Mechanical Manipulation of Chemical Reactions: Reactivity Switching

Feb 10, 2014 - Photoswitches incorporated into molecular frameworks have been used since a long time to trigger chemical processes on demand. Here, it...
0 downloads 11 Views 735KB Size
Letter pubs.acs.org/JPCL

Mechanical Manipulation of Chemical Reactions: Reactivity Switching of Bergman Cyclizations Martin Krupička,*,†,§ Wolfram Sander,‡ and Dominik Marx*,† †

Lehrstuhl für Theoretische Chemie, Ruhr−Universität Bochum, 44780 Bochum, Germany Lehrstuhl für Organische Chemie II, Ruhr−Universität Bochum, 44780 Bochum, Germany



S Supporting Information *

ABSTRACT: Photoswitches incorporated into molecular frameworks have been used since a long time to trigger chemical processes on demand. Here, it is shown how mechanophores can be used as switches in order to drastically change the reactivity of a neighboring functional group as a function of external stress. The reactivities of cyclic enediynes, which are highly toxic agents when undergoing Bergman cyclization, roughly correlate with the distance between the bond-forming carbons in many cases. It is demonstrated how this distance, and thus enediyne reactivity, can be tuned upon applying mechanical stress. Depending on suitable substitution patterns, chemically inert species can be turned into highly reactive ones and vice versa, thus extending the concept of photoswitching to mechanoswitching. Moreover, depending on the derivative, it is found that C1−C5 cyclization becomes energetically preferred over the Bergman (C1−C6) pathway at nano-Newton forces, thus leading to a force-induced switch in selectivity in such cases. SECTION: Molecular Structure, Quantum Chemistry, and General Theory

E

to cyclization;3,4 the rich interplay of structure, electronic structure, and reactivity is comprehensively reviewed in ref 5. The thermal reaction has been thoroughly studied using various wave function13−15 as well as DFT3,4,16−19 approaches. Although the activation enthalpy and d do not feature a simple linear correlation, a critical range for spontaneous cyclization from about d = 2.9 up to 3.4 Å has been found.3,4 The traditional way to tune the reactivity of cyclic enediynes consists in chemical modification of the reactant, where a rich behavior is found depending on ring size, ring type, and functionalization,2,20 or even upon introducing conformational control as a trigger for activation.21,22 The possibility to apply in a controlled fashion external mechanical forces site-specifically to covalent bonds has already been exploited in seminal experiments using different techniques such as force-clamp atomic force microscopy (AFM), sonication of polymer−functionalized mechanophores, and the molecular force−probe concept.6−8,23−31 Using such techniques, changes of intramolecular structure on the angstrom scale can be achieved if mechanical forces in the appropriate nano-Newton (nN) range are applied, thus enabling covalent mechanochemistry (CMC).9−11 This suggests the basic idea to control the reactions of those molecules where the reactivity strongly depends on structural changes in the sub-Å range.12 In the following, we describe in the sense of generic showcases several cyclic enediynes that have been suitably functionalized for CMC manipulations. Using elec-

nediyne toxins are an important class of anticancer drugs acting as DNA-cleaving agents upon undergoing the Bergman cyclization.1,2 Their reactivity depends in a most crucial way on molecular structure, the distance between the reactive carbons being a simple but crude descriptor;3,4 see ref 5 for a thorough discussion of strain effects and the associated references. As of recently, the controlled distortion of molecules upon applying external mechanical forces became experimentally accessible,6−8 thus opening the field of covalent mechanochemistry.9−11 One such powerful approach is based on functionalizing the force-sensitive functional unit by attaching long polymer chains, which transduce mechanical forces generated by sonication.7,10 Here, considering cyclic enediynes as such “mechanophores”,12 it will be shown that their reactivity toward the Bergman reaction can be tailored at will. Upon suitable functionalization, enediynes with high barriers, thus being stable at room temperature, can be made reactive by sonication, whereas highly reactive species can be stabilized. This opens the doorway to turn benign enediynes into highly reactive species due to mechanoswitching triggered by sonication akin to photoswitching triggered by light. The thermally induced Bergman cyclization is the key step in the action of anticancer antibiotics of the enediyne class. These molecules can bind to DNA, where attack of a nucleophile induces a sequence of reactions that ultimately result in the generation of highly reactive 1,4-didehydrobenzene diradicals, which are the active species cleaving DNA via hydrogen atom abstraction.2 For cyclic enediynes, the size of the ring containing the reactive carbons and their distance d have been shown to be roughly correlated with phenomenological stabilities (see, e.g., Table 4 in ref 2) and activation enthalpies © 2014 American Chemical Society

