Product Distribution from Precursor Bite Angle Variation in Multitopic

Subscriber access provided by UNIV OF SCIENCES PHILADELPHIA

Product Distribution from Precursor Bite Angle Variation in Multitopic Alkyne Metathesis: Evidence for a Putative Kinetic Bottleneck Timothy P. Moneypenny, Anna Yang, Nathan P. Walter, Toby J. Woods, Danielle L. Gray, Yang Zhang, and Jeffrey S. Moore J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b02184 • Publication Date (Web): 19 Apr 2018 Downloaded from http://pubs.acs.org on April 19, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Product Distribution from Precursor Bite Angle Variation in Multitopic Alkyne Metathesis: Evidence for a Putative Kinetic Bottleneck Timothy P. Moneypenny, II,1,2 Anna Yang,1 Nathan P. Walter,2,3 Toby J. Woods,4 Danielle L. Gray,4 Yang Zhang,2,3 and Jeffrey S. Moore*,1,2 1

Department of Chemistry, University of Illinois at Urbana—Champaign, Urbana, Illinois 61801, United States 2 Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana−Champaign, Urbana, Illinois 61801, United States 3 Department of Nuclear, Plasma, and Radiological Engineering, University of Illinois at UrbanaChampaign, Urbana, Illinois 61801, United States 4 School of Chemical Sciences, University of Illinois at Urbana—Champaign, Urbana, Illinois 61801, United States

ABSTRACT In the dynamic synthesis of covalent organic frameworks and molecular cages, the typical synthetic approach involves heuristic methods of discovery. While this approach has yielded many remarkable products, the ability to predict the structural outcome of subjecting a multitopic precursor to dynamic covalent chemistry (DCC) remains a challenge in the field. The synthesis of covalent organic cages is a prime example of this phenomenon, where precursors designed with the intention of affording a specific product may deviate dramatically when the DCC synthesis is attempted. As such, rational design principles are needed to accelerate discovery in cage synthesis using DCC. Herein, we test the hypothesis that precursor bite angle contributes significantly to the energy landscape and product distribution in multitopic alkyne metathesis (AM). By subjecting a series of precursors with varying bite angles to AM, we experimentally demonstrate that the product distribution, and convergence towards product formation, is strongly dependent on this geometric attribute. Surprisingly, we discovered that precursors with the ideal bite angle (60º) do not afford the most efficient pathway to the product. The systematic 1 ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 35

study reported here illustrates how seemingly minor adjustments in precursor geometry greatly affect the outcome of DCC systems. This research illustrates the importance of fine-tuning precursor geometric parameters in order to successfully realize desirable targets. INTRODUCTION Dynamic covalent chemistry (DCC)1,

2

is a powerful synthetic method to construct shape-

persistent molecular architectures including macrocycles,3-5 cages,6-12 and covalent organic frameworks.13,

14

The ability of DCC reactions to proceed via thermodynamic control –

ultimately allowing for error correction and convergence towards thermodynamically stable products – provides a means of synthesizing large, complex molecules in one step with high yields on gram scales.1 Implicit in this thermodynamically controlled scenario is the fact that DCC systems explore a multitude of structures and molecular arrangements before finally reaching an energy minimum and corresponding product distribution. Furthermore, DCC provides the ability to re-equilibrate the product distribution by simply adjusting the reaction conditions (e.g., concentration, temperature, or solvent). It is no coincidence that DCC has emerged as the most viable candidate to construct large, complex 2D and 3D molecular architectures comprised solely of covalent bonds. One field in particular – the synthesis of covalent molecular cages – has benefitted from well-known DCC reactions and continues to be a heavily-researched area. The typical approach in this field is to use a heuristic method to design precursors with reasonable geometric attributes required for a targeted structure. With this approach, the topicity (number of reactive functional groups), geometry, and directionality of precursors allow the chemist to propose a reasonable thermodynamically-stable structural outcome. While such an approach is tempting, it occasionally fails to produce the desired outcome or yields no 2 ACS Paragon Plus Environment

Page 3 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


discernible products at all. Although the field of metal-ligand coordination complexes has found great success using heuristic methods,15,

16

the construction of cages via DCC has additional

kinetic considerations and difficulties that must be overcome to achieve reliable architecture formation. In a practical sense, these difficulties arise from finite catalyst lifetimes, slower rates of reactions, and potential kinetic bottlenecks in comparison to coordination complexes.1 Whereas kinetically labile metal-ligand coordination systems siphon through intermediates on the order of seconds,17, 18 DCC systems may take up to several days.6, 19-22 Consequently, slow reaction rates and kinetic bottlenecks may result in extended intermediate lifetimes, affording unanticipated structures including intractable mixtures and complex precipitates that impede the approach toward equilibrium or halt it completely.6, 12, 23, 24 In the context of multitopic (greater than two reactive sites) DCC systems where kinetic factors may intervene, it is useful to frame the problem of reaching the desired product not in terms of a single reaction pathway but rather in terms of a reaction energy landscape, where each point on the landscape corresponds to a molecular entity with a specific internal free energy.25 As such, the DCC reaction is more accurately described as a process which progresses through an ensemble of various structural intermediates and reaction pathways en route toward the thermodynamic product distribution. In this context, the systems that fail to reach a single discrete product suggest that the energy landscape governing the reaction is in either one of two categories: 1) the landscape is too flat—leading to an analogous “Levinthal’s paradox”25 where there is insufficient time to reach the thermodynamic product, or 2) the landscape is too rough— leading to premature kinetic bottlenecks at intermediate stages. It is because of these scenarios that precursor design considerations must take into account not only the thermodynamics of

