id
stringlengths 24
24
| idx
int64 0
402
| paragraph
stringlengths 106
17.2k
|
|---|---|---|
67174e2d83f22e4214c6cc2c
| 5 |
The acceptors were synthesized following the procedures shown in Scheme 2. 2 was converted into the corresponding sulfonyl amide 3 by first converting it into the sulfonyl chloride, followed by quenching with ammonium hydroxide. Next, both enantiomers of 3 were reacted with troc-chloride. Interestingly, if the reaction is left for 20 hours, a ditrocylation occurs yielding compound 4-(S) as the major product, whereas, if the reaction time is reduced to 30 minutes compound 5-(R) is obtained through monotrocylation. However, following a procedure from Huang et al. it was possible to selectively remove only one of the troc-groups from 4-(S) to obtain 5-(S).
|
67174e2d83f22e4214c6cc2c
| 6 |
With the chiral acceptors in hand, the glycosylations in Figure were carried out, with both 1-α or 1-β as the donor, and the reactions were followed by in situ IRspectroscopy. The kinetic trace from a distinct peak at 1670 cm -1 from the donor was used to fit different integrated rate laws, and it was found that the reactions seem to follow the rate law for a bimolecular reaction with different reactants of different concentrations (See SI Section 3.1 for details). The fitted kinetic traces are shown in the plot in Figure , and the calculated rate constants can be found in Table . No major differences in the rates are observed, although the glycosylations with 5-(R) are slightly slower than the glycosylations with 5-(S). The rate of the glycosylations with 5-(R) seems to be independent of the donor configuration, whereas, for the glycosylations with 5-(S), 1-α seems to be a better match for the acceptor. Despite no major differences in the rate between the glycosylations with 1-α and the glycosylations with 1-β, a great difference in anomeric selectivity is observed. For the glycosylations with 1-α only the β-product is observed but for the glycosylations with 1-β a mixture of anomers is observed, which is in agreement with previous observations. 14 Table : Rate constants determined from second order integrated rate law plot in Figure and α : β ratio of N -glucoside product with standard deviations.
|
67174e2d83f22e4214c6cc2c
| 7 |
The peaks give a similar kinetic trace when normalized, and the decreasing and increasing peaks are mirrored through a horizontal mirror plane, with one exception. For the glycosylations with 1-β the peak at 1590 -1 shows a different behavior as it initially increases and afterward decreases until it reaches a plateau. Upon closer inspection of the peaks as a function of time (Figure B) it is evident that this difference is a consequence of an additional overlapping peak, which is not observed for the glycosylations with 1-α (right plot Figure ). It is seen that for the glycosylations with 1-β (left plot Figure ) the peak has a shoulder to the left, which initially increases and reaches a maximum after 2 hours (purple line), and after that starts decreasing to disappear after 6 hours (black line). This indicates that an intermediate is formed and builds up during the glycosylation with the β-TCA.
|
67174e2d83f22e4214c6cc2c
| 8 |
To further investigate this, the reactions were followed by NMR. As the peak trends were nearly identical for the two acceptors, only reactions with the (S)-acceptor were investigated by NMR. Two reactions were carried out in NMR tubes, one with the 1-α donor and one with the 1-β donor (Figure ). 1 to 2 spectra were recorded each hour and the stacked spectra are shown in Figure . For the reaction with 1-α, a relatively simple reaction picture is observed (3B left), where the signal from the α-TCA (1-α) decreases, while the signal from the β-product (6β) increases. Further, a small formation of hemiacetal side-product is observed. A more complicated picture is observed for the reaction with the β-TCA as seen in the stacked spectra to the right in Figure . As expected, the signal from the β-TCA decreases over time, and formation of the α-product (6-α) is observed. In addition, a significant amount of hemiacetal and β-product is formed during the reaction. It is also seen that α-TCA is formed during the reaction, which would inevitably react to form the β-product. Isomerization of the β-TCA has previously been observed for this type of reaction. Interestingly, a signal, that initially increases and afterward decreases, is observed to the left of the signal from the α-product (Figure right). As this behavior is similar to the behavior observed by IR, it is likely that it is the same intermediate giving rise to both peaks. However, it should be noted that the findings from the NMR experiments cannot be completely transferred to the observations in the IR experiments, as the concentration is lower, 0.044 M in the NMR experiments vs 0.2 M in the IR experiments, and also the solvent has been changed from CH 2 Cl 2 to CDCl 3 . In Figure , the integral of the peaks from the observed species is plotted as a function of time. From these plots the difference in product distribution is very pronounced, indicating that one major pathway is dominating for the reaction with 1-α, whereas for the reaction with 1-β there are multiple competing pathways.
|
67174e2d83f22e4214c6cc2c
| 9 |
The identity of the unknown intermediate was still not clear after the NMR experiments, and therefore the glycosylation with the 1-β and the (R)-acceptor in Figure was carried out again, this time with the aim of trapping the intermediate. Thus the reaction was stopped after two hours, as the IR data suggested that the concentration of the intermediate would be at its maximum at this point. An NMR spectrum of the crude reaction mixture was recorded after a basic aqueous workup. The spectrum and experimental details can be found in SI (3.1.6). Both α-TCA, α-product, β-product, and hemiacetal were observed in the crude reaction mixture, but not the unknown intermediate. However, an observed J 1,2 of 3.40 Hz for the experiments in Figure indicates that the intermediate is an α-glycoside. In addition, an HSQC spectrum recorded during the kinetic NMR experiments (SI Figure , as it would be in accordance with the NMR data, and similar O-glycosides have been observed in self-promoted N -glycosylations with amide acceptors, as stable intermediates (Figure ). The distribution of O-glycoside and N -glycoside was shown to be solvent independent and the O-glycoside would rearrange to the N -glycoside upon heating. Thus, it is expected that the α-O-glycoside rearranges to the N -glycoside. Furthermore, the O-glycoside in Figure and similar glycosides are reported to give rise to a peak around 1627 cm -1 in IR-spectra, which is similar to the peak observed from the unknown intermediate (1600 -1625 cm -1 ).
|
67174e2d83f22e4214c6cc2c
| 10 |
A computational study was initiated to probe the mechanism further. The study comprises two parts: one focuses on investigating the importance of the conformation of the TCA donor and the second focuses on locating a possible mechanistic pathway. To reduce the computational cost the benzyl groups were replaced with methyl groups. It was previously suggested by Nielsen et. al. that the differences in the degree of stereoselectivity between the α-TCA and the β-TCA are caused by the conformation of the imidate leaving group i.e. whether the imidate nitrogen is pointing "up" or "down". The full explanation can be found in the original paper by Nielsen et al., and an illustration of this initial explanation can also be found in SI (Section 4). To investigate this hypothesis a relaxed scan of the TCA group was carried out i.e. a geometry optimization for different frozen dihedral angles. The scans are illustrated in Figure . By looking at the TCA donors it is evident that a rotation around the imidate oxygen and the imidate carbon would lead to a change in the direction of the imidate nitrogen, this corresponds to changing the dihedral angle between the anomeric carbon, the imidate oxygen, the imidate carbon and the imidate nitrogen, here denoted (C-1)-O-C=N (blue arrow in Fig- ). However, for the β-TCA a rotation around the anomeric bond (red arrow in Figure ) would also lead to a change in the imidate nitrogen's position relative to the carbohydrate ring "plane", which is not the case for the α-TCA. Hence, two scans for the β-TCA were performed, while only one scan was performed for the α-TCA. It is seen in Figure top, that the conformation with the nitrogen pointing down is not a stable conformation for the α-TCA and is instead the transition state for the rotation around the bond. This is in line with early MM2 calculations by Schmidt. When rotating around the anomeric bond in the β-TCA a different picture is observed, as multiple minima are observed on the potential energy surface. It is seen that the global minimum (starting/end geometry) has the nitrogen pointing down with a slight twist toward the endocyclic oxygen. A local mini-mum with nitrogen pointing upwards is also observed at a dihedral angle of 60 degrees. For the (C-1)-O-C=N scan of the β-TCA a shallow minimum with nitrogen pointing up is also observed at 125 degrees. Interestingly, upon further rotation, the scan ends with an optimized geometry resembling the local minimum found in the anomeric bond scan of the β-TCA. In total three minima are found for the β-TCA, a global which has the nitrogen pointing down (Conformation 1 Figure ), and two local which have the nitrogen pointing up (Conformation 2 and Conformation 3 Figure ). To investigate the influence of the different conformations the Boltzmann distribution was calculated, and the probability of each conformation (p i ) can be found in Figure . It is seen that the distribution is heavily skewed towards Conformation 1. Based on these results it seems unlikely that two competing pathways starting from different conformations of the two TCA donors are the reason for the observed difference in selectivity as initially proposed. However, a difference in the favorability of trajectory for attack in the formed ion pairs, which was also a part of the initial explanation, could still play a role. To investigate this we sought to locate a (or multiple) possible mechanistic pathways on the potential energy surface.
|
67174e2d83f22e4214c6cc2c
| 11 |
We used a non-chiral sulfonyl carbamate acceptor 14 for the mechanistic study to lower the computational cost and a possible mechanistic pathway was successfully located for the glycosylations with the α-TCA donor and a depiction can be found in Figure . The structures of the located stationary points are illustrated in 2D in Figure .