Received: December 5, 2013 Accepted: February 10, 2014 Published: February 10, 2014 905

dx.doi.org/10.1021/jz402644e | J. Phys. Chem. Lett. 2014, 5, 905−909

The Journal of Physical Chemistry Letters

Letter

tronic structure calculations as a function of constant external force, F0, it will be demonstrated that, indeed, experimentally accessible mechanical forces of about 1−2 nN are able to significantly alter the reaction barrier to Bergman cyclization, ΔE‡(F0). These proof of concept calculations suggest to explore potential applications of suitably tailored enediynes in future experimental studies. The effect of external tensile forces on reactant structures can be roughly estimated using computationally economical force field methods because the major part of the generated strain energy is absorbed by changes of the soft dihedral and bond angles. The strict electronic requirements of transition states (TS) and the changes of electronic structure upon approaching the TS, on the other hand, do not allow for such a simplified molecular mechanics treatment, but require instead structure optimization using proper electronic structure methods. Here, optimization is done using our constant-force (i.e., isotensional or so-called EFEI) approach,32 which is applied to the three functionalized enediyne species depicted in Figure 1 that have

Figure 2. Mechanochemical effects on energetics and structures as a function of constant force F0 upon streching the enediynes 1 (circles, red lines), 2 (squares, green lines), and 3 (diamonds, blue lines) sketched in 1. (a) Change in activation energy, ΔΔE‡(F0) = ΔE‡(F0) − ΔE‡(0); (b) Change in C−C distance, ΔΔd(F0) = Δd(F0) − Δd(0), where Δd(F0) = dRS(F0) − dTS(F0) obtained from d in the reactant state (RS) and transition state (TS); (c) C−C distances, d, in RS (solid lines, left axis scale) and TS (dashed lines, right axis scale); (d) Change in the electronic structure contribution to the activation energy, ΔΔE‡ES(F0), defined in analogy to ΔΔE‡(F0) in panel a.

Supporting Information reveals that the activation entropy is essentially force-independent and small for all three species. The pulling mode (i), as expected, elongates the reactant ground state structure, thus decreasing the carbon−carbon distance d and bringing the reactive centers closer together considerably (solid red line in Figure 2c). Most interestingly, d remains essentially constant in the force-transformed transition state of 1 (dashed red line). This can be understood since quite strict requirements for cycloaromatization reactions have to be fulfilled in the transition state. As a net result, the compression of d when moving from the reactant to the transition state decreases significantly upon stretching (red line in Figure 2b). For the overall reaction of species 1 this means that it is the reactant structure that gets “mechanically distorted” toward the transition state structure, the latter not changing much. Pictorially speaking, the reactant essentially climbs along the reaction coordinate toward the transition state with the activation energy thus continuously decreasing with increasing force as quantified in Figure 2a (red line). In stark contrast to facilitating cyclization as a result of applying an external force, species 2 and thus the second pulling mode (ii) leads to a significant increase of the activation energy according to Figure 2a (green line) with a concurrently increasing distance d between the reactive carbons (solid green line in Figure 2c). The structure of the transition state, however, remains invariant also in this case (dashed green line in Figure 2c). Thus, the observed increase in activation energy can be explained by the direct mechanical action: the difference in distance between reactant and transition state increases (green line in Figure 2b) upon stretching, and therefore the activation energy rises.

