The energy diagram (Figure 8.11) reveals a pathway for a unidirectional rotary cycle (Figure 8.12) where the atropodiastereomerizations barriers are lowered by the formation of a six membered palladacycle bridging the biphenyl and directed by the stereogenic sulfoxide. The (Z)-7a‡ barrier is lowered by the formation of
trans-(P)-7b from (P)-7a by the oxidative addition of a Pd(0) reagent (i) which allows for a facile helix inversion (ii). Iodine can be used to facilitate oxidation of (M)-7b to a Pd(IV) species followed by rapid reductive elimination which
(P)-7a (Z)-7a‡ (M)-7a (E)-7a‡ (P)-7a i (P)-7b ii (M)-7b iii iv (M)-7c v (P)-7c vi 0 25 50 75 100 125 150 175 200 225 7a trans-7b cis-7b (P)-7c (M)-7c
227
CHEMICALLY DRIVEN ROTARY MOTOR WITH AUTONOMOUS DIRECTIONALITY
8
reintroduces the iodine (iii)[21] affording (M)-7a in which the phenyls have undergone a 180° rotation with respect to each other. C-H activation of (M)-7a using a Pd(II) reagent allows for the formation of (M)-7c which subsequently is able to undergo atropodiastereomerizations (v) with a significantly lower barrier than (E)-7a‡ to afford (P)-7c. The directional behaviour of (v) is analogous to that
of (ii) though with a fairly higher barrier of activation probably due to the steric and electronic repulsion of the halides. From (P)-7c, (P)-7a can be obtained by hydrogenation using a hydrogen donor such as hydrazine, isopropanol or formic acid. This completes the chemically driven 360° rotational cycle in which the stereogenic sulfur provides the system with autonomous directionality with a clockwise rotation of the phenyl moiety.
Figure 8.12. Rotary cycle of 7. i) oxidative addition, ii) atropodiastereomerization, iii) reductive
elimination, iv) C-H activation, v) atropodiastereomerization, vi) hydrogenation
The rotation of this system was attempted practically using conditions as indicated in Figure 8.12. Where the first 180° posed no significant practical problems, the second 180° based on C-H activation could not be achieved for this system. In the palladacycle formation using Pd(OAc)2 on a mixture of (P)-7a and (M)-7a little reactivity was observed under a variety of conditions. Increasing the reaction temperature ultimately led to other reactions possibly due to the formation of metallic palladium(0) or merely via degradation pathways. In the reaction of
7a7c the (P)-diastereoisomer has to rotate 11.5 degrees over the I-C-C-F dihedral
angle and (M)-7a has to undergo a 40.9° rotation. Putting the halides in close proximity might be the reason for a high barrier of activation for the C-H addition. To lower this barrier while avoiding major changes to the steps in the rotational cycle, either the fluorine could be exchanged for a substituent with a similar steric hindrance though lowered electronic repulsion, or a change of iodide for a smaller halide.
228
8
Figure 8.13. Scheme for ground state rotation around the biphenyl bond of 8 with the optimized
geometries for minima and transition states (indicated by ‡) displayed to the sides of structures as viewed from the top through the biphenyl C-C bond and the lower phenyl in the z-x plane.
229
CHEMICALLY DRIVEN ROTARY MOTOR WITH AUTONOMOUS DIRECTIONALITY
8
Change in Design and Computational method
To improve the system a change of halide was chosen in which the iodine atom was exchanged for a bromine. At the same time a change was made to optimize the computational method. Long-range corrected functionals and dispersion-corrected DFT significantly improve the results when intermolecular interactions are concerned, but also show improvements in the calculated energies and thermochemistry of single molecules.[22–28] For a review on these new functionals and dispersion correction see the work of Grimme.[29] Grimme’s D3 London Dispersion correction provides the most accurate results though has not been implemented in the software package used. B3LYP is outperformed by most other functionals amongst which double hybrids (B2PLYP-D3) perform markedly well all-round. Consistent and satisfactory results are also obtained using the M06-2X and ωB97X-D functionals of which experience has shown the ωB97X-D functional to perform outstandingly.[30]
Starting from the optimized geometries of 7, the iodines were exchanged for bromines and optimizations were performed with the selected ωB97X-D functional with the same split valence basis set (def2-SVP[15] and dhf-SVP[16] with ECP[17] for Pd) in DCM using IEFPCM[19][20]. Single point energy calculations were performed using ωB97X-D/def2-dhf-TZVPwith ECP on Pd in DCM (IEFPCM).[15][16] The resulting geometries of the conformations of 8 (Figure 8.13) marginally differ from those found for 7. The transition states were not readily obtained but required additional scanning over the dihedral angle dictated by the negative vibrational modes of the transition states of 7. The three pathways for the atropodiastereomerization of 8 are depicted vertically in Figure 8.13 connected by theoretical palladium additions and eliminations. The energies calculated with the triple zeta valence basis set (Table 8.4) reveal changes in the activation barriers and equilibrium constants with respect to 7. Since two changes have been made to the system (the functional and the molecule) it is inappropriate to attempt to assign the changes in energy levels to either the change in functional or molecule.