3 ACS Paragon Plus Environment


Page 4 of 35

products but also that the reaction energy landscape provides pathways to these products on reasonable time scales. As it stands in the field of synthesizing covalent molecular cages via DCC, there is a need for a deeper understanding of how the DCC energy landscape depends on geometric attributes of precursors. Only when a dynamic system proceeds along a smooth energy landscape, or a landscape which does not allow intermediates to irreversibly fall into deep energy minima, will the thermodynamic product be achieved in appreciable yields. Accordingly, precursors must be designed with geometric attributes that direct product distributions along smooth energy lanscapes.26 While several covalent organic cages have been prepared via DCC,9, 27 there are few reports on unsuccessful attempts from which much could be learned. Additionally, while there are no clear set of design principles to form molecular cages, the literature hints that subtle environmental effects or changes in precursor geometric attributes result in the non-intuitive formation of cage products with unexpected architectures. For example, Warmuth and coworkers discovered a solvent dependent cage-forming system using imine condensation, where one precursor selectively formed an octahedron, a tetrahedron, or a square anti-prism depending on the solvent used in the reaction.28 Cooper and co-workers developed a strategy to determine odd-even effects in the synthesis of imine cages where the products from the reactions with an even α,ωalkanediamine carbon chain length afforded tetrahedral [4+6] cages and the products from the reactions with an odd chain length afforded [2+3] cage structures.29 Zhang and co-workers devised a study on altering the size of precursor units. They designed multiple tritopic precursors with ca. 90º carbazole arms, varied the size of the central panel, and observed that the size of the building blocks plays a significant role in determining the resulting structural outcome.


Page 5 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Precursors with smaller panels tended to form D2h tetramers, while precursors with larger panels tended to form dimers.30 Klotzbach and Beuerle discovered the shape-selective formation of bipyramidal, tetrahedral, or cubic organic cages when diboronic acids with bite angles of 60º, 120º, and 180º are reacted, respectively, in the condensation with catechol-functionalized tribenzotriquinacenes.12 These results collectively indicate that multitopic DCC energy landscapes are complex and their product distributions are affected by various reaction parameters that are not always intuitive. To understand multitopic DCC landscapes more completely, we sought to perform systematic studies that experimentally discern design principles for tetrahedral organic cages synthesized via dynamic alkyne metathesis (AM).26, 31 Since this example forms a tetrahedral cage product in nearly quantitative yields, we supposed it would be a key system to investigate the effects of minor adjustments in precursor geometry on the DCC energy landscape. Taking inspiration from work by Fujita,32-34 we hypothesized that precursor bite angle (or the angle between phenylenecontaining arms) significantly alters the dynamic energy landscape and resulting product distribution in multitopic AM. Herein, we test this bite angle hypothesis by synthesizing a series of analogous tritopic precursors with different bite angles—one with a tight bite angle, one with a medium bite angle, and one with a bite angle that matches (within the limits of our measurement technique) the ideal angle of a tetrahedron (60º) —to deliberately alter the DCC energy landscape.25 Subjecting each precursor to identical AM conditions and characterizing the resulting mixtures demonstrates that precursors with ideal bite angles are not the most efficient at forming a tetrahedral molecular cage. Monitoring reaction progress at various time points via gel permeation chromatography (GPC) provides evidence that slight changes in bite angle induce restructuring of the dynamic



Page 6 of 35

AM energy landscape.26 Our results demonstrate that even minor adjustments in precursor geometry significantly bias the dynamic AM energy landscape. Cage formation via AM is thus reminiscent of other reactive systems that venture through multiple intermediates, some of which are on a pathway to the target, while others have strayed from a pathway directed to the target.32, 34

RESULTS AND DISCUSSION Synthesis and Characterization of Precursors. A series of tritopic precursors with varying bite angles were synthesized: PSulf, PCarb, and POxy (Figure 1b), with small, medium, and large bite angles, respectively. The original precursor, PCarb, was synthesized according to a previously reported procedure26 and represents the medium bite angle precursor. The small bite angle precursor, PSulf, was synthesized in high yields from a four-step synthesis, with the key step involving a three-fold nucleophilic aromatic substitution of 1,3,5-tribromo-2,4,6-triethylbenzene using sodium thiophenylate in 1,3-dimethyl-2-imidazolidinone at 140 ºC (Scheme 1).35 The large bite angle precursor, POxy, was synthesized in good yields from an eight-step synthesis (Scheme 1). The key synthetic steps involve a three-fold nucleophilic aromatic substitution of trihydroxy intermediate, 6, with 4-fluoro-1-nitrobenzene using cesium carbonate in dimethylformamide, and iodine-promoted transformation of the tris-(1-aryl-3,3-dialkyltriazene) intermediate, 9, with iodine in diiodomethane.36 The 4-tert-butylbenzyl moities in the POxy precursor were installed for the purposes of solubility. In order to accurately compare the bite angles of each precursor, we utilized single crystal Xray diffraction (XRD) analysis (Figure 2). Single crystals of PSulf were grown from slow evaporation of a 1:1 mixture of dichloromethane and methanol and crystallized in the monoclinic space group P21/n. The X-ray crystal structure exhibits an alternating up-down orientation of the