Figure 1. Functionalized cyclic enediynes: reactants 1, 2, and 3, and core structure 4 of their reaction products (see text). Red triangles mark the carbons sites to which the force of fixed magnitude and opposite direction is applied collinearly and the distance d between the reactive carbons is shown by dashed lines with arrows.

been selected for computational convenience; the constantforce approach is explained and contrasted to others in a recent review.11 In particular, collinear forces of increasing magnitude and opposite direction are applied to the two carbon sites (symbolized by the red triangles) to which the forcetransducing polymer chains can be attached; the issue of force-transduction along such chains has been examined earlier.33,34 For each constant force, full optimization of reactant, transition state, and product structures is performed, which provides us, beyond any further modeling, with their forceinduced structural distortions and the associated (activation) energies; obviously the “thermal case” and thus the usual activation energy is obtained in the limit of zero force. Three representative modes for applying constant tensile stress were designed as shown in Figure 1: (i) pulling along the molecule to bring the reactive atoms closer together using species 1, (ii) pulling across the molecule based on 2 to make their approach more difficult, and (iii) pulling diagonally via molecule 3 which should not much affect the reaction. The obtained changes in activation energies and the changes in C1−C6 distance, d, from reactant to transition state as a function of increasing force are compiled in Figure 2; comparison of the absolute activation energies to the activation free energies in Figure 3 of the 906

dx.doi.org/10.1021/jz402644e | J. Phys. Chem. Lett. 2014, 5, 905−909

The Journal of Physical Chemistry Letters

Letter

Diagonal pulling (iii) provides us with yet another scenario: this time, stretching does not substantially alter the structure of the reactant 3 (solid blue line in Figure 2c) but instead now has a surprisingly profound distortive effect on the transition state structure (dashed blue line in Figure 2c). It causes a decrease of the decisive C−C distance d with increasing force, as the TS is substantially deformed by the unsymmetrically applied external force. It is noted in passing that the formed C−C bond in the bicyclic product 4 does not change much as a function of force for all three pulling scenarios: it remains virtually constant for species 1 (changing by ca. −0.0006 Å from 0 to 2 nN) and increases only by at most 0.02 Å for the derivatives 2 and 3. The distortions of the molecular skeletons along the reaction pathways are depicted in Figure 3 for 1 and 2 and in the inset of Figure 2 for 3, all obtained by optimizing the structures at a constant force of 2 nN along the pathways. Figure 4. Activation energies for Bergman cyclization (see the solid line in the inset that connects the reactive carbon sites, C1 and C6, for species 3) and for the C1−C5 cyclization pathway (see the dotted line therein) as a function of force for species 1 (red circles, top panel), 2 (green squares, middle), and 3 (blue diamonds, bottom) using solid and dashed lines, respectively.

leading to a force-induced switch in selectivity. Although the activation energy for this alternative process is almost insensitive to force, it becomes slightly preferred for external forces exceeding roughly 1.5 nN due to the aforementioned systematic destabilization of the Bergman pathway as shown in Figure 4. This competing behavior is not found for the other two derivatives, 1 and 2, where Bergman cyclization is stable up to the highest force considered. The presented findings demonstrate that the reactivity of cyclic enediynes with respect to highly reactive 1,4-didehydrobenzenes as intermediates, can be manipulated significantly upon stretching appropriately functionalized species. Most interestingly, mechanical stretching forces could be used to activate enediynes which are stable at room temperature, thus “switching on” biological activity. However, could such mechanoswitching be realized in experiment? Polymer swelling techniques generate only very low forces10 and are thus not successful to influence the reactivity of enediynes.12 On the other hand, it is well-established that sonication10 is able to generate forces in the nN range as required for CMC11 and thus should be successful in activating the mechanoswitch. Another promising approach is to activate the mechanoswitch using a photoswitch such as stiff stilbene being at the heart of the molecular force probe technique.27,28,31 Thus, exploiting the potential of cyclic enediynes to yield highly reactive intermediates as a result of mechanical switching offers a wealth of promising routes to be realized experimentally.