For the basis rotor 8a the (P,S)-8a diastereoisomer was found to be lower in energy than (M,S)-8a with an equilibrium ratio at room temperature of 81 to 19, respectively, though an enormous barrier for racemization at room temperature was found for the anti-folded transition state (E,S)-8a‡ (t½rac = 2.87 million years). To
obtain a reasonable rate of isomerization, i.e. a half-life of one hour, the system would require a temperature of approximately 234 °C. To overcome these tremendous barriers a six membered palladacycle bridging the biphenyl obtained by palladium insertion could be formed.
230
8
Table 8.4. Gibbs free energies (kJ·mol−1) and dihedral angles of 8
(B97X-D/def2-dhf-SVP//def2-dhf-TZVP with ECP(Pd) in DCM (IEFPCM))
Compound Gibbs free energy (kJ·mol−1) S-C-C-(X) dihedral angle
(P,S)-8a 0 109° (Br) (Z,S)-8a‡ 177 −14.0° (Br) (M,S)-8a 3.5 −74.1° (Br) (E,S)-8a‡ 155 186° (Br) trans-(P,R)-8b 17.6 37.5° (Pd) trans-(Z,R)-8b‡ 82.9 2.87° (Pd) trans-(M,R)-8b 0 −38.0° (Pd) cis-(P,R)-8b 23.1 37.9° (Pd) cis-(Z,R)-8b‡ 83.5 −0.28° (Pd) cis-(M,R)-8b 4.9 −37.1° (Pd) (P,R)-8c 15.0 38.3° (Pd) (M,R)-8c‡ 114 −29.5° (Pd) (P,R)-8c‡ 126 28.9° (Pd) (M,R)-8c 0 −38.8° (Pd)
Oxidative addition by a palladium(0) species into the C-Br bond of (P,S)-8a leads to the formation of (P,R)-8b with the trans configuration on palladium being in favour over the cis by 9 to 1. This reduces the barrier for atropodiastereomerization by means of thermal helix inversion of the biphenyl directed by the stereogenic sulfur to only 65.3 kJ·mol−1 which predicts a half-life of 49 ms at rt readily producing (M,R)-8b (trans to cis ratio remains 9 to 1). Reintroduction of the bromine atom (vide infra) gives 8a as (M,S)-8a in which the phenyls have undergone a 180° rotation with respect to each other. C-H activation of (M,S)-8a with a palladium(II) species gives rise to (P,R)-8c in which neither the relative helical nor point-chirality have changed but due to palladium taking precedence on sulfur and over bromine both absolute chirality descriptors change. This palladacycle assists atropodiastereomerization through the anti-folded transition state (M,R)-8c‡ to give (M,R)-8c. The half-life of helix inversion of (P,R)-8c at rt
is lowered from millions of years to 14 hours, which is a remarkable increase in rate, even though it is a smaller improvement than trans-(Z,R)-8b‡ was for
(Z,S)-8a‡. A hydrogenation of the palladium-carbon bond regenerates (P,S)-8a
completing the 360° rotational cycle for which the energy diagram is summarized in Figure 8.14.