Page 7 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


central hexasubstituted benzene ring, with up-oriented ethyl groups and down-oriented phenylene arms.37-39 The angle between benzylic edges and the central benzene moiety (using a centroid) was found to be 103.4º on average (Figure S57). To determine the bite angle, three vectors were computed from two atoms on each arm of the precursor (see Figure S59). The angles between vectors were calculated and averaged to give the bite angle for PSulf of 31º. Single crystals of PCarb were grown by slow diffusion of methanol into a solution of ethyl acetate and crystallized in the monoclinic space group P21/n. The crystal structure exhibits the expected alternating up-down orientation of hexasubstituted benzenes. The X-ray crystal structure shows that the angle between the benzylic edges and the central benzene moiety was 118.6º on average (Figure S56). In the fashion described previously (see above), a bite angle of 51º was calculated. Lastly, single crystals of POxy grew by slow diffusion of methanol into a solution of dichloromethane and crystallized in the hexagonal space group P-3c1. This structure also exhibits the expected alternating up-down configuration about the center hexasubstituted benzene. The angle between the benzylic edges and the central benzene moiety was found to be 123.2º on average (Figure S58). The calculated bite angle of 60º matches (within the limits of our measurement technique) the ideal angle for a tetrahedron. While we acknowledge that precursor bite angles may vary slightly in solution, and that crystal packing may bias the measured angle, we believe the calculation from solid state structures provides a reasonable approximation of the average bite angle for each precursor. Single Precursor AM Experiments. To test the bite angle hypothesis, each precursor was subjected to AM with identical conditions. Initial AM experiments were performed at 70 ºC with 10 mM precursor concentration in 1,2,4-trichlorobenzene (TCB) for 12 hours using a



Page 8 of 35

molybdenum catalyst ([Mo], 5 mol %), triphenylsilanol ligand (30 mol %), and molecular sieves (5 Å, 800 mg/mmol of propynyl groups) to sequester the 2-butyne byproduct.40 The results of these experiments are illustrated in Figure 3. Specifically, AM of PSulf affords a product distribution exhibiting a mixture of ill-defined oligomers, as shown by the GPC trace (Figure 3a). The 1H NMR spectrum of this mixture exhibits broad resonances, typical of oligomerization (Figure S48). These observations indicate that within the reaction conditions, there is no preferred discrete molecular entity that forms upon subjecting PSulf to AM for 12 hours at 70 ºC. Subjecting PCarb to AM at 70 ºC for 12 hours affords a very narrow product distribution, and forms exclusively the desired tetrahedral cage (Figure 3b).26 The GPC trace exhibits one sharp peak with a narrow PDI of 1.02. The 1H NMR exhibits sharp peaks, indicative of a highly symmetric, discrete product. Characterization by MALDI-MS and single crystal XRD analysis confirms the formation of a tetrahedral cage26 in near quantitative yield. Repeating this experiment at room temperature for 12 hours affords an identical product distribution with similar yield. Subjecting POxy to AM at 70 ºC affords a product distribution that exhibits signs of both oligomers and a discrete structure as shown by the GPC trace in Figure 3c. The 1H NMR spectrum of this mixture exhibits sharp peaks indicative of a highly symmetric molecule as well as broad resonances at the baseline due to the presence of oligomers (Figure S49). Flash chromatography of this mixture affords a new cage, TdOxy, in 23% yield (Figure 4a). Characterization of the resulting product by NMR and MALDI-MS reveals a highly symmetric compound with a mass corresponding to a tetramer (Figure 4), confirming the formation of TdOxy.


Page 9 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Although these initial experiments indicate that the product distribution of multitopic AM is affected by precursor bite angle, they only provide data for a single set of reaction conditions. To examine how each system behaves over time, the reaction progress was monitored by GPC. Each precursor was subjected to AM using similar conditions described above except at room temperature and in chloroform, and aliquots of the reactions were taken at various time points. The aliquots were then quenched with hydrated chloroform and the product mixtures were characterized via GPC. The results obtained from these experiments are shown in Figure 5. Room temperature reaction conditions were chosen to better sample early reaction progress. Chloroform was used as the solvent to allow direct injection into the GPC. In our hands, there were negligible solvent effects on product distribution during AM of these precursors under these reaction conditions (see Figure S12). In the case of PSulf, the system progresses through a series of intermediates within the first three minutes of the reaction as shown in Figure 5a. As the reaction proceeds, however, the product distribution broadens and enters a regime where formed intermediates are unable to locate a single thermodynamically-stable product. As such, the reaction results in a broad product distribution of ill-defined oligomers within eight hours of reaction time. In the case of PCarb (Figure 5b), initially this precursor forms smaller and larger oligomers, similar to PSulf. As the reaction progresses, however, all of the oligomeric intermediates funnel toward TdCarb, until the entirety of the dynamic mixture converges within 4 hours. The POxy system starts in a similar fashion as the previous two cases, initially forming a mixture of oligomers (Figure 5c). Some intermediates are able to form TdOxy and the concentration of cage increases steadily over time. Other intermediates, however, enter a regime where they continuously grow larger and these intermediates are unable to locate the cage product within eight hours.