Figure 3. Superimposed optimized structures along the reaction pathway at a constant external force of 2 nN from reactant to product for species 1 in panel a, and for species 2 in panel b; the same rendering is used in the inset of Figure 2b.

The unexpected behavior of diagonal pulling requires closer analysis in order to rationalize the observed effect. The decomposition of the total energy into the electronic structure energy term and the external force term enables one to assess the activation energy change that is exclusively associated with the structural distortion, ΔE‡ES, i.e., without considering the mechanical work that is responsible for this structural response of both reactant and transition states in the first place. Obviously, having the general concept of force-transformed potential energy surfaces32 in mind,11 the molecule is mechanically always forced away from the optimal structure for a given stationary point and, therefore, the change in the electronic energy term is always positive. Inspecting the electronic energy contribution to the activation energy reveals the key difference: it increases for (ii) and decreases strongly for (i), whereas it remains essentially constant for scenario (iii) according to 2d. It is the asymmetrical distortion of TS (see Supporting Information), which is responsible for the very pronounced increase in activation energy observed for species 3. Interestingly, it is also found that only for species 3 another reaction channel, namely, C1−C5 cyclization (see dotted line in the inset of Figure 4), becomes energetically favored over the Bergman (i.e., C1−C6) pathway beyond a critical force, thus



COMPUTATIONAL METHODS All calculations were performed using Gaussian 0935 with the unrestricted B3LYP functional together with the TZVP basis set using our in-house modification for performing the isotensional (EFEI) calculations32 at a set of constant forces. A fully quantitative electronic structure treatment of the 1,4didehydrobenzene intermediate of the Bergman cyclization calls for sophisticated multireference correlation methods. However, it has been shown that computationally economic unrestricted broken symmetry DFT, as used here as a function of force, provides useful energies and structures.15,17,36,37 907