|
67174e2d83f22e4214c6cc2c
| 12 |
The reaction is initiated by formation of a reactant complex with a hydrogen bond between the carbamate hydrogen and imidate nitrogen. The complex undergoes a proton transfer via TS1, to form an intermediate ion pair (INT1). This ion pair can change conformation to a new intermediate ion pair (INT2), where the TCA-group is rotated so that one of the protonated imine hydrogens forms a hydrogen bond to the oxygen at C-2 and the carbamate surrounds the TCA-donor. Then the ion pair passes through TS2, the transition state for leaving group departure, which yields the oxocarbenium-like ionpair INT3. INT3 has the deprotonated carbamate nitrogen in the right position for attack of the anomeric carbon to give the β-N -glucoside product via TS3. It is seen from the reaction coordinate, that INT1 is higher in energy than TS1, which is not physically possible. However, looking at Table S.4 (Section 6.1), it is evident that the pure DFT energy is higher for the transition state, but when including zero-point vibrational energies and thermodynamic contributions, it becomes lower. This is most likely due to approximations made when calculating the zero-point vibrational and thermodynamic contributions. Hence, TS1 and INT1 must be very similar in energy, which is also illustrated by their almost identical structure, in line with the Hammond postulate. Multiple attempts to locate a more S N 2-like pathway failed, and we therefore suggest that the reaction mainly proceeds via this more S N 1-like pathway. This is also in line with previous observations by Pinna et al. In Figure the transition states with displacement vectors, which show the motion of the molecule during the transition state, are displayed. It is seen that during the proton transfer (TS1), the motion is localized to the proton, whereas for TS2, the anomeric carbon and the endocyclic oxygen move away from the leaving group to form the oxocarbenium-ion like intermediate, while the leaving group is departing. In TS3 the vectors illustrate the movement of the anomeric carbon and carbamate nitrogen towards each other. This is in accordance with an S N 1-like reaction initiated by a proton transfer.
|
67174e2d83f22e4214c6cc2c
| 13 |
As the kinetic study showed formation of multiple intermediates and products for the glycosylation with the β-TCA it was decided that a similar study for this reaction would not be conducted, as it would require the pathway for the formation of all intermediates and products to be located on the potential energy surface. Due to the similar rate constants calculated for the glycosylation with the α-TCA and the β-TCA we hypothesize that the reaction between the β-TCA and the acceptor, and all side reactions, occurs through a similar mechanism to that in Figure and possibly via the same intermediate.
|
67174e2d83f22e4214c6cc2c
| 14 |
Donor Catalyst T ( From Figure it is seen that the donor and acceptor are very close in space after INT2 is formed. This leads to a favorable trajectory for attack by the deprotonated acceptor on the opposite side of the leaving group, which results in inversion of the stereochemistry. This is also in line with a loss of inversion when the glycosyl donor has an axial C-2. Thus to try to explain the loss of selectivity and the observed side products in glycosylations with the β-TCA we decided to search for similar ion pairs between the β-TCA and the acceptor. Two ion pairs were located, and these are shown in Figure . Due to the β-configuration, the protonated imidate cannot hydrogen bond to the O-2 oxygen as in INT2. Instead, the protonated imidate is hydrogen bonded to the endocyclic oxygen in both INTβa and INTβb, thus the carbamate is further away in space than for the corresponding ion pair with the α-TCA (INT2). As a consequence, the carbamate can be orientated, so that an attack would lead to retention of stereochemistry (INTβa). Additionally, even when the carbamate is positioned in a way that would lead to inversion (INTβb) the carbamate nitrogen is further away from the anomeric carbon (5 Å) compared to the distance in INT2 (3.4 Å) which might leave room for competing nucleophiles and explain the increased amount of side products. To investigate this hypothesis, we decided to conduct a competition experiment.
|
67174e2d83f22e4214c6cc2c
| 15 |
The experiments in Table were carried out to see if the carbamate would act as a catalyst rather than a self-promoter in the presence of an excess of competing nucleophile. We first confirmed that the competing nucleophile, isopropanol, does not react with the donor without addition of the carbamate "catalyst". This was confirmed by Entry 1, as only the starting material was recovered. When 0.2 equivalent of the carbamate is added the major product becomes the α-O-glycoside at both 0 • C and room temperature (Entry 2 and 3). The formation of the O-glycoside is highly α-selective suggesting that the carbamate is still a part of an associated ion pair and positioned so that either the carbamate or leaving group is blocking for attack from the upper side of the ring. However, the carbamate is not a good catalyst since N -glycoside product is also observed and thereby is consumed during the reaction. Such species formed by an irreversible reaction between a catalyst and a glycosyl donor have earlier been coined the rebound product by Miller. By comparing Entry 3 and Entry 4 it is seen that if the donor is the α-TCA the major product is the β-N -glycoside product with the β-O-glycoside product as a minor product. This is in line with the α-TCA carbamate ion pair being tighter than the β-TCA carbamate ion pair. Intrigued by the catalytic, albeit suboptimal, properties of the carbamate we hypothesized that if the formation of the rebound product (i.e. the N -glycoside) could be suppressed a new procedure for catalytic α-selective O-glycosylation could be developed. We tried several different carbamate and alcohol combinations but for all of them, the rebound product was still observed (SI Section 5). However, the insights from the mechanism of selfpromoted N -glycosylations have led us on the track of the discovery of new selective catalysts for glycosylation. We envision that if the formation of a nucleophilic counterion upon activation of the TCA is prevented either by steric hindrance or by decreasing the acidity of the catalyst so that the TCA is activated but without full proton transfer, stereospecific catalysis can be obtained. This is illustrated in Figure . Very acidic catalysts can form stable counterions, which give rise to unstable rebound products that are subject to anomerization, and the stereochemical information from the donor is often lost. This can, for instance, be observed with triflates 6, 8 or paratoluenesulfonic acid. If the acidity is decreased, the counterion becomes less stable and the rebound product becomes stable, making the reaction between the donor and counterion irreversible. In this way, the activator is no longer a catalyst but instead a self-promoter. By decreasing the acidity to the point where full protonation does not occur, but the TCA is still activated enough for reaction with another nucleophile might lead to transition back into a catalytic window with stereospecificity.
|
67174e2d83f22e4214c6cc2c
| 16 |
From all of the above findings, we propose the mechanistic pathway in Figure for self-promoted Nglycosylations. The high stereoselectivity observed for the α-TCA is explained by the formation of an ion pair with a very favorable trajectory, where the leaving group is hydrogen bonded to O-2 (A). A similar ion pair cannot be formed by the β-TCA, where the leaving group instead will be hydrogen bonded to the endocyclic oxygen, which means that the nucleophilic glycosyl acceptor will be in an unfavorable position for attack (B). Furthermore, a cavity, which allows for attack by competing nucleophiles, is created, leading to the formation of multiple competing pathways.
|
67174e2d83f22e4214c6cc2c
| 17 |
In summary, the mechanism of self-promoted Nglycosylations has been thoroughly investigated. Both IR and NMR experiments have been performed, and the importance of the conformation of the TCA donors has been investigated through computational studies. Furthermore, a possible mechanistic pathway for the reaction of a sulfonyl carbamate acceptor with an α-TCA donor has been located on the potential energy surface using DFT calculations. Experiments with a competing nucleophile were carried out, which helped substantiate the mechanism, but also revealed that the N -sulfonyl carbamate acceptors have some catalytic properties. In the proposed mechanistic scenario, a favorable trajectory for attack after protonation of the α-TCA compared to a less favorable trajectory for attack of the protonated β-TCA explains the loss of stereoselectivity and increased amount of side products for glycosylation with the β-TCA compared to glycosylation with the α-TCA. Altogether, these findings provide valuable insight into both the mechanism, the scope, and the limitations of self-promoted N -glycosylations. Additionally, the study provides insight into the mechanism of glycosylation in general, and some hints of how to design stereospecific catalysts have been revealed.
|
610975b8393cc9d4bb50d355
| 0 |
In chemistry and materials science, the feedback between computer simulations and lab experiments is often crucial. On the one hand, computer simulations are frequently needed to understand, interpret or support some experimental findings. On the other hand, lab experiments are the "ground truth" to validate computational methods. As we will show, although the physics is the same, there are numerous practical barriers for a seamless exchange between simulations and experiments that make a tight integration more difficult than perhaps necessary. This is notable as computer simulations have evolved to the extent that "routine" simulations can provide feedback which might be most valuable if the feedback loop between experiment and simulation is as short as possible.
|
610975b8393cc9d4bb50d355
| 1 |
It is hard to make "routine" calculations straightforwardly available in a mixed experimental-simulation environment. Running simulations requires the management of a sometimes highly complex software environment in a heterogeneous hardware landscape, e.g., when test runs are executed on a local laptop or workstation and production runs are performed on a high-performance computing (HPC) cluster with completely different architecture.
|
610975b8393cc9d4bb50d355
| 2 |
Furthermore, one needs to transfer input data to the appropriate compute system and prepare them in a way that it can be processed by the employed simulation and analysis tools. Further, to perform correct simulations, one needs to define the right settings and numerical input parameters for the simulations, with at least some minimal understanding of the limitations and assumptions of the employed physical models and algorithms. Manually designing and successfully implementing such a pipeline is not only error-prone, but requires the consultation of a computational expert. If the DFT optimisation is run manually by an experienced computational researcher, the direct link to the experiment will be lost. This is a further disadvantage in addition to the extra time the computational researcher needs to spend on a "routine" simulation.
|
610975b8393cc9d4bb50d355
| 3 |
In this work, we aim to remove the barrier that is imposed by the need to "provide the right input" and make it possible to carry out "routine" calculations in an integrated experimental and computational setting. In this vision, carrying out a simulation should be comparable to sending out an experimental sample to a central facility for routine characterisation. One specifies which calculations need to be done, but the transfer, pre-processing and creation of input data is fully automated.
|
610975b8393cc9d4bb50d355
| 4 |
In the following, we assume that the experimental group uses an electronic lab notebook (ELN) to collect and store all experimental data on a particular sample. We demonstrate how such an ELN can be connected to a simulation platform that gives easy access to routine but state-of-the-art simulations. Such a connection ensures that experimental results can be compared to simulations on a routine basis. We have chosen the cheminfo ELN as the ELN and AiiDAlab 6 as the simulation platform. Both platforms are open-source and allow storing data provenance, meaning that we have a complete record of how simulations and experiments have been performed and influenced each other. The cheminfo ELN makes sure that all required data are sent to AiiDAlab along with the request for a simulation. Once the simulations are done, the results are automatically linked to the sample and interactively visualised in a web browser. The results contain metadata and the unique identifiers of the calculations, enabling users to trace back at a later time the full provenance of their simulations performed by AiiDAlab.