Page 10 of 35

The results from these experiments provide evidence to validate the hypothesis that precursor bite angle governs multitopic AM energy landscapes. As shown by PSulf, it was observed that tightening the bite angle directs intermediates to off-target pathways. The result is a broad, illdefined product distribution with undetectable cage formation under the reaction conditions and allotted reaction time. Precursor PSulf thus has a bite angle that is simultaneously too tight to favor the exclusive formation of a tetrahedron and too wide to favor the exclusive formation of a closed dimer under these conditions. Thus, the dynamic system drifts into an oligomeric regime of the energy landscape. The results from POxy are counterintuitive to heuristic design principles based on the geometrically ideal bite angle. Although the bite angle measured for POxy was closest to the ideal angle for a tetrahedron, the early-time product distribution indicates that the system initially ventures off the target pathway. Additionally, while some intermediates are able to form TdOxy, the majority of the dynamic system continues to progress along off-target pathways within eight hours. Hence, precursor POxy provides an example where the bite angle still allows the system to converge toward a discrete cage product, but the rate of convergence is decreased relative to PCarb. If the reaction is allowed to proceed for 24 hours, the product distribution ultimately funnels toward TdOxy in near-quantitative yields (See Figure S11). Finally, precursors with a slightly tighter bite angle than the mathematically ideal angle for a tetrahedron provide the most direct convergence to the tetrahedral cage.33 While the bite angle of PCarb is ca. 9º tighter than the optimum for a tetrahedron, the dynamic mixture achieves nearly quantitative yields of the cage within four hours at room temperature. We deduce that the dynamic energy landscape in this case has a "funneled" surface, allowing the intermediates to


Page 11 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


siphon toward the most thermodynamically-stable product while avoiding other off-target pathways.25 Precursor Mixing Experiments. Although PSulf was not able to form detectable quantities of a tetrahedral cage during AM after 8 hours of reaction time, we hypothesized it is possible to incorporate it into a cage structure using a mixture of the other precursors. To test this hypothesis, we performed mixing experiments between the various precursors. Precursor mixing experiments were performed by reacting PSulf and PCarb, PSulf and POxy, PCarb and POxy, and all three in one pot. The experiments were performed with 10 mM concentration of precursors (total), using 5 mol % [Mo], 30 mol % Ph3SiOH, and 5 Å MS in TCB for 12 hours at room temperature. The GPC traces from the mixtures afforded by these experiments are shown in Figure 6. We first investigated the product distribution formed from mixing PCarb and POxy. Subjecting a 1:1 by mole mixture of PCarb and POxy to AM afforded a well-behaved product distribution containing a molecular weight between that of TdCarb and TdOxy (Figure 6a). Characterization of this mixture by 1H NMR indicated that a distribution of mixed cage products was obtained (Figures S53 and S55). For simplicity, cage composition is denoted in molar ratios by [PSulf:PCarb:POxy]. MALDI-MS analysis confirmed the formation of three mixed cage species [0:3:1], [0:2:2], and [0:1:3] (Figure S14). A similar result was obtained when precursors POxy and PSulf were mixed in a 1:1 molar ratio and subjected to AM (Figure 6b). Characterization by MALDI-MS confirmed the formation of TdOxy ([0:0:4]) and the mixed [1:0:3], [2:0:2], and [3:0:1] cages (Figure S13). Additionally, precursors PCarb and PSulf were mixed in various molar ratios (1:3, 2:2, and 3:1, respectively) and subjected to AM and the product distributions were characterized (Figure 6c). Characterization of each mixture via MALDI-MS confirmed the



Page 12 of 35

formation of TdCarb ([0:4:0]) and the [3:1:0], [2:2:0], and [1:3:0] mixed cages (Figures S15 and S16). As a final experiment, all three precursors were mixed in equimolar concentrations and subjected to AM. As shown by the GPC trace in Figure 6d, a broad product distribution was obtained. Characterization by MALDI-MS confirmed the formation of various cages, including the [2:1:1], [1:2:1], and [1:1:2] mixed cages (Figures S17 and S18). All combination of cages were detected from mixing experiments using MALDI-MS except for the cage comprised solely of PSulf precursors. The obtained results allow us to deduce characteristics of dynamic AM. Although bite angle deviations greatly affect the DCC energy landscape, they do not elicit self-sorting of the various multitopic components within the reaction parameters. Secondly, precursor PSulf is incorporated into cage architectures provided the other precursors in the structure can accommodate the strain. In each mixing experiment, PSulf was successfully incorporated into a molecular cage architecture. DFT Calculations. To further understand the process of cage formation, we performed DFT calculations to determine the relative structural enthalpies (∆E) between the final open tetramer (OT) intermediate (one edge open) and the tetrahedral cage (Td) for each precursor system. In our calculations, the OT intermediate contained two methyl-capped acetylene moieties and the Td cages were all minimized with one 2-butyne molecule as the side product. The DFT calculations were performed in VASP using the PBE functional, with 2000 minimization steps. The results from these calculations, as shown in Figure 7, allow rationalization of the experimentally observed product distributions for each system. In the PSulf system, the corresponding closed cage, TdSulf, is 32.5 kcal/mol higher in enthalpy than the open tetramer (OTSulf). Thus, the final “cage-closing” step in the PSulf system results in significant structural