dx.doi.org/10.1021/jz402644e | J. Phys. Chem. Lett. 2014, 5, 905−909

The Journal of Physical Chemistry Letters



Letter

(13) Cramer, C. J. Bergman, Aza-Bergman, and Protonated AzaBergman Cyclizations and Intermediate 2,5-Arynes: Chemistry and Challenges to Computation. J. Am. Chem. Soc. 1998, 120, 6261−6269. (14) Galbraith, J. M.; Schreiner, P. R.; Harris, N.; Wei, W.; Wittkopp, A.; Shaik, S. A Valence Bond Study of the Bergman Cyclization: Geometric Features, Resonance Energy, and Nucleus-Independent Chemical Shift (NICS) Values. Chem.−Eur. J. 2000, 6, 1446−1454. (15) Schreiner, P. R.; Navarro-Vazquez, A.; Prall, M. Computational Studies on the Cyclizations of Enediynes, Enyne-Allenes, and Related Polyunsaturated Systems. Acc. Chem. Res. 2004, 38, 29−37. (16) Gräfenstein, J.; Hjerpe, A. M.; Kraka, E.; Cremer, D. An Accurate Description of the Bergman Reaction Using Restricted and Unrestricted DFT: Stability Test, Spin Density, and On-Top Pair Density. J. Phys. Chem. A 2000, 104, 1748−1761. (17) Gräfenstein, J.; Kraka, E.; Filatov, M.; Cremer, D. Can Unrestricted Density-Functional Theory Describe Open Shell Singlet Biradicals? Int. J. Mol. Sci. 2002, 3, 360−394. (18) Navarro-Vázquez, A.; Prall, M.; Schreiner, P. R. Cope Reaction Families: To Be or Not to Be a Biradical. Org. Lett. 2004, 6, 2981− 2984. (19) Santos, J. C.; Andres, J.; Aizman, A.; Fuentealba, P.; Polo, V. A Theoretical Study on the Reaction Mechanism for the Bergman Cyclization from the Perspective of the Electron Localization Function and Catastrophe Theory. J. Phys. Chem. A 2005, 109, 3687−3693. (20) Alabugin, I. V.; Manoharan, M.; Kovalenko, S. V. Tuning Rate of the Bergman Cyclization of Benzannelated Enediynes with Ortho Substituents. Org. Lett. 2002, 4, 1119−1122. (21) Semmelhack, M. F.; Gallagher, J. J.; Minami, T.; Date, T. The Enol-Keto Trigger in Initiating Arene Diradical Formation in Calicheamicin/Esperamicin Analogs. J. Am. Chem. Soc. 1993, 115, 11618−11619. (22) Semmelhack, M. F.; Wu, L.; Pascal, R. A.; Ho, D. M. Conformational Control in Activation of an Enediyne. J. Am. Chem. Soc. 2003, 125, 10496−10497. (23) Lenhardt, J. M.; Black, A. L.; Craig, S. L. gem-Dichlorocyclopropanes as Abundant and Efficient Mechanophores in Polybutadiene Copolymers under Mechanical Stress. J. Am. Chem. Soc. 2009, 131, 10818−10819. (24) Liang, J.; Fernández, J. M. Mechanochemistry: One Bond at a Time. ACS Nano 2009, 3, 1628−1645. (25) Huang, Z.; Yang, Q.-Z.; Khvostichenko, D.; Kucharski, T. J.; Chen, J.; Boulatov, R. Method to Derive Restoring Forces of Strained Molecules from Kinetic Measurements. J. Am. Chem. Soc. 2009, 131, 1407−1409. (26) Wiggins, K. M.; Hudnall, T. W.; Shen, Q.; Kryger, M. J.; Moore, J. S.; Bielawski, C. W. Mechanical Reconfiguration of Stereoisomers. J. Am. Chem. Soc. 2010, 132, 3256−3257. (27) Huang, Z.; Boulatov, R. Chemomechanics with Molecular Force Probes. Pure Appl. Chem. 2010, 82, 931−951. (28) Huang, Z.; Boulatov, R. Chemomechanics: Chemical Kinetics for Multiscale Phenomena. Chem. Soc. Rev. 2011, 40, 2359−2384. (29) Brantley, J. N.; Wiggins, K. M.; Bielawski, C. W. Unclicking the Click: Mechanically Facilitated 1,3-Dipolar Cycloreversions. Science 2011, 333, 1606−1609. (30) Wiggins, K. M.; Syrett, J. A.; Haddleton, D. M.; Bielawski, C. W. Mechanically Facilitated Retro [4 + 2] Cycloadditions. J. Am. Chem. Soc. 2011, 133, 7180−7189. (31) Kucharski, T. J.; Boulatov, R. The Physical Chemistry of Mechanoresponsive Polymers. J. Mater. Chem. 2011, 21, 8237−8255. (32) Ribas-Arino, J.; Shiga, M.; Marx, D. Understanding Covalent Mechanochemistry. Angew. Chem., Int. Ed. 2009, 48, 4190−4193. (33) Ribas-Arino, J.; Shiga, M.; Marx, D. Mechanochemical Transduction of Externally Applied Forces to Mechanophores. J. Am. Chem. Soc. 2010, 132, 10609−10614. (34) Dopieralski, P.; Anjukandi, P.; Ruckert, M.; Shiga, M.; RibasArino, J.; Marx, D. On the Role of Polymer Chains in Transducing External Mechanical Forces to Benzocyclobutene Mechanophores. J. Mater. Chem. 2011, 21, 8309−8316.

ASSOCIATED CONTENT

S Supporting Information *

Additional analyses, coordinates and complete ref 35. This material is available free of charge via the Internet at http:// pubs.acs.org.



AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Present Address §

M.K.: Max-Planck-Institut für Chemische Energiekonversion, Stiftstrasse 34-36, 45470 Mülheim an der Ruhr, Germany. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We gratefully acknowledge financial support by the Reinhart Koselleck Grant “Understanding Mechanochemistry” (MA 1547/9). The calculation were carried out using resources from NIC Jülich, BOVILAB@RUB, and Rechnerverbund− NRW.