231
CHEMICALLY DRIVEN ROTARY MOTOR WITH AUTONOMOUS DIRECTIONALITY
8
Figure 8.14. Energy diagram of 8. i) oxidative addition, ii) atropodiastereomerization, iii) reductive
elimination, iv) C-H activation, v) atropodiastereomerization, vi) hydrogenation
The proposed rotary system has experimentally been realized by Collins et al. as shown in Figure 8.15.[31] The chemical molecular motor behaves fully as predicted, however, a few practical modifications have been applied to allow for the experimental realization. The most notable modifications regard the ligands on palladium for which the highest yielding option was chosen over the least computationally costly ligand, naturally. (S,M)-8a was converted into (S,P)-8a by C-H activation using palladium(II) acetate followed by hydride transfer from sodium triacetoxyborohydride and carbon–hydrogen reductive elimination completing the first 180 degrees of the rotational cycle. Addition of tricyclohexylphosphine followed by bromination using N-bromosuccinimide (NBS) allowed for the continued rotation through a second 180 degrees of
(S,P)-8a to (S,M)-(S,P)-8a. It should be noted that no additional palladium is added; the initially
in situ-formed palladium(0) is used as a palladium source in the second 180 degrees rotation which adds an exciting feature to the rotational cycle of this chemical molecular motor. Not only does this system rotate unidirectionally around it biphenyl bond, it does so by ‘cycling’ through the oxidation states of catalytic palladium with hydrogenation and bromination agents as its chemical fuels.
(P)-8a (Z)-8a (M)-8a (E)-8a (P)-8a i (P)-8b ii (M)-8b iii iv (P)-8c v (M)-8c vi 0 25 50 75 100 125 150 175 200 225
232
8
Figure 8.15. Chemical structures and reaction scheme for an integrated cycle based on switching
palladium(II) and palladium(0) redox states for unidirectional 360 degree rotation of (S,M)-8 into (S,P)-8 into (S,M)-8. i) C-H activation, ii) atropodiastereomerization, iiii) hydrogenation iv) oxidative addition, v) atropodiastereomerization, vi) reductive elimination. Adapted from ref. [31].
This system exhibits chemically fuelled rotation around a biphenyl bond in a single molecule with autonomous directionality and it takes us a major step further towards the ultimate goal of a fully autonomous chemically-fuelled molecular motor. To achieve such autonomy, the system needs to be able to perform in the presence of its fuel without an operator’s interference. Theoretically this can already be achieved in the current system, by the omission of the C-H activation. The barrier for atropodiastereomerization of (M)-8a is biased by 22 kJ·mol−1 in the desired direction (Table 8.4, Figure 8.14), additionally, it is the minor diastereoisomer in its equilibrium with (P)-8a. Therefore, at the appropriate temperature (M)-8a is expected to rotate 180 degrees unidirectionally towards
(P)-8a without the need for a C-H activation pathway after which the second 180
degrees rotation can be achieved by the oxidative addition-bromination pathway. The barrier for the unassisted atropodiastereomerization of 8a is nonetheless high and at the required temperatures might allow for undesirable degradation pathways. These can be suppressed by lower barriers which require structural modifications; however, these might in turn significantly alter the behaviour of the molecular machine.
Conclusion
Systems with the potential to behave as chemically driven molecular motors with autonomous directionality have been studied computationally. Changes were made to the design of a theoretically functioning system following experimental results
233
CHEMICALLY DRIVEN ROTARY MOTOR WITH AUTONOMOUS DIRECTIONALITY
8
which revealed problems in the execution and therefore proof of the rotational cycle. Additionally, the computational method was altered to account for dispersion and long range interactions. The theoretical study of the modified design revealed an almost identical behaviour with respect to the initial design. The in silico functioning chemically driven molecular motor with autonomous directionality was shown to perform similarly as such in vitro by the use of powerful organometallic reactivity principles. This achievement represents an important advance in the development of complex molecular machines and highlights the use of quantum chemistry in the design of such systems and in the analysis of their behaviour.
Acknowledgements
This work was conceived and experimentally realized by Beatrice Collins,[31] who is also thanked for her review of this chapter.
References
[1] P. D. Boyer, Nature 1999, 402, 247–249, doi:10.1038/46193. [2] T. R. Kelly, H. De Silva, R. a Silva, Nature 1999, 401, 150–152,
doi:10.1038/43639.
[3] T. R. Kelly, R. A. Silva, H. De Silva, S. Jasmin, Y. Zhao, J. Am. Chem. Soc. 2000,
122, 6935–6949, doi:10.1021/ja001048f.
[4] B. J. Dahl, B. P. Branchaud, Tetrahedron Lett. 2004, 45, 9599–9602, doi:10.1016/j.tetlet.2004.10.147.