Page 13 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


strain energy for TdSulf and is not enthalpically favored. To further illustrate this point, the average bite angle in the calculated structure for TdSulf of 45.1º (Figure 7c) is ca. 14º larger than the bite angle calculated from the crystal structure of PSulf, indicating that the precursor must widen its bite angle significantly for successful Td cage formation. In the PCarb system, the tetrahedral cage (TdCarb) is only 4.2 kcal/mol higher in enthalpy than the corresponding open tetramer (OTCarb), indicating that the cage-closing step results in minor cage strain. This strain energy is evident in the crystal structure of TdCarb,26 where the acetylene moieties are slightly bent inwards toward the center ca. 17º from a linear geometry. Lastly, in the POxy system, the tetrahedral cage (TdOxy) is 0.6 kcal/mol lower in enthalpy than the corresponding open tetramer (OTOxy). These results are not surprising, as this precursor was designed to have a bite angle that closely resembled the ideal angle for a tetrahedral cage. Our calculations support the argument that enthalpic considerations alone are insufficient to describe the experimental observations from DCC reactions. Comparison between the PCarb system and the POxy system demonstrates this point: although TdOxy is more enthalpically favored to form than TdCarb (with respect to their corresponding OT intermediates), our experimental results show that TdCarb forms via a more efficient pathway. We hypothesize that this discrepancy is reconciled by kinetic considerations in that there are significant differences in the relative energy barriers for formation of the last metallacyclobutadiene intermediate in the final “cage-closing” step. If correct, this final step becomes a kinetic bottleneck on the DCC energy landscape and this putative bottleneck is governed by the precursor bite angle. Further molecular modeling analysis will be performed in future studies to test this hypothesis. CONCLUSION



Page 14 of 35

In conclusion, a systematic study investigating the effects of precursor bite angle on the DCC energy landscape of multitopic AM was performed. Three precursors were synthesized with bite angles ranging from 31º to 60º, and each was subjected to dynamic AM. It was observed that tightening the bite angle, as illustrated by AM of PSulf, effectively biased the DCC energy landscape and prevented the system from funneling exclusively toward a discrete architecture. On the other hand, widening the bite angle to the ideal tetrahedron geometry, as evidenced by POxy, surprisingly biased the DCC energy landscape so as to lengthen the required time for the system to locate the desired cage product. It was also observed that precursors with angles slightly tighter than the ideal bite angle for a tetrahedron (PCarb) afford the desired cage architecture most efficiently. DFT calculations of relative structural enthalpies allowed us to validate that there exists a discrepancy between our experimental observations and simple, idealized geometric predictions. Given added complexities in DCC systems such as kinetic bottlenecks, an intuitive approach toward designing DCC molecular cage precursors is thus not always predictive. Detailed pathway analyses are likely to supply the missing insight to overcome the design challenges in DCC. Our results lend credence to the underlying systemic issues facing the synthesis of 3D architectures via DCC, where variations in precursor geometry lead to significant deviation of product distributions away from discrete products. Such behavior considerably challenges rational design and limits synthesis of novel covalent organic cage targets. These experiments demonstrate that precursor bite angle dominates energy landscapes and product distributions in multitopic AM. Geometric control of dynamic energy landscapes provides insight toward essential precursor design parameters for successful architecture formation. Our results suggest that to synthesize novel covalent organic cages in high yields via multitopic AM, one must


Page 15 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


design and synthesize precursors with slightly smaller bite angles than that of the theoretically optimum angle for the targeted structure. We envision the information learned from this study will alleviate difficulties in novel architecture synthesis via other DCC chemistries and contribute to rational precursor design rules that allow a priori targeting of structures with specific architectural features. Future efforts will focus on quantifying the observations made here with a pathway model that accounts for the sensitivity in product distribution to the kinetic variations resulting from geometric attributes of the precursor. Such a model is poised to obviate trial-and-error practices for synthesizing novel molecular cages via multitopic DCC. ASSOCIATED CONTENT The Supporting Information is available free of charge on the ACS Publications website at DOI: . See the Supporting Information for experimental conditions and procedures, syntheses, and compound characterizations including GPC, 1H NMR, 13C NMR, MALDI-MS, and single crystal XRD analysis. Single crystal crystallographic data for this paper have been deposited at the Cambridge Crystallographic Data Centre under deposition numbers CCDC 1590139 (PSulf), CCDC 1590140 (PCarb), and CCDC 1590138 (POxy). These data can be obtained free of charge from www.ccdc.cam.ac.uk/data_request/cif. AUTHOR INFORMATION Corresponding Author *[email protected] Notes The authors declare no competing financial interests. ACKNOWLEDGEMENTS



Page 16 of 35

This work was supported by the National Science Foundation CHE under Grant Number 1610328. T. P. M. II gratefully acknowledges the Arnold and Mabel Beckman Foundation for a Beckman Institute Graduate Fellowship and the Drickamer family for a Harry G. Drickamer Fellowship. T. P. M. II thanks Prof. Semin Lee for helpful discussions. The authors thank the School of Chemical Sciences NMR Lab and Mass Spectrometry Lab and the George L. Clark Xray Facility. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8.

9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33.