|
610975b8393cc9d4bb50d355
| 5 |
The cheminfo ELN ensures that all experimental data are automatically converted into a FAIR-compliant format (typically, JCAMP-DX ). The metadata containing the type of experiments and sample are added in a standardised format (cheminfo.github.io/data_schema/). Due to this standardisation and the availability of a representational state transfer (REST) application programming interface (API), the cheminfo ELN is ideally suited for integration with other services.
|
610975b8393cc9d4bb50d355
| 6 |
Typically, an experimental structure is initially uploaded to the ELN, e.g., as the crystal structure derived from a Rietveld refinement or the molecular structure as the educt or product of synthesis (see Figure ). In the ELN, users have now access to a button that we developed as part of the current integration work, that initiates the transfer of the structure to their personal AiiDAlab instance using a URL redirect (Figure ). This request will contain the sample's universally unique identifier (UUID), the database uniform resource identifier (URI), the username, and the structure name as URL query parameters (see Supplementary Figure ). With this information, AiiDAlab can then employ the REST API of the cheminfo ELN to request the structure file (Figure ). This structure file can subsequently be used as an input for simulations or other calculations. Once these are completed, the results can be submitted back to the ELN, with reference to the initial structure from which they were launched.
|
610975b8393cc9d4bb50d355
| 7 |
A key requirement for our implementation is that the integration is both seamless and secure. For this reason, we use the API token mechanism implemented by the rest-on-couch package . For this, AiiDAlab first checks if it already owns a suitable token to connect to the ELN and otherwise prompts the user to set up the connection, if that is not the case. To this aim, the user is asked to provide the ELN instance's address and type (cheminfo in our case) and the access token, which can be obtained via the "Request token" button within the AiiDAlab interface. A click on the button will open a page from the cheminfo ELN in an iframe. Before printing the token, the ELN first checks if the user is already authenticated and, if not, it redirects to the authentication page. After the authentication, the user is redirected back to the token-generation page, where the token is shown. This mechanism ensures that the browser contexts remain separated, but we can still use the existing session cookie stored in the browser to authenticate the session in the ELN. The separation of browser contexts also implies that users need to copy and paste the token from the iframe into an input field on the same page. At the same time, the reuse of the session cookie implies that users do not have to log in again typing username and password. The token will then be stored in the AiiDAlab file system of the user. We consider the security risk of that as acceptable, since the access to the AiiDAlab file system is already protected by the authentication mechanism of AiiDAlab.
|
610975b8393cc9d4bb50d355
| 8 |
As a note, we also considered passing tokens via URL parameters, but discarded this idea. The URL parameters are secure sockets layer (SSL) encrypted for the transport but might be logged in plain text on the servers and the browser history. We could also not simply provide the token with a RESTful POST request from the ELN to the AiiDAlab, as the latter currently implements no REST API, but efforts are ongoing to enable it. We emphasize that, to disable the integration, users can simply delete the token using a custom frontend page provided in the ELN, or delete the token in the ELN setup page of AiiDAlab.
|
610975b8393cc9d4bb50d355
| 9 |
Figure . The ELN frontend with the "submit to AiiDAlab button". The cheminfo ELN, with which we implemented this prototype, has different views for different datatypes. In the "crystal structure" view our integration now displays a "submit to AiiDAlab button" (highlighted in the top right) that sends the UUID of the structure, as well as the database URI and user identifiers (encoded as URL parameters) to a specific AiiDAlab URL (the specific instance can be selected with a dropdown menu). The URL points to a custom Jupyter notebook that we developed, that contains code to automatically extract the URL parameters generated by the ELN, initialise the subsequent steps and guide the user through them.
|
610975b8393cc9d4bb50d355
| 10 |
In the ELN, most interfaces are implemented using the visualizer library , which comes with powerful plotting tools and type renderers. Computed results, if they are added to the database in a compatible form, will appear in the same way as experimental results -just as if they were created by another analytical instrument. For example, simulated adsorption isotherms can seamlessly be overlapped with experimental ones and processed using the same tools.
|
610975b8393cc9d4bb50d355
| 11 |
After clicking the "submit to AiiDAlab button" in the ELN, users will land on a page of the AiiDAlab (see Figure ) that allows opening the imported structure in an AiiDAlab application. Among others, we provide links to the applications that offer automated simulations via the CP2K and QUANTUM ESPRESSO packages that are popular density-functional theory (DFT) codes. Those are used for geometry and cell optimisation, and for the calculation of electronic properties such as band structures. Additionally, AiiDAlab offers applications for force-field based simulations of gas adsorption isotherms and the calculation of geometric properties of nanoporous materials using RASPA and Zeo++ , respectively.
|
610975b8393cc9d4bb50d355
| 12 |
Structures often contain atomic overlaps, disorder, floating solvents, or missing hydrogens that could not be resolved in the experiments. A structure file with these characteristics is usually sufficient in an experimental context, but is typically not suitable to serve as direct input for simulations. Since experimental researchers are typically not aware Tabs organise different applications (Geometry/cell optimisation and electronic properties, pore geometry analysis, isotherm calculation) that contain links to different tools (e.g., QUANTUM ESPRESSO and CP2K as DFT engines) using for any given applications. of such additional constraints, we developed a tool that flags the most commonly encountered issues, focusing on the particular case of metal-organic frameworks, and allows users to directly fix them with automated procedures.
|
610975b8393cc9d4bb50d355
| 13 |
Most of these checks and fixes operate on the structure graph, which we derive using bond heuristics implemented in pymatgen . Based on the structure graph, unbound solvent can then be identified as the connected components that do not cross the periodic boundaries. Over or undercoordination can be identified with hard-coded heuristics for common coordination environments. Based on the latter, we can also return coordinates at which hydrogen atoms would be expected. In the frontend, this is implemented at two different levels (see Figure ). First, a dedicated tab in the structure editor allows to automatically add missing hydrogens and select errors to be fixed (Figure ). Second, a final check indicates potential issues that should still be fixed (Figure ).
|
610975b8393cc9d4bb50d355
| 14 |
To capture the full provenance of the interaction between the computational and lab experiments, we store additional parameters in both the AiiDA and the ELN database. The AiiDA database is used by the AiiDA workflow engine to orchestrate the workflows, to which AiiDAlab provides the frontend. To this aim, we extended the ELN database schema to include the source type (simulation, experiment, literature), the database URI, and the UUID of the data node from the AiiDA database. Correspondingly, in the AiiDA database, we store the ELN database URI, the type of the ELN, the sample UUID, the attachment filename, and the data type. This information is stored in the so-called "extras" of a node representing the structure in the AiiDA database (of type CifData or StructureData).
|
610975b8393cc9d4bb50d355
| 15 |
In our opinion, the key difficulty for the integration of different services is that there are still no generally accepted standard schemas for the storage and transfer of scientific data. Most ELNs do not impose a data schema or offer a a b Figure . Automated checks help to ensure that structures are ready for simulations. a Many issues can be directly fixed from the structure editor. For example, hydrogen atoms can be automatically added for common coordination geometries. b Coloured checkmarks and crosses indicate the checks performed and potential issues. In this particular case, the checks detected a floating atom in the centre of the pore (which can then automatically be removed).
|
610975b8393cc9d4bb50d355
| 16 |
standardised API to access specific characterisation data. Furthermore, those that do define a schema currently use different schemas that are not directly interoperable. This means that, in practice, one needs to manually create mappings between the data schemas of the different platforms that need to interoperate (e.g., here, the AiiDAlab platform and the ELNs). To simplify this process with the cheminfo ELN, we have been developing a Python package (cheminfopy) that provides a Python interface for the same abstractions (sample, reaction, and attachments) that are used in the database schema of the cheminfo ELN (see Supplementary Note 3). Using this interface, attributes of samples (e.g., boiling point, molecular mass) in the ELN can be accessed as simple Python object attributes. Since AiiDAlab already implements parsers for the simulation outputs, the only missing step we implemented was to map the AiiDA data objects representing simulation outputs to JCAMP-DX files, the preferred format for spectra in the cheminfo ELN. The conversion has been implemented in the aiidalab-eln package.
|
610975b8393cc9d4bb50d355
| 17 |
As a concrete example of our integration protocol, we compare the powder X-ray diffractogram (PXRD) of a synthesised metal-organic framework (MOF) with the one obtained from a DFT-optimised structure. Specialised workflows have been developed to optimise the geometry of a MOF using the crystal structure as the starting point 2 . These workflows involve a combination of different optimisation strategies and typically take several hours to converge. Once we have such a DFT-optimised structure, there are several calculations we can do. For example, we can predict the PXRD pattern (see Supplementary Note 5). Large deviations with respect to the experimental PXRD are often a strong indication that the proposed structural model is not stable and might not be the correct one. Importantly, this also clearly indicates that this structural model is not relevant for subsequent simulations of other Figure . Overlay of an experimental and a simulated isotherm. In the ELN, the simulated isotherm can be visualized and processed in the same way as an experimental one. The only difference between the simulated the experimental isotherm being the metadata that, in the case of the simulated isotherm, contains the UUID of the output in the AiiDAlab database. Adsorption isotherms that vastly differ in saturation loading can indicate an unsuccessful activation of the material. spectra. A geometry optimisation might also help in the refinement process, for instance, by providing information about the location of the hydrogen atoms. Once we have such a DFT optimised crystal structure, AiiDAlab can be used to predict many experimental properties. For example, an important application of MOFs is gas storage and separation . A key observable for the performance in these applications is the gas adsorption isotherm. One also needs to consider the partial charges in the gas adsorption simulations for polar molecules . Since for "routine" simulations one does not want to worry about how charges are computed, the AiiDAlab application automatically derives the partial charges after the geometry optimisation if an isotherm simulation is requested by the user. After the simulation, one can overlay the simulated isotherm with the experimental data in the ELN, as shown in Figure . Deviations between the experiment are often due to the inclusion of guest molecules or reduced crystallinity. For this reason, it is important to compare the experimental results with the simulations. Unfortunately, this comparison does not happen routinely at the moment, even though the gas adsorption simulation can be considered as "routine" simulations. We anticipate that the direct integration of atomistic and molecular simulations into an ELN will make such simulations routine.