REFERENCES

(1) Nicolaou, K. C.; Ogawa, Y.; Zuccarello, G.; Schweiger, E. J.; Kumazawa, T. Cyclic Conjugated Enediynes Related to Calicheamicins and Esperamicins: Calculations, Synthesis, and Properties. J. Am. Chem. Soc. 1988, 110, 4866−4868. (2) Nicolaou, K. C.; Dai, W. M. Chemistry and Biology of the Enediyne Anticancer Antibiotics. Angew. Chem., Int. Ed. Engl. 1991, 30, 1387−1416. (3) R. Schreiner, P. Cyclic Enediynes: Relationship between Ring Size, Alkyne Carbon Distance, and Cyclization Barrier. Chem. Commun. 1998, 483−484. (4) Schreiner, P. R. Monocyclic Enediynes: Relationships between Ring Sizes, Alkyne Carbon Distances, Cyclization Barriers, and Hydrogen Abstraction Reactions. Singlet-Triplet Separations of Methyl-Substituted p-Benzynes. J. Am. Chem. Soc. 1998, 120, 4184− 4190. (5) Mohamed, R. K.; Peterson, P. W.; Alabugin, I. V. Concerted Reactions That Produce Diradicals and Zwitterions: Electronic, Steric, Conformational, and Kinetic Control of Cycloaromatization Processes. Chem. Rev. 2013, 113, 7089−7129. (6) Wiita, A. P.; Ainavarapu, S. R. K.; Huang, H. H.; Fernandez, J. M. Force-Dependent Chemical Kinetics of Disulfide Bond Reduction Observed with Single-Molecule Techniques. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 7222−7227. (7) Hickenboth, C. R.; Moore, J. S.; White, S. R.; Sottos, N. R.; Baudry, J.; Wilson, S. R. Biasing Reaction Pathways with Mechanical Force. Nature 2007, 446, 423−427. (8) Yang, Q.-Z.; Huang, Z.; Kucharski, T. J.; Khvostichenko, D.; Chen, J.; Boulatov, R. A Molecular Force Probe. Nat. Nanotechnol. 2009, 4, 302−306. (9) Beyer, M. K.; Clausen-Schaumann, H. Mechanochemistry: The Mechanical Activation of Covalent Bonds. Chem. Rev. 2005, 105, 2921−2948. (10) Caruso, M. M.; Davis, D. A.; Shen, Q.; Odom, S. A.; Sottos, N. R.; White, S. R.; Moore, J. S. Mechanically-Induced Chemical Changes in Polymeric Materials. Chem. Rev. 2009, 109, 5755−5798. (11) Ribas-Arino, J.; Marx, D. Covalent Mechanochemistry: Theoretical Concepts and Computational Tools with Applications to Molecular Nanomechanics. Chem. Rev. 2012, 112, 5412−5487. (12) Hickenboth, C. R.; Rule, J. D.; Moore, J. S. Preparation of Enediyne-Crosslinked Networks and Their Reactivity under Thermal and Mechanical Conditions. Tetrahedron 2008, 64, 8435−8448. 908

dx.doi.org/10.1021/jz402644e | J. Phys. Chem. Lett. 2014, 5, 905−909

The Journal of Physical Chemistry Letters

Letter

(35) Frisch, M. J.; et al. Gaussian 09, revision C.01; Gaussian, Inc.: Wallingford, CT, 2009. (36) Marquardt, R.; Balster, A.; Sander, W.; Kraka, E.; Cremer, D.; Radziszewski, J. G. 1,4-Didehydrobenzol. Angew. Chem. 1998, 110, 1001−1005. (37) Sander, W. m-Benzyne and p-Benzyne. Acc. Chem. Res. 1999, 32, 669−676.

909

dx.doi.org/10.1021/jz402644e | J. Phys. Chem. Lett. 2014, 5, 905−909