S. J. Rowan, S. J. Cantrill, G. R. L. Cousins, J. K. M. Sanders and J. F. Stoddart, Angew. Chem. Int. Ed., 2002, 41, 898952. Y. Jin, C. Yu, R. J. Denman and W. Zhang, Chem. Soc. Rev., 2013, 42, 6634-6654. W. Zhang and J. S. Moore, J. Am. Chem. Soc., 2005, 127, 11863-11870. A. J. Gallant and M. J. MacLachlan, Angew. Chem. Int. Ed., 2003, 42, 5307-5310. S. W. Sisco, B. M. Larson and J. S. Moore, Macromolecules, 2014, 47, 3829-3836. M. Mastalerz, ChemComm., 2008, DOI: 10.1039/B808990F, 4756-4758. D. Xu and R. Warmuth, J. Am. Chem. Soc., 2008, 130, 7520-7521. T. Tozawa, J. T. A. Jones, S. I. Swamy, S. Jiang, D. J. Adams, S. Shakespeare, R. Clowes, D. Bradshaw, T. Hasell, S. Y. Chong, C. Tang, S. Thompson, J. Parker, A. Trewin, J. Bacsa, A. M. Z. Slawin, A. Steiner and A. I. Cooper, Nat. Mater., 2009, 8, 973-978. M. Mastalerz, Angew. Chem. Int. Ed., 2010, 49, 5042-5053. J. W. Colson, A. R. Woll, A. Mukherjee, M. P. Levendorf, E. L. Spitler, V. B. Shields, M. G. Spencer, J. Park and W. R. Dichtel, Science, 2011, 332, 228-231. C. Yu, Y. Jin and W. Zhang, Chem. Rec., 2015, 15, 97-106. S. Klotzbach and F. Beuerle, Angew. Chem. Int. Ed., 2015, 54, 10356-10360. A. P. Côté, A. I. Benin, N. W. Ockwig, M. O'Keeffe, A. J. Matzger and O. M. Yaghi, Science, 2005, 310, 1166-1170. S.-Y. Ding and W. Wang, Chem. Soc. Rev., 2013, 42, 548-568. S. R. Seidel and P. J. Stang, Acc. Chem. Res, 2002, 35, 972-983. T. R. Cook and P. J. Stang, Chem. Rev., 2015, 115, 7001-7045. T.-Z. Xie, K. J. Endres, Z. Guo, J. M. Ludlow, C. N. Moorefield, M. J. Saunders, C. Wesdemiotis and G. R. Newkome, J. Am. Chem. Soc., 2016, 138, 12344-12347. X. Lu, X. Li, K. Guo, T.-Z. Xie, C. N. Moorefield, C. Wesdemiotis and G. R. Newkome, J. Am. Chem. Soc., 2014, 136, 18149-18155. A. G. Slater, M. A. Little, PulidoA, S. Y. Chong, HoldenD, ChenL, MorganC, WuX, ChengG, ClowesR, M. E. Briggs, HasellT, K. E. Jelfs, G. M. Day and A. I. Cooper, Nat. Chem., 2017, 9, 17-25. A. R. Stefankiewicz and J. K. M. Sanders, ChemComm, 2013, 49, 5820-5822. N. Ponnuswamy, F. B. L. Cougnon, J. M. Clough, G. D. Pantoş and J. K. M. Sanders, Science, 2012, 338, 783-785. F. B. L. Cougnon, N. Ponnuswamy, N. A. Jenkins, G. D. Pantoş and J. K. M. Sanders, J. Am. Chem. Soc., 2012, 134, 19129-19135. M. W. Schneider, I. M. Oppel and M. Mastalerz, Chem. Eur. J, 2012, 18, 4156-4160. J. Sun and R. Warmuth, ChemCommun, 2011, 47, 9351-9353. K. A. Dill and H. S. Chan, Nat. Struct. Mo.l Biol., 1997, 4, 10-19. S. Lee, A. Yang, T. P. Moneypenny and J. S. Moore, J. Am. Chem. Soc., 2016, 138, 2182-2185. T. Hasell and A. I. Cooper, Nat. Rev. Mater., 2016, 1, 16053. X. Liu and R. Warmuth, J. Am. Chem. Soc., 2006, 128, 14120-14127. K. E. Jelfs, E. G. B. Eden, J. L. Culshaw, S. Shakespeare, E. O. Pyzer-Knapp, H. P. G. Thompson, J. Bacsa, G. M. Day, D. J. Adams and A. I. Cooper, J. Am. Chem. Soc., 2013, 135, 9307-9310. Q. Wang, C. Yu, C. Zhang, H. Long, S. Azarnoush, Y. Jin and W. Zhang, Chem. Sci, 2016, 7, 3370-3376. A. Fürstner, Angew. Chem. Int. Ed., 2013, 52, 2794-2819. J. Bunzen, J. Iwasa, P. Bonakdarzadeh, E. Numata, K. Rissanen, S. Sato and M. Fujita, Angew. Chem. Int. Ed., 2012, 51, 3161-3163. M. Yoneya, S. Tsuzuki, T. Yamaguchi, S. Sato and M. Fujita, ACS Nano, 2014, 8, 1290-1296.


Page 17 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


34. 35. 36. 37. 38. 39. 40.

Q.-F. Sun, J. Iwasa, D. Ogawa, Y. Ishido, S. Sato, T. Ozeki, Y. Sei, K. Yamaguchi and M. Fujita, Science, 2010, 328, 1144-1147. J. H. R. Tucker, M. Gingras, H. Brand and J.-M. Lehn, J. Chem. Soc., Perkin Trans. 2, 1997, DOI: 10.1039/A608455I, 1303-1308. Z. Wu and J. S. Moore, Tetrahedron Lett., 1994, 35, 5539-5542. S. Simaan, J. S. Siegel and S. E. Biali, J. Org. Chem., 2003, 68, 3699-3701. J. Kim, Y. K. Kim, N. Park, J. H. Hahn and K. H. Ahn, J. Org. Chem., 2005, 70, 7087-7092. T. Terada, T. Wakimoto, T. Nakamura, K. Hirabayashi, K. Tanaka, J. Li, T. Matsumoto and K. Tatsumi, Chem. Asian J., 2012, 7, 920-929. J. Heppekausen, R. Stade, R. Goddard and A. Fürstner, J. Am. Chem. Soc., 2010, 132, 11045-11057.