|
610975b8393cc9d4bb50d355
| 18 |
To improve the interoperability of the cheminfo ELN with other services and tools we are currently migrating the data schema of the cheminfo ELN to JSON-schema (via TypeScript types, github.com/cheminfo/ cheminfo-types), which would allow us to programmatically validate requests to the ELN. In addition, we are also implementing a REST API for AiiDAlab which would also enable ELNs, and other services, to send data to and request data from the AiiDAlab. Clearly, our work highlights the need for a collective effort from the community of ELN developers for the development of a standardised API specification, comparable to OPTIMADE , for standardised access to data across simulation and experimental databases.
|
64ad8d8e6e1c4c986b368d4f
| 0 |
Pyrrovobasine ( ) is an indole alkaloid with a pseudo-dimer structure isolated from Voacanga africana stem bark extracts, mainly in tropical Africa, by Beniddir and co-workers in 2022 (Figure ). It is structurally characterized by a pyrraline structure on the complex linkage of tryptophan and tryptamine, which is unusual for a natural product. The total synthesis of pyrrovobasine (1) has not yet been reported. In this work, we report the first total synthesis of pyrrovobasine (1). This method features an inexpensive and efficient strategy towards pyrrovobasine (1) on a gram scale in a short process. It can be widely applied to synthesizing other related natural products and analogs.
|
64ad8d8e6e1c4c986b368d4f
| 1 |
The synthetic strategy of pyrrovobasine ( ) is shown in Figure . Since pyrrovobasine (1) has a pyrraline structure, which is relatively rare as a natural product, the key to this synthetic route is how to introduce this pyrraline structure. Considering the instability of the pyrraline skeleton, we planned to introduce a pyrraline derivative into a tryptamine derivative at the end of the total synthesis. There are several reports on the synthesis of alkyl pyrralines. 2 Alkylation of pyrrole derivatives is the simplest and most reliable reaction, although strong basic conditions are generally used. On the other hand, it has been reported that the epimerization of the methyl ester on the Vobasine skeleton proceeds readily under basic conditions. Based on the above, we planned to synthesize the pyrraline skeleton using a pyranone derivative, which proceeds under weakly acidic conditions in the late stage of the synthesis. The regioselectivity and stereoselectivity of the tryptamine derivative in the Vobasine skeleton must be controlled, and we speculated that the regioselectivity of the tryptamine derivative would be C2-selective based on its electron density. Interestingly, the Me-group of methyl ester on the Vobasine skeleton shows a signal around 2.5 ppm in 1 H NMR. 1 This peak is shifted to a higher field than that of the normal methyl ester. This is presumably due to the shielding effect of the indole moiety of the Vobasine skeleton, which covers the methyl ester moiety, and thus shifts it to a higher magnetic field. Therefore, we speculate that the methyl ester on the Vobasine backbone acts as a steric hindrance and that the introduction of tryptamine derivatives proceeds by avoiding the methyl ester, allowing the stereoselective introduction of tryptamine derivatives.
|
64ad8d8e6e1c4c986b368d4f
| 2 |
The plausible biosynthetic pathway of pyrrovobasine (1) is shown in Figure . After removal of the alcohol moiety of vobasinol (2), the C2 position of the tryptamine derivative 4 with the pyrraline structure is linked to the activated vobasinol derivative 3 to form pyrrovobasine (1). It is known that sugar derivatives such as hexose react with amine residues of amino acids and proteins by Maillard-type reactions. Tryptamine derivative 4 is thought to be formed by the enzymatic decarboxylation of L-tryptophan followed by the reaction of the amino group with hexose in a Maillard-type reaction. Based on the biosynthetic pathway described above, we planned to introduce the tryptamine and pyrraline skeletons at the end of the synthesis.
|
64ad8d8e6e1c4c986b368d4f
| 3 |
The synthetic route of pyrrovobasine ( ) is shown in Scheme 1. The key to this synthesis is the short and scalable synthesis of the vobasinol skeleton 2, a key intermediate in the biosynthetic pathway. Therefore, the use of an inexpensive and scalable reaction that can be performed on a decagram scale is critical for the large-scale synthesis of pyrrovobasine (1). Therefore, we decided to perform the C-C bond formation reaction mainly in the early stage of the synthesis. In addition, we avoided using expensive transition metals such as Pd in the C-C bond formation reaction as much as possible, especially in the early stage of the large-scale synthesis. First, D-tryptophan was used as the starting material and N-propargylated by reaction with propargyl bromide 5. The resulting compound was then reacted with α-ketoglutaric acid 6, affording the cyclized compound 7 in 49% (2 steps) (trans/cis = 2:1). Compound 7 was successfully used for the synthesis of tetracyclic compound 8 in 88% yield (40 g) via Dieckmann condensation, which proved to be very useful in the synthesis of 1,3-dicarbonyl compounds. It is worth noting that the tetracyclic compound 8 can be synthesized on a 40-gram scale from the starting material D-tryptophan in a 3-step procedure without any protecting group using inexpensive reagents. Subsequent decarboxylation of 8 followed by Boc protection led to the synthesis of 9.
|
64ad8d8e6e1c4c986b368d4f
| 4 |
Our next attempt was the construction of the pentacyclic compound 12. The development of an inexpensive and scalable method for the preparation of vobasinol skeleton 2 is one of the most important aspects of this synthetic route. Several synthetic approaches for vobasinol scaffold 2 have been reported. However, some of them require expensive Pd catalysts, which have dramatically increased in recent years, and are not suitable for the decagram scale, or require activation of the ketone moiety of the starting material, or several derivatization steps after the cyclization reaction. In view of this, we investigated various cyclization reactions of 9. Interestingly, 12 was found to be easily synthesized by the reaction of 9 with Mn(OAc)3 at room temperature in acetic acid (49%, E/Z = 1:1). The reaction mechanism involves the coordination of 9 to Mn(OAc)3 to form the Mn-enolate 10. Subsequently, 10 undergoes enolate oxidation to form 11, which has a radical at the α-position of the ketone. Then 11 is subjected to a radical addition with a nearby alkyne to form 12-Z and 12-E. This concise and short-step strategy has made it possible to prepare pentacyclic compound 12, a common intermediate for many vobasine alkaloids, from inexpensive tryptophan on a decagram scale in only 6 steps. Notably, 12-Z and 12-E are separable and can be derivatized to related natural products. The obtained 12-Z with Boc group was deprotected to achieve a simple formal synthesis of koumidine with antitumor activity (see Supporting Information). 10 On the other hand, 12-E was converted to 13 by the Wittig reaction. The resulting 13 was hydrolyzed to the aldehyde with trifluoroacetic acid, and the thermodynamically stable stereochemistry of the methyl ester 14 was obtained using iodine, KOH, and MeOH, followed by Boc protection.
|
64ad8d8e6e1c4c986b368d4f
| 5 |
The synthetic challenge for the total synthesis of pyrrovobasine (1) is the epimerization of the thermodynamically stable methyl ester stereochemistry in 14 to the kinetic product methyl ester 16. We first tried to epimerize 14 under different basic conditions, but the desired kinetic product methyl ester 16 was not observed under any of the conditions. On the other hand, Zhang and Yang et al. reported the epimerization of the thermodynamically stable methyl ester using Ir-catalyzed photo-epimerization. Therefore, 14 was treated with LDA and iodine to give 15, which was found to be unstable and used immediately for the next reaction without purification. Then 15 was successfully converted to 16 by UV irradiation at 365 nm in the presence of an Ir-catalyst at -50 ºC followed by deprotection of the Boc group. Although this epimerization procedure works well, the resulting kinetically dominated product 16 was difficult to separate from the thermodynamically dominated product methyl ester, which was also generated in situ during the photo-epimerization reaction from 15. Therefore, the crude product 16 was used in the next reaction without further purification.
|
64ad8d8e6e1c4c986b368d4f
| 6 |
The introduction of the tryptamine derivative 4 with the pyrraline structure to 16 is the last task necessary to complete the total synthesis of 1. Because of the possible instability of the pyrraline scaffold, we planned to introduce the pyrraline moiety step by step after the introduction of protected tryptamine 19 to 18. The bioinspired activation of the tertiary amine moiety on 16 was inspired by the pioneering work of Han et al. In prior studies, it was demonstrated that the activation of tertiary amine moiety followed by C-N bond cleavage could take place using (bromodifluoromethyl)trimethylsilane. When the condition using (bromodifluoromethyl)trimethylsilane was applied to 16, 17 was successfully detected in the reaction mixture by LC-MS. It is also worth noting that the BF4 salt 17 was a stable solid and could be easily isolated by Celite ® filtration followed by concentration. This stable BF4 salt 17 is a useful intermediate that could be used for the synthesis of other related natural products. The reaction mixture of 17 was then acidified with HCl/MeOH to cleave the C-N bond and protected tryptamine 19 was added to react with intermediate 18 to give 20. The resulting 20 was successfully subjected to N-methylation by reductive amination to give 21 in 17% (5 steps). Finally, 21 was deprotected with Zn to afford a primary amine and then reacted with pyranone derivative 22 2 in AcOH at 50ºC to complete the total synthesis of 1.06 g of pyrrovobasine (1).