[5] Y. Lin, B. J. Dahl, B. P. Branchaud, Tetrahedron Lett. 2005, 46, 8359–8362, doi:10.1016/j.tetlet.2005.09.151.
[6] G. Bringmann, T. Hartung, Angew. Chem., Int. Ed. 1992, 31, 761–762, doi:10.1002/anie.199207611.
[7] S. P. Fletcher, F. Dumur, M. M. Pollard, B. L. Feringa, Science 2005, 310, 80–82, doi:10.1126/science.1117090.
[8] G. Bringmann, M. Breuning, H. Endress, D. Vitt, K. Peters, E.-M. Peters,
Tetrahedron 1998, 54, 10677–10690, doi:10.1016/S0040-4020(98)00618-8.
[9] G. Bringmann, D. Vitt, J. Kraus, M. Breuning, Tetrahedron 1998, 54, 10691– 10698, doi:10.1016/S0040-4020(98)00619-X.
[10] F. R. Leroux, A. Berthelot, L. Bonnafoux, A. Panossian, F. Colobert, Chem. - Eur.
J. 2012, 18, 14232–14236, doi:10.1002/chem.201202739.
[11] T. Wesch, F. R. Leroux, F. Colobert, Adv. Synth. Catal. 2013, 355, 2139–2144, doi:10.1002/adsc.201300446.
[12] C. K. Hazra, Q. Dherbassy, J. Wencel-Delord, F. Colobert, Angew. Chem., Int. Ed. 2014, 53, 13871–13875, doi:10.1002/anie.201407865.
[13] A. D. Becke, Phys. Rev. A 1988, 38, 3098–3100, doi:10.1103/PhysRevA.38.3098. [14] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789,
234
8
[15] F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys. 2005, 7, 3297–3305, doi:10.1039/b508541a.
[16] F. Weigend, A. Baldes, J. Chem. Phys. 2010, 133, 174102, doi:10.1063/1.3495681. [17] K. A. Peterson, D. Figgen, M. Dolg, H. Stoll, J. Chem. Phys. 2007, 126, 124101,
doi:10.1063/1.2647019.
[18] K. A. Peterson, B. C. Shepler, D. Figgen, H. Stoll, J. Phys. Chem. A 2006, 110, 13877–13883, doi:10.1021/jp065887l.
[19] S. Miertuš, E. Scrocco, J. Tomasi, Chem. Phys. 1981, 55, 117–129, doi:10.1016/0301-0104(81)85090-2.
[20] E. Cancès, B. Mennucci, J. Tomasi, J. Chem. Phys. 1997, 107, 3032, doi:10.1063/1.474659.
[21] K. Muñiz, Angew. Chem., Int. Ed. 2009, 48, 9412–9423, doi:10.1002/anie.200903671.
[22] J. Da Chai, M. Head-Gordon, J. Chem. Phys. 2008, 128, 1–15, doi:10.1063/1.2834918.
[23] Y. K. Kang, B. J. Byun, J. Comput. Chem. 2010, 31, 2915–2923, doi:10.1002/jcc.21587.
[24] S. Grimme, S. Ehrlich, L. Goerigk, J. Comput. Chem. 2011, 32, 1456–1465, doi:10.1002/jcc.21759.
[25] L. Goerigk, S. Grimme, Phys. Chem. Chem. Phys. 2011, 13, 6670–6688, doi:10.1039/c0cp02984j.
[26] K. S. Thanthiriwatte, E. G. Hohenstein, L. A. Burns, C. D. Sherrill, J. Chem.
Theory Comput. 2011, 7, 88–96, doi:10.1021/ct100469b.
[27] E. Masson, Org. Biomol. Chem. 2013, 11, 2859–2871, doi:10.1039/C3OB26704K. [28] Y. Liu, J. Zhao, F. Li, Z. Chen, J. Comput. Chem. 2013, 34, 121–131,
doi:10.1002/jcc.23112.
[29] S. Grimme, Wiley Interdiscip. Rev. Comput. Mol. Sci. 2011, 1, 211–228, doi:10.1002/wcms.30.
[30] J.-D. Chai, M. Head-Gordon, Phys. Chem. Chem. Phys. 2008, 10, 6615–6620, doi:10.1039/b810189b.
[31] B. S. L. Collins, J. C. M. Kistemaker, E. Otten, B. L. Feringa, Nat. Chem. 2016, 8, 860–866, doi:10.1038/nchem.2543.
235