Page 18 of 35

TOC Graphic


Page 19 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Main Text Figures:

Figure 1: (a) Synthesis of a tetrahedral organic cage via AM. TCB = 1,2,4-trichlorobenzene; MS = molecular sieves. (b) A series of analogous tritopic precursors synthesized in this study. The bite angle is defined as the angle between phenylene-containing arms and is determined via Xray crystallography. R = 4-tert-butylbenzene.



Page 20 of 35

Scheme 1: Synthesis of PSulf and POxy precursors.a

aReagents

and conditions: (a) bromine (5 equiv), Fe powder (0.3 equiv), 10 min, rt, 98 %; (b) sodium thiophenolate (7 equiv), DMI, 140 ºC, 3 d, 72 %; (c) NBS (3.3 equiv), DCM/MeCN 2:1 (v/v), 0 ºC, 15 min then rt, 2 d, 74 %; (d) propynyl magnesium bromide (5 equiv), Pd(dppf)Cl2 (0.06 equiv), THF, 65 ºC, 2 d, 90 %; (e) paraformaldehyde (9 equiv), HBr in acetic acid (9.4 equiv), 85 ºC, 38 %; (f) 1-bromo-4tert-butylbenzene (5 equiv), Mg (10 equiv), benzene, 100 ºC, 16 hr, 70 %; (g) BBr3 (4.3 equiv), DCM, 0 ºC → rt, 14 hr, 99 %; (h) 1-fluoro-4-nitrobenzene (5 equiv), Cs2CO3 (5 equiv), DMF, 90 ºC, 6 hr, 84 %; (i) Pd/C (0.10 mass equiv), hydrogen atmosphere, EtOAc, rt, 24 hr, 99 %; (j) HCl (12 equiv), NaNO2 (4.5 equiv), MeCN/H2O 2:1 (v/v), 0 ºC, 30 mins; (k) K2CO3 (15.6 equiv), pyrrolidine (7.5 equiv), MeCN/H2O 1:1 (v/v), 0 ºC → rt, 1 hr, 93 % over two steps; (l) iodine (3.1 equiv), diiodomethane, 80 ºC, 4 hr, 90 %; (m) propyne atmosphere, CuI (0.1 equiv), Pd(PPh3)2Cl2 (0.08 equiv), Et3N/THF 5:1 (v/v), rt, 12 hr, 77 %. DMI = 1,3-dimethyl-2-imidazolidinone, NBS = N-bromosuccinimide, BBr3 = boron tribromide.


Page 21 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2: X-ray crystal structures of (a) PSulf, (b) PCarb, and (c) POxy shown as ellipsoids at the 50% confidence level. Hydrogen atoms and the 4-tert-butylbenzyl moieties for POxy are removed for clarity.



Page 22 of 35

Figure 3: GPC chromatograms of the product distributions afforded after AM of (a) PSulf, (b) PCarb, and (c) POxy at 70 ºC for 12 hours.


Page 23 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4: Purification of the product distribution afforded from subjecting POxy to AM at 70 ºC. (a) Structure of TdOxy. (b) GPC traces before (blue) and after (black) purification. (c) MALDIMS spectrum of purified product using a DCTB matrix. (d) 1H NMR (400 MHz, d-chloroform) comparison of POxy (top) and TdOxy (bottom). The red dotted box corresponds to the propynyl proton resonance.



Page 24 of 35

Figure 5: GPC chromatograms taken after various time points of AM of (a) PSulf, (b) PCarb, and (c) POxy at room temperature with 5 mol % [Mo] catalyst. Red arrows denote cage products.


Page 25 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 6: GPC chromatograms of the product distribution afforded from precursor mixing experiments. (a) GPC traces of TdCarb (blue), TdOxy (red), and the 1:1 mixture of PCarb and POxy (pink). (b) GPC traces of TdOxy (red), the product distribution from AM of PSulf (black), and the 1:1 mixture of POxy and PSulf (orange). (c) normalized (by area) GPC traces of TdCarb (black), the 1:3 mixture (pink), the 1:1 mixture (blue), the 3:1 mixture (red) of PCarb and PSulf, respectively, and the product distribution of PSulf (green). (d) GPC traces of TdCarb (blue), TdOxy (red), the product distribution after AM of PSulf (black), and the 1:1:1 mixture of PSulf, PCarb, and POxy (green).



Page 26 of 35

Figure 7: Results from DFT calculations. (a) Representative “cage closing” metathesis reaction starting from the open tetramer of PSulf (OTSulf) to form TdSulf. (b) Plot of calculated relative structural energies, ∆E (relative to the open tetramer) vs. bite angle (calculated from DFT) of each system. All values in kcal/mol. (c) Overlay of the energy minimized structures of TdSulf (blue), TdCarb (orange), and TdOxy (green). Hydrogen atoms are removed for clarity. From these structures, the bite angles (ϴ) were calculated as 45.1º, 52.8º, and 58.5º for TdSulf, TdCarb, and TdOxy respectively. DFT, PBE functional, 2000 minimizations steps.