|
64ad8d8e6e1c4c986b368d4f
| 7 |
The synthesis described here was enabled by user-friendly reactions and strategies that are inexpensive and scalable. The early stages of this synthetic route focused primarily on the short steps, decagram-scale synthesis of the key intermediate, the pentacyclic compound 12. In the process, a 3-step synthesis of the tetracyclic compound 8 without protecting groups was achieved. In addition, a direct radical coupling of ketone 9 with Mn(OAc)3 was discovered and a decagram-scale synthesis of pentacyclic compound 12, a key intermediate for many vobasine alkaloids, was achieved in only 6 steps. Finally, we developed an efficient bioinspired method for introducing tryptamine, which has a pyrraline structure that is relatively rare in natural products and achieved the first total synthesis and gram-scale supply of pyrrovobasine (1). This methodology is anticipated to be useful for future investigations into the biological activity and structureactivity relationships of pyrrovobasine (1).
|
665de4dc418a5379b0ff1734
| 0 |
Human immunodeficiency virus type 1 (HIV-1) is responsible for acquired immunodeficiency syndrome (AIDS). During the progression of this disease, the vital cells that are effective in the immune system -helper T cells (CD4+ T cells, T cells that contain CD4), macrophages, and dendritic cells (which present antigenic material to the cell surface of T cells)are infected by HIV. Consequently, the number of CD4+ T cells decreases, leading to cell death in the host. When the number of CD4+ T cells reaches a critical point, cell-mediated immunity is lost and the body becomes sensitive to infections, resulting in AIDS .
|
665de4dc418a5379b0ff1734
| 1 |
Synthesized in the early phase of the viral infection, Tat is secreted from T-infected cells circulating in the bloodstream . It is then taken up into uninfected cells where Tat localizes in the nucleus and binds to the long terminal repeat of HIV (LTR) TAR on the viral RNA hairpin. In this way, Tat acts as a regulatory protein of viral gene expression, enhancing the transcription of all HIV-1 genes . This is commonly referred to as the primary role of the Tat protein .
|
665de4dc418a5379b0ff1734
| 2 |
Additionally, numerous pathological conditions are also associated with the Tat protein. For example, it inhibits phagocytosis in macrophages by binding to phosphoinositol-4,5-biphosphate (PI4,5P2) on cell membranes and alters the activity of potassium channels in cardiomyocytes . Some others involve triggering neurodegeneration , disrupting the blood-brain barrier , and inducing apoptosis in T cells . In summary, this protein performs many tasks in the extracellular and intracellular milieu and with that, it rightly has earned the reputation of being the main viral toxin of the HIV-1 virus .
|
665de4dc418a5379b0ff1734
| 3 |
How can a single protein be involved in this many seemingly unrelated biological functions? This stems from the fact that Tat is an intrinsically disordered protein with no welldefined secondary structures . As a result of this structural plasticity that allows it to adopt many conformations in solution, Tat can interact with a large repertoire of proteins and receptors in infected and non-infected cells . Amongst them, a prominent one is p53 which is known as a cellular tumor suppressor . In a cell, p53 is a homotetrameric transcription factor that positively and negatively regulates the expression of diverse genes, whereby it governs diverse downstream events, including glycolysis , autophagy , the regulation of oxidative stress , angiogenesis and so forth. Furthermore, this regulatory protein serves a function in the cellular response to numerous cellular stresses and ultimately coordinates the cell cycle and apoptosis. Taken together, p53 is decisive in the survival and death of abnormal cells .
|
665de4dc418a5379b0ff1734
| 4 |
In the context of AIDS, the interaction of Tat/p53 is perplexing because their roles appear to be somehow contradictory. As such, p53 functions in a way to induce apoptosis whereas Tat's goal is to assure the survival of HIV. In this respect, it was speculated that the interaction between these two proteins indeed involves a switching mechanism. In the early phase of the disease, p53 binds to Tat and presumably inhibits it . As Tat subsequently accumulates in the cellular milieu, it suppresses the expression of p53 without having any direct contact with it. In summary, one can suggest that these proteins indeed crosstalk, which may result in the transformation of cells by HIV-1 or activation of HIV-1 replication . Despite its significance, the studies aiming at elucidating how these two proteins interact are currently limited. In an early work, the interaction between these two proteins was confirmed through the yeast two-hybrid systems, implying a specific binding of these two proteins . A more recent and conspicuous study, aiming to clarify the molecular foundation of these interactions, revealed that p53 in fact binds to Tat protein through its tetramerization domain.
|
665de4dc418a5379b0ff1734
| 5 |
Using peptide mapping technique and fluorescence anisotropy, the same study unveiled that the tetramerization domain in p53 specifically binds to residues 1-35 and 47-57 within the Tat protein . The results reported hereby are certainly very significant; however, a detailed investigation of the atomistic interactions formed between these proteins is fundamental for more profound interpretation of experimental studies, considering the biological implications of these interactions.
|
665de4dc418a5379b0ff1734
| 6 |
In this study, Tat (PDB code: 1K5K, model 1) and p53 (PDB code: 2J0Z) were docked through four different web servers: ZDOCK , FRODOCK , HawkDock and HPEPDOCK . By visual inspection with VMD, ten top ranked complex structures from each server were compared with those from other servers and consensus binding poses were identified . As a result of this analysis, four different models for the Tat-p53 complex were determined.
|
665de4dc418a5379b0ff1734
| 7 |
The MD simulation files were prepared through Charmm GUI and subsequently, the simulations were run for 1000 ns using NAMD (version 2.14) . For each Tat-p53 model structure three replica simulations were carried out. Electrostatic interactions were calculated using the particle-mesh Ewald method with a grid spacing of 1 Å . Simulations were performed with a time step of 2 fs, with all interactions calculated at every time step. The temperature and pressure were kept at 303.15K and 1 atm, respectively. Atomic coordinates were recorded at every 20 ps.
|
665de4dc418a5379b0ff1734
| 8 |
The stability of the complexes was investigated with RMSD analysis using the VMD program. In this analysis, simulation trajectory is aligned on p53 initial structure, and then RMSD calculation was performed for the Tat protein to see the stability of the Tat binding. A python library PyContact is used to extract information about the atomistic interactions formed between p53 and Tat . Hydrogen bond analysis between p53 and Tat was performed using VMD program with default parameters (3Å for Donor-Acceptor distance and 20 for Angle Cutoff degree).
|
665de4dc418a5379b0ff1734
| 9 |
In investigating the interaction between Tat and the tetramerization domain of p53, our approach initiates with the docking of the two proteins utilizing three docking servers (refer to the experimental section for comprehensive details). Subsequently, the best ten models generated by each server are scrutinized in comparison to those produced by the other servers, according to the position of Tat relative to that of p53. Following to this comparative analysis, models exhibiting similarities are filtered, thereby reducing the total count of models to four (Figure ). This is achieved through the visual examination of Tat/p53 complexes with VMD. The names of these four distinct models in this manuscript and the servers they originate from are delineated in Table .
|
665de4dc418a5379b0ff1734
| 10 |
Next, we performed three 1 µs long three replica simulations to investigate the stability of the predicted complex structures. In some of the simulations, Tat remained stable in its initial position, whereas in some other simulations the position of Tat has changed considerably. Through MM/PBSA analysis, it is observed that the binding energies of 10 out of the 12 models fell within the range of approximately -1 kcal/mol to -15 kcal/mol. However, there are two exceptions, namely replica 1 and 2 of model_ZDOCK (refer to Table ). It should be noted that the binding energy of replica 2 is positive, indicating a thermodynamically unfavorable binding pose. Conversely, that of replica 1 is found to be -46.6 kcal/mol, suggesting this to be the most favorable one among all the considered replicas. Figure presents snapshots for the conformation of p53 and Tat at the beginning and at the end of the simulations.
|
665de4dc418a5379b0ff1734
| 11 |
With these replicas in hand, we obtained important insights regarding the interactions between the tetramerization domain of p53 and Tat, with particular emphasis on the binding regions in both proteins. In the research conducted by Gabizon and colleagues, two specific domains within the Tat protein were suggested as directly involved in complexation: Tat(47-57) also known as the basic patch and Tat(1-31) . These regions exhibit contrasting amino acid compositions. As such, Tat(47-57), characterized by a high abundance of lysine and arginine residues, exhibits a net charge of +8, whereas Tat(1-31), primarily composed of cysteine residues and a hydrophobic core, carries a total charge of only +2. For the tetramerization domain of p53, residues 326-355 have been identified as the specific domain directly engaging with Tat. In their investigation, Gabizon and colleagues noted that the initial interaction between these two proteins entails long-range electrostatic interactions, thereby corroborating the involvement of highly charged Tat(47-57) domain in the binding process. Later, this initial interaction is fine-tuned by some additional ones with lower ionic character, implying that Tat(1-31) becomes the main region in Tat to interact with p53 (for further details, the reader should refer to the original paper). As for the tetramerization domain of p53, the researchers reported that the binding region in p53 spans residues Glu326 to Arg342.
|
665de4dc418a5379b0ff1734
| 12 |
To determine which domains of Tat and p53 interact, we analyzed the last 200 ns of model_ZDOCK's replica 1, which exhibits the highest binding energy. This analysis was performed using the PyContact tool. In this replica, we see that both N-domain and the basic of Tat bind to p53, as suggested by Gabizon. Some interactions are: (i) hydrogen bonds between Tyr32-Thr329, Tyr32-Gln331, Cys34-Thr329, (ii) salt-bridges between Arg49-Glu343, Arg52-Glu339, Lys50-Glu343 (See Table ). Additionally, it is noteworthy that the C-terminal domain of Tat contributes minimally to its binding with p53. Specifically, only one residue, Glu87, forms a salt bridge with Lys351. Collectively, our results indicate that the interaction between the two proteins primarily involves the basic patch (residues 47-57) and the N-terminal region of Tat (residues 1-40). This partially aligns with a previous observation suggesting that Tat (residues 73-86) may also be involved in complex formation with p53 . what dictates the binding energy was the number of hydrogen bonds between two proteins, in that more hydrogen bonds would translate into higher binding energy between proteins. However, hydrogen bond analyses revealed that the number of hydrogen bonds formed between proteins in each replica is consistently close, indicating that it cannot be the sole factor governing the binding energy (Figures ). For this reason, further detailed structural analysis is certainly needed to substantiate any conclusive finding. The last finding pertains to the binding conformation of p53. To this respect, Gabizon and colleagues elucidate that the negatively charged residues Glu343, Glu346, and Glu349 within the p53(326-355) region mediates the initial contact between the two proteins. This is unsurprising given the high charge content of the Tat(47-57) region. Model_ZDOCK's replica 1 also confirms hydrogen bonds and salt bridges between p53(Glu343) and Tat(Arg49) residues in the last 200 ns. However, Glu346 and Glu349 do not appear to play a role during the complexation between the two proteins according to our model.
|
665de4dc418a5379b0ff1734
| 13 |
In terms of how p53 functions and how it interacts with other proteins, the tetramerization domain of the protein is central as it strictly controls the delicate balance between the active and the latent DNA-binding conformations of p53. Thus, this domain ensures that p53 adopts the appropriate conformation for the overall activity within the cell .
|
665de4dc418a5379b0ff1734
| 14 |
Considering how the tetramerization domain modulates protein-protein interactions, it is no surprise that this domain is targeted by HIV-Tat protein. Of course, the full biological function of p53-Tat interaction is not fully understood yet; however, it is implied that this may enhance Tat potency in transcribing its target viral genes. This notion is bolstered by the observation that both p53 and Tat proteins are, in fact, nuclear proteins acting as transcription factors . Considering the importance of Tat for the pathogenesis, investigating the interaction between Tat and p53 will certainly contribute to better understanding of the biochemistry of Tat and the deleterious effects of HIV disease within cells. Driven by this motivation, we investigated the complexation of p53 and Tat proteins, using molecular docking and MD simulations. In the MD simulations of Tat-p53 complexes, we observed that our most favorable model aligns with the experimental findings previously reported by Gabizon et al., highlighting the significance of non-ionic interactions observed during the complexation of these two proteins. Of greater significance, it appears that the interactions primarily occur via Tat the basic patch and N-terminal domains. Overall, these finding reported herein may lead to the discovery of anti-HIV agents with previously unprecedented activity.
|
6527281545aaa5fdbbcd7258
| 0 |
In the search for next-generation optoelectronic devices, there has been a growing interest in donor-acceptor (D-A) materials, including D-A co-crystals and D-A copolymers, for their application in organic solar cells and organic lightemitting diodes. D-A complexes, by definition, exhibit charge transfer (CT) in their ground and select excited states. In the search for structure-function relationships to integrate into high-throughput screening and machine learning protocols, several studies have assessed ways to predict the degree of CT (DCT), or ionicity parameter, in the S0 state of D-A materials from molecular quantities such as orbital energies, vibrational frequencies, and geometric parameters. Early evidence shows a relationship between DCT in S0 to the magnitude of effective CT integrals, commonly used in models of charge transport. The DCT in the first electronically excited state of D-A dimers, S1, has emerged as a key quantity for predicting radiative and non-radiative lifetimes in D-A materials, including intersystem crossing rates and fluorescence lifetimes. These lifetimes are particularly difficult to compute directly using D-A dimer models, as energy levels and transition dipoles ofter differ substantially between the molecular cluster and material. A variety of DCT metrics for excited-state calculations have been put forward, as reviewed recently. We construct a similarity metric, η, as follows. First, define a molecular orbital
|
6527281545aaa5fdbbcd7258
| 1 |
where Sµν = ⟨χµ | χν ⟩ is the overlap matrix involving displaced basis functions. We compute Oij twice: once between the HOMO of the isolated donor molecule (HOMO i ) and the HOMO of the donor molecule within the complex (HOMO c ), using ghost functions to ensure that both calculations have the same basis functions; and second, between the LUMO of the isolated acceptor molecule (LUMO i ) and the LUMO of the acceptor molecule within the complex (LUMO c ). We define η as the average of these two quantities:
|
6527281545aaa5fdbbcd7258
| 2 |
The two geometries are maximally oriented using the Kabsch algorithm, in order to maximize the overlap. Each computation of η requires three ground state calculations to obtain the orbitals of the isolated donor molecule, the orbitals of the isolated acceptor molecule, and the orbitals of the D-A complex. The calculation of Oij is performed in a locally-modified version of Q-Chem. To assess the correlation between η and the S1 DCT of D-A dimers, we screened 31 D-A complexes with donor and acceptor molecules shown in Figure and whose S1 states are dominated by a HOMO → LUMO transition. The donor and acceptor molecules chosen are augmented from a data set recently chosen in a screening of S0 DCT. The donor molecules exhibit a diversity of molecular structures, while the acceptor molecules are 7,7,8,8-tetracyanoquinodimethane (TCNQ) and its fluorinated derivatives, F x TCNQ. Geometries are optimized with Gaussian G16 at the CAM-B3LYP/6-31+G(d,p) level of theory with Grimme dispersion. We calculated the DCT using natural bond orbital (NBO) population analysis implemented in Gaussian G16 and transition density matrix (TDM) analysis in Theodore. We investigate the impact of geometries used to compute η by plotting η vs. S1 DCT in two ways. First, we optimize both the isolated monomers and their dimer complex and compute η vs. S1 DCT (Figure , left). Second, we optimize the dimer complex and take the geometries of the isolated monomers to be the same as in their dimer complex (Figure , right). In each case, the S1 DCT is computed using TDM analysis. Notably, we observe significant variations in the η values when the monomers were optimized, as evidenced by the outliers shown in the left plot of Figure . Additionally, the R 2 value from linear regression is 0.66 when comparing optimized monomers vs. optimized dimers, while the R 2 value is 0.96 when monomer geometries are unrelaxed from those found in the dimer complex. In both cases, our analysis reveals a positive linear correlation between the S1 DCT and η, indicating that S1 DCT is large when the HOMO (LUMO) orbital of the donor (acceptor) retains its character from the isolated molecule. That the unrelaxed monomer geometries provide superior performance presents certain practical advantages, including avoiding the computational cost associated with optimizing the monomer geometries. Moreover, computing η at a single geometry eliminates the necessity of evaluating the atomic overlap integrals at displaced geometries. This simplifies the calculation of η, as Sµν becomes the atomic basis self-overlap matrix, which is commonly printed in the output of electronic structure programs. 31 To assess the sensitivity of η to the dimer geometry, bridging the gap between D-A dimers and D-A co-crystals, we perform an analysis of the S1 DCT vs. η for geometries from experimental crystallographic data (where available) and compare to the results found in Figure . S1 DCT is again computed using TDM analysis. Figure shows the results obtained using the geometries from experiment in comparison to the dimer geometries, and indicates that the positive linear correlation still holds when the D-A geometries are taken from experimental crystal structures. In fact, when linear regressions are performed separately, the trend lines are almost indistinguishable. In Table , we report the value of η computed at each geometry (dimer complex vs. experiment) and the percent deviation between the two. The values of η obtained from these two geometries have very small percent deviations, from 0.1% to 6% with an average of 1.3%. This shows that η can bridge between different types of molecular structures (isolated dimer and experimental crystal), indicating reliability of the metric for different data sources. This is challenging for predictive metrics that rely on orbital energies. We next assess two methodologies for calculating the correlation of S1 DCT with η. Using unrelaxed molecular geometries taken from the dimer complexes, Figure plots η vs. S1 DCT using either TDM analysis (as in Figure ), or alternatively using NBO analysis. The linear regression analysis of the TDM-analyzed data (R 2 = 0.96) is significantly improved compared to that of the NBO-analyzed data (R 2 = 0.57). The poor fit in the latter case is due to the method's inability to treat delocalized electron transfer. Conversely, TDM provides a spatial mapping of the electron-hole pair associated with an electronic transition between two states and can successfully treat such delocalized electron transfer. While both methods are commonly used to calculate the charges in molecules, we recommend TDM analysis over NBO.
|
6527281545aaa5fdbbcd7258
| 3 |
In the left panel of Figure , the HOMO for DMeO-BTBT, the LUMO for TCNQ, and the HOMO and LUMO of their dimer complex are visualized; this D-A dimer has the greatest S1 DCT in the data set. The visual similarity between the monomer orbitals and those in the complex is apparent, and the localization onto donor and acceptor moieties in the complex is striking. In the right panel of Figure , the HOMO of DPTTA HOMO, the LUMO of F4TCNQ, and HOMO and LUMO of their dimer complex are visualized; this D-A dimer has the smallest S1 DCT in the data set. While there is significant visual similarity between the monomer MOs and those in the complex, there is also substantial delocalization of the HOMO onto the acceptor molecule and similar delocalization of the LUMO onto the donor. To maximize S1 DCT, the electron density must be localized on the electron donor in the HOMO and transferred completely to the acceptor LUMO. The value of η quantifies the degree to which the isolated HOMO (LUMO) of the donor (acceptor) correlates with that in the complex, which predicts S1 DCT.
|
6527281545aaa5fdbbcd7258
| 4 |
We have shown that the S1 (HOMO→ LUMO) DCT can be predicted by a novel metric, η, that computes the average similarity between a donor (acceptor) molecule's HOMO (LUMO) and that of the corresponding orbital in the D-A complex. We find that η exhibits a positive linear correlation with S1 DCT for a set of 31 D-A pairs. In choosing molecular geometries to compute η, we find that in comparing orbitals between isolated donor and acceptor molecules and their corresponding D-A complexes, one should use the same molecular geometries in the isolated molecules as in the D-A complex. Alternatively, experimental crystal structure data for the D-A complex can be used instead of optimized D-A dimer geometries. This allows flexibility in input data for inclusion in high-throughput screening and machine learning protocols. Lastly, we compare two methods for determining DCT, NBO and TDM analysis, and find that TDM is more reliable due to its ability to treat electron delocalization. Future work will aim to generalize the η metric to characterize the DCT of other electronic states, with the goal of uncovering further orbital structurefunction relationships.