Page 27 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1: (a) Synthesis of a tetrahedral organic cage via AM. TCB = 1,2,4-trichlorobenzene; MS = molecular sieves. (b) A series of analogous tritopic precursors synthesized in this study. The bite angle is defined as the angle between phenylene-containing arms and is determined via X-ray crystallography. R = 4-tertbutylbenzene. 232x165mm (150 x 150 DPI)

ACS Paragon Plus Environment


Scheme 1: Synthesis of PSulf and POxy precursors.a aReagents and conditions: (a) bromine (5 equiv), Fe powder (0.3 equiv), 10 min, rt, 98 %; (b) sodium thiophenolate (7 equiv), DMI, 140 ºC, 3 d, 72 %; (c) NBS (3.3 equiv), DCM/MeCN 2:1 (v/v), 0 ºC, 15 min then rt, 2 d, 74 %; (d) propynyl magnesium bromide (5 equiv), Pd(dppf)Cl2 (0.06 equiv), THF, 65 ºC, 2 d, 90 %; (e) paraformaldehyde (9 equiv), HBr in acetic acid (9.4 equiv), 85 ºC, 38 %; (f) 1-bromo-4-tert¬-butylbenzene (5 equiv), Mg (10 equiv), benzene, 100 ºC, 16 hr, 70 %; (g) BBr3 (4.3 equiv), DCM, 0 ºC → rt, 14 hr, 99 %; (h) 1-fluoro-4-nitrobenzene (5 equiv), Cs2CO3 (5 equiv), DMF, 90 ºC, 6 hr, 84 %; (i) Pd/C (0.10 mass equiv), hydrogen atmosphere, EtOAc, rt, 24 hr, 99 %; (j) HCl (12 equiv), NaNO2¬ (4.5 equiv), MeCN/H¬2O 2:1 (v/v), 0 ºC, 30 mins; (k) K2CO3 (15.6 equiv), pyrrolidine (7.5 equiv), MeCN/H2O 1:1 (v/v), 0 ºC → rt, 1 hr, 93 % over two steps; (l) iodine (3.1 equiv), diiodomethane, 80 ºC, 4 hr, 90 %; (m) propyne atmosphere, CuI (0.1 equiv), Pd(PPh¬3)2Cl2 (0.08 equiv), Et3N/THF 5:1 (v/v), rt, 12 hr, 77 %. DMI = 1,3-dimethyl-2-imidazolidinone, NBS = Nbromosuccinimide, BBr3 = boron tribromide. 202x166mm (150 x 150 DPI)


Page 28 of 35

Page 29 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3: GPC chromatograms of the product distributions afforded after AM of (a) PSulf, (b) PCarb, and (c) POxy at 70 ºC for 12 hours. 120x96mm (220 x 220 DPI)



Figure 6: GPC chromatograms of the product distribution afforded from precursor mixing experiments. (a) GPC traces of TdCarb (blue), TdOxy (red), and the 1:1 mixture of PCarb and POxy (pink). (b) GPC traces of TdOxy (red), the product distribution from AM of PSulf (black), and the 1:1 mixture of POxy and PSulf (orange). (c) normalized (by area) GPC traces of TdCarb (black), the 1:3 mixture (pink), the 1:1 mixture (blue), the 3:1 mixture (red) of PCarb and PSulf, respectively, and the product distribution of PSulf (green). (d) GPC traces of TdCarb (blue), TdOxy (red), the product distribution after AM of PSulf (black), and the 1:1:1 mixture of PSulf, PCarb, and POxy (green). 231x114mm (150 x 150 DPI)


Page 30 of 35

Page 31 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5: GPC chromatograms taken after various time points of AM of (a) PSulf, (b) PCarb, and (c) POxy at room temperature with 5 mol % [Mo] catalyst. Red arrows denote cage products. 338x158mm (150 x 150 DPI)



Figure 4: Purification of the product distribution afforded from subjecting POxy to AM at 70 ºC. (a) Structure of TdOxy. (b) GPC traces before (blue) and after (black) purification. (c) MALDI-MS spectrum of purified product using a DCTB matrix. (d) 1H NMR (400 MHz, d-chloroform) comparison of POxy (top) and TdOxy (bottom). The red dotted box corresponds to the propynyl proton resonance. 365x287mm (150 x 150 DPI)


Page 32 of 35

Page 33 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2: X-ray crystal structures of (a) PSulf, (b) PCarb, and (c) POxy shown as ellipsoids at the 50% confidence level. Hydrogen atoms and the 4-tert-butylbenzyl moieties for POxy are removed for clarity. 331x95mm (150 x 150 DPI)



Figure 7: Results from DFT calculations. (a) Representative “cage closing” metathesis reaction starting from the open tetramer of PSulf (OTSulf) to form TdSulf. (b) Plot of calculated relative structural energies, ∆E (relative to the open tetramer) vs. bite angle (calculated from DFT) of each system. All values in kcal/mol. (c) Overlay of the energy minimized structures of TdSulf (blue), TdCarb (orange), and Td¬Oxy (green). Hydrogen atoms are removed for clarity. From these structures, the bite angles (ϴ) were calculated as 45.1º, 52.8º, and 58.5º for TdSulf, TdCarb, and TdOxy respectively. DFT, PBE functional, 2000 minimizations steps. 270x544mm (150 x 150 DPI)


Page 34 of 35

Page 35 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table of Contents Image 181x77mm (150 x 150 DPI)


Product Distribution from Precursor Bite Angle Variation in Multitopic

Recommend Documents