|
61114276abc9e2220469dc9e
| 0 |
Cyclo carbon (C 18 ) is a novel all-carboatomic molecule that has been generated and characterized recently . This allotrope of carbon has attracted considerable attention in the field of chemistry and material science since it has been observed in condensed phase , and a large number of theoretical explorations on its electronic structure , aromaticity , optical and spectroscopic properties , molecular interaction , potential applications in molecular devices , as well as multiple properties of its analogues have been reported in succession. However, there is currently no way to obtain an isolate product of C 18 and its pure-carbon analogues due to their extreme chemical activity.
|
61114276abc9e2220469dc9e
| 1 |
In fact, since Hoffmann put forward the idea of C 18 in 1966 , several strategies have been proposed to try to achieve this unique allotropic carbon, including synthesis by eliminating the heteroatomic substituents of its various cyclocarbon precursors through methods such as retro-Diels-Alder reaction , decomplexation , decarbonylation , and [2+2] cycloreversion . One of the two recent pioneering experiments to prepare C 18 was achieved by sequentially removing a pair of carbonyl groups [(-CO) 2 ] from a cyclocarbon oxide C 18 -(CO) 6 , producing the C 18 -(CO) and C 18 -(CO) 2 molecules in turn, up to the final product C 18 . Actually as early as 30 years ago, Diederich and co-workers have successfully prepared C 18 -(CO) 6 molecule and determined it by X-ray diffraction with the same purpose . However, although cyclocarbon oxides are very special compounds with rich-carbon features, maybe blinded by the pursuit of pure all-carboatomic rings, they have not received the attention they deserve. And yet we believe that, at least till now, studying the easily synthesized and relatively stable cyclocarbon oxides C 18 -(CO) n (n = 6, 4, and 2) should have more practical significance than investigating the very reactive and hardly attainable C 18 .
|
61114276abc9e2220469dc9e
| 2 |
Our previous theoretical studies on all-carboatomic C 18 showed that, due to the special sp-hybrid form of its carbon atoms, the molecule exhibits an unusual in-plane π electron system (π in ) in addition to the out-of-plane π system (π out ) that are common in conjugated molecules . It can be expected that the cyclocarbon oxides should possess the special electronic structure that interests us like C 18 . In addition, it is a meaningful topic to explore how the presence of -CO groups affects the electron delocalization and aromaticity of the cyclocarbon oxides. So, in this work, the bonding character and electron delocalization of three C 18 precursors, C 18 -(CO) n (n = 6, 4, and 2), are analyzed in detail using a variety of wavefunction analysis methods based on reliable quantum chemistry calculations. Their respective aromatic characteristics will be discussed and compared in terms of the number of -CO groups.
|
61114276abc9e2220469dc9e
| 3 |
The anisotropy of the current-induced density (ACID) analysis was realized by the ACID code based on the output file of Gaussian program, and the maps were generated by POV-Ray render . The gauge-including magnetically induced currents (GIMIC) analysis was finished via the GIMIC code based on the formatted check point file of Gaussian program, and the maps and animations were rendered by ParaView visualization program . Other electronic structure and wavefunction analyses were performed with the Multiwfn 3.8(dev) code based on the wavefunctions produced by DFT calculations. The isosurface maps of various orbitals and real space functions were rendered by means of Visual Molecular Dynamics (VMD) software based on the cube files exported from Multiwfn. The colored contour maps of various real space functions were plotted directly via the Multiwfn code.
|
61114276abc9e2220469dc9e
| 4 |
Previous theoretical studies on the molecular structure of C 18 have demonstrated that only the exchange-correlation functionals with more than 25% Hartree-Fock exchange can reliably exhibit the ground-state geometry of the molecule , and the B97XD/def2-TZVP level has been proved to robustly reproduce the high-precision computational result and experimental observation. Hence, we still adopt the same strategy as our previous works on C 18 to optimize the geometry of precursors C 18 -(CO) n (n = 6, 4, and 2).
|
61114276abc9e2220469dc9e
| 5 |
The geometric structures of the precursors C 18 -(CO) n (n = 6, 4, and 2) obtained at the B97XD/def2-TZVP level are shown in Fig. , and the corresponding Cartesian coordinates are listed in Table . The optimized structures of all precursor molecules are found to be strictly planar with no imaginary frequencies, and the point group symmetries of C 18 -(CO) n with n = 6, 4, and 2 are D 3h , C 2v , and C 2v , respectively. Fig. summarizes the calculated structural parameters and available crystal data of C 18 -(CO) 6 molecule. The maximum absolute deviations of bond lengths and bond angles between theoretical calculation and experimental measurement are only 0.03 Å and -2.3º , respectively, which shows that our calculations accurately reproduced the crystal structure of the C 18 -(CO) 6 . This also emphasizes that the exchange-correlation functional of B97XD is a reliable choice for the studying of C 18 and its substitutes.
|
61114276abc9e2220469dc9e
| 6 |
From a structural point of view, the C 18 -(CO) n (n = 6, 4, and 2), like C 18 ring, present a polyynic skeleton with alternating long and short bonds. Based on the geometry and wavefunction generated at the level of B97XD/def2-TZVP, we simulated the scanning tunneling microscope (STM) image of the precursors C 18 -(CO) n (n = 6, 4, and 2) in the gas phase, as displayed in Fig. . It can be seen that the tunneling current is relatively prominent over the short C-C bonds, corresponding to gray or white on the STM image. In contrast, almost no tunneling current is detected on the long C-C bonds, because the local density-of-state (LDOS) of the highest occupied molecular orbitals (HOMOs) show nodal planes in these regions. This character of STM of cyclocarbon oxides is very similar to that of C 18 , exhibiting similarity between their electronic structures . For details and meanings of STM image simulation, please refer to the supplementary material of Ref. .
|
61114276abc9e2220469dc9e
| 7 |
Bond order is an important concept and a quantitative index for characterizing chemical bonds, which essentially reflects the number of electron pairs shared by two bonding atoms. Here, we calculated various bond orders of C 18 -(CO) n (n = 6, 4, and 2) via different methods, and the bond lengths and bond orders of each precursor are collectively plotted in Figs. S3-S5. It is found that there is a good correlation between bond length and various bond orders of each species, that is, the longer the C-C bond length is, the smaller the C-C bond order is, and vice versa. Similarly to the case of C 18 , the bond orders of the long and short bonds in C 18 -(CO) n (n = 6, 4, and 2) calculated by all methods are larger than 1.0 and smaller than 3.0, respectively, suggesting that some of the electrons from the short C-C bonds, which are formally triplet bond, are delocalized to the long C-C bonds. The existence of electron delocalization in C 18 -(CO) n (n = 6, 4, and 2) implies that it is possible for them to generate induced ring current in external magnetic field, which is the topic we will focus on below.
|
61114276abc9e2220469dc9e
| 8 |
In addition to the commonly known π out molecular orbitals (MOs), molecules containing successive sp-hybrid carbons usually have a special conjugated π in orbitals located in the molecular plane . The π out and π in MOs of C 18 -(CO) n (n = 6, 4, and 2) molecules are listed in Figs. S6-S8, which will be involved in the following wavefunction analysis.
|
61114276abc9e2220469dc9e
| 9 |
The interaction region indicator (IRI) is a simple real space function that can clearly reveal both chemical bonds and weak interactions between atoms or molecular fragments . The isosurfaces and color-filled maps of IRI function for the π MOs, referred to as IRI-, are depicted in Fig. . It can be seen that the IRI-π isosurface of the C-C bonds in the ring except for those adjacent to the -CO groups is annular, indicating that these C-C bonds have dual π (i.e. π out and π in ) interaction features. However, the strength of π interaction of two types of C-C bonds is obviously different. The annular IRI-π isosurface of the short C-C bonds is broader and bluer than that of the long C-C bonds, providing intuitive evidence for the larger electron density in the π interaction region of the short ones, so the π interaction of the short bonds is stronger. In contrast, since the corresponding IRI-π isosurface is obviously not annular, only out interactions exist on the C-C bonds adjacent to the -CO groups.
|
61114276abc9e2220469dc9e
| 10 |
The MOs of C 18 -(CO) n can be considered as the result of mixing the MOs of the -CO and C 18 -fragments. The analysis of the interaction between the fragment MOs in As shown in the previous IRI- analysis and the subsequent electron delocalization analyses, after the combination of (CO) 2 and C 18 , the in electron conjugation is destroyed around the junction region between the two fragments. Two underlying reasons are: (1) the introduction of the -CO groups causes obvious distortion on the C 18 structure, especially makes severe in-plane bending of the C-C-C angles in the carbon ring near the -CO groups, which leads to evident difficulty in in electron conjugation over this region; (2) as can be seen in Fig. , the MOs of -(CO) 2 and the in orbitals of C 18 -produce a significant mixing, resulting in new occupied MOs of C 18 -(CO) 2 clearly display the character of bonds between the -(CO) 2 and C 18 -moieties. Since electrons are well known to be strongly localized, the originally globally delocalized in electrons in C 18 can no longer delocalize over the bonds where carbon ring binds to the two -CO groups.
|
61114276abc9e2220469dc9e
| 11 |
The localized orbital locator (LOL) and electron localization function (ELF) As the number of -CO groups in precursors decreases from n = 6 to 4 and then 2, the value of bifurcation of the LOL- out isosurfaces gradually increases, which means that the out electron delocalization of the molecule enhances sequentially. However, the isosurfaces of LOL- in are truncated near the -CO groups, and the hindrance to the electron delocalization of the molecule with more -CO substituents is more severe. Therefore, in MOs in C 18 -(CO) n (n = 6, 4, and 2) only exhibit a local rather than global delocalization, and C 18 -(CO) 2 shows a relatively more complete delocalization than its analogues. Similarly to that of out systems, the electron delocalization of in MOs on the short C-C bonds is also more pronounced than that on the long bonds. By comprehensively considering the electron delocalization characteristics of π out and π in MOs, we can infer that the overall electron delocalization of C 18 -(CO) n (n = 6, 4, and ) of isosurface except for the breakpoint near -CO groups.
|
61114276abc9e2220469dc9e
| 12 |
The color-filled maps of LOL- out at 0.5 Å above the ring as well as LOL- in on the ring plane of C 18 -(CO) n (n = 6, 4, and 2) molecules are plotted in Fig. for presenting the π-electron delocalization from another perspective. These contour maps show, more clearly, the difference in electron delocalization around long and short C-C bonds in the carbon ring of different molecules as well as the truncation feature of LOL- in near the -CO groups, which are consistent with the above conclusions drawn based on isosurface maps.
|
61114276abc9e2220469dc9e
| 13 |
Generally, a closed circle molecule with electron delocalization over it will produce a unidirectional ring current under the induction of an external magnetic field, and the strength of the induced current reflects the strength of its aromaticity. Next, we will discuss and compare the aromaticity of C 18 -(CO) n (n = 6, 4, and 2) by analyzing their response to external magnetic field.
|
61114276abc9e2220469dc9e
| 14 |
The anisotropy of current-induced density (ACID) is an intuitive and universally applicable method for visual analysis of electron delocalization by generating molecular ring current induced under a given external magnetic field , which can be further decomposed into the contributions of different types of molecular orbitals to gain a deeper insight. The ACID isosurfaces of π out and π in electrons, respectively referred to as ACID- out and ACID- in , in C 18 -(CO) n (n = 6, 4, and 2) are plotted in Fig. . It can be seen that the π out electrons of all three precursors show complete and obvious diatropic ring currents, which conform to the left-hand rule with the direction of the external magnetic field. The unidirectional ring currents caused by π out electrons are basically distributed at the region above/below the ring. This observation provides strong evidence for the out-of-plane aromaticity of the C 18 -(CO) n (n = 6, 4, and 2). Furthermore, the ACID- out isosurface of the precursors widens and the ring current density increases with the decrease of number of -CO groups in the molecule, indicating that the molecular aromaticity created by π out electrons gradually strengthens in the order of C 18 -(CO) 6 < C 18 -(CO) 4 < C 18 -(CO) 2 . The ACID isosurface of the π in electron system, namely ACID- in , in the molecules shows the truncation of the in-plane ring current near the -CO groups. In particular, the increase of the number of -CO groups results in an enhanced hindrance of global ring current and leads to more appearance of multidirectional local induced currents.
|
61114276abc9e2220469dc9e
| 15 |
The ACID diagrams that take into account all π (i.e., π out and π in ) electrons in the system are shown in Fig. , from which it can be seen that the induced current intensity of these precursors is consistent with the extent of electron delocalization obtained from the previous analysis, namely C 18 -(CO) 6 < C 18 -(CO) 4 < C 18 -(CO) 2.
|
61114276abc9e2220469dc9e
| 16 |
The gauge-including magnetically induced current (GIMIC) is another effective method to reveal the induced current of chemical systems under action of external magnetic field . The GIMIC maps on the ring plane of C 18 -(CO) n (n = 6, 4, and 2), which provides a complementary perspective on aromaticity from the ACID isosurface, are displayed in Fig. . It is clear that the significant diatropic/paratropic currents formed inside/outside the ring for each precursor, from which we can draw the same conclusion as that from ACID analysis that the precursor molecules are all considered to be aromatic.
|
61114276abc9e2220469dc9e
| 17 |
The isosurface and color-filled map of the ZZ component of ICSS, referred to as ICSS ZZ , for the three precursors C 18 -(CO) n (n = 6, 4, and 2) are shown in Fig. . It can be seen that there is a shielding area inside the ring protruding in the direction perpendicular to the ring plane, surrounding by a toroidal deshielding area outside the ring. The shielding/deshielding characteristics of C 18 -(CO) n (n = 6, 4, and 2) once again indicate that they should be considered as a system with remarkable aromaticity, since typical aromatic systems share the same feature of ICSS ZZ . From the figure it is obvious that the peripheral deshielding region of each molecule is interrupted by the -CO groups, and the more -CO groups there are, the more severe the interruption is. The reason for this is clearly explained by the previous orbital interaction analysis, that is the existence of the -CO groups breaks the conjugation of in electrons and hence the corresponding in-plane induced current. The strong interference of magnetic shielding characteristics is therefore generated. Note that from color-filled map in Fig. , the color inside/outside the ring become redder/bluer with decrease of the number of -CO groups, displaying stronger shielding/deshielding effect. This observation clearly demonstrates that from C 18 -(CO) 6 to C 18 -(CO) 4 then C 18 -(CO) 2 , the molecule becomes more and more aromatic. The ICSS ZZ values calculated at 1 Å above the center of the ring, referred to as ICSS ZZ , is quite robust and popular for quantitatively determining aromaticity of ring systems. The ICSS ZZ (1) values of C 18 -(CO) n (n = 6, 4, and 2) are observed to be 4.0, 7.8, and 15.3 ppm for n = 6, 4, and 2, respectively, which are lower than that of C 18 and benzene (29.9 ppm) , indicating that a relative weak but still distinct aromaticity in the molecules. The order of aromaticity strength determined by the ICSS ZZ (1) is also fully consistent with the judgment results from all above visual analyses.
|
61114276abc9e2220469dc9e
| 18 |
The bonding character, electron delocalization, and aromaticity of C 18 precursors, C 18 -(CO) n (n = 6, 4, and 2), are theoretically explored in depth by using DFT calculations and a variety of wavefunction analyses methods. The optimized geometry and various bond orders indicate that the cyclocarbon skeleton of the precursors shows an alternating structure of long and short C-C bonds, which is akin to C 18 . The results of MO analysis and IRI analysis show that these molecules have two sets of π-conjugated electron systems, respectively described by π out and π in MOs, which provide different contributions to molecular aromaticity. The orbital interaction analysis reveals the essential reason why the introduction of -CO groups destroys the π in conjugated system. Based on the analysis of LOL- and ELF- functions, the characteristics of electron delocalization in π out and π in MOs were confirmed. The ACID, GIMIC, and ICSS ZZ methods graphically revealed the prominent induced ring current or magnetic shielding effect of electrons in precursors induced by external magnetic field. All the analysis methods based on the wavefunction of quantum chemistry calculations or response to external magnetic field came to exactly the same conclusion, that is, the three precursor molecules all have π out electron global delocalization and π in electron local delocalization, so they can be definitely classified as aromatic species, and the aromatic strength increases with the decrease of number of -CO groups.
|
646e91bd4f8b1884b7396688
| 0 |
The incorporation of carboxyl groups onto olefins is one of the most useful and industrially applied reactions of the modern era (e.g., hydrocarboxylation, Figure ). The reverse process of removing a carbonyl group (deformylation/decarboxylation) is less studied, yet could serve as a convenient tool to install an olefin when the carboxylic acid is readily available, offering alternative disconnections for retrosynthetic analysis. For example, olefin 1 could be envisioned to arise from a sequence of Diels-Alder 8 and functional group manipulations or, perhaps more attractively, from readily available tranexamic acid in a single operation (following Cbz protection). One potential platform to access this reactivity would be via an electrochemically generated carbocation produced by decarboxylation (Figure ). Such reactivity is embodied by the classic Hofer-Moest reaction, a transformation that generates these reactive intermediates through a strongly oxidative process. Although there are numerous examples that utilize this electrosynthetic tactic, the vast majority proceed only on activated systems (allylic, benzylic, and a-heteroatom). Recent variants, such as decarboxylative etherification and fluorination, have extended the scope of this reaction largely to tertiary systems and select secondary examples.
|
646e91bd4f8b1884b7396688
| 1 |
Electrochemical decarboxylation of secondary, let alone primary carboxylic acids, is considered extremely challenging even in these state-of-the-art protocols. The recently developed rapid alternating polarity (rAP) variant of the Kolbe reaction is a notable exception as it enables decarboxylation of primary carboxylic acids with high chemoselectivity; however, this reaction favors carbon radical formation over the carbocation. To the best of our knowledge, a general electrochemical decarboxylative olefination has yet to be described. In this communication, we disclose a modification of the rAP-Kolbe conditions, now allowing facile decarboxylation of unactivated primary, secondary, and tertiary carboxylic acids without the need for preactivation (e. g., redox-active ester formation), enabling access to olefins via carbocations. Our preliminary studies started from slightly modified electrochemical decarboxylations for etherification and flurionation 15 (Table ). In an attempt to favor olefination, the alcohol nucleophile was excluded from the etherification conditions (entry 1). However, oxidation of the model secondary acid was sluggish, and no olefin product was observed. Fluorination conditions led to trace conversion to olefin 1, in good agreement with the original report that this method is not applicable to secondary carboxylic acids (entry 2). rAP Kolbe conditions gave much better conversion when compared to the previous methods, resulting in moderate yield of 1 (entry 3). To further improve the reaction, graphite electrodes were evaluated based on their well-established ability to access carbocations over carbon radicals. However, switching to graphite electrodes resulted in low conversion with 50 ms rAP (entry 4). This large drop in conversion could be explained by the order of magnitude higher electrode capacitance of graphite over reticulated vitreous carbon (RVC) (i.e., electricity was largely consumed by charge-discharge cycles of the electrical double layer). To account for such a large electrode capacitance, a much longer pulse was applied (5 s instead of 50 ms), resulting in a significant improvement in conversion and 67% yield of the desired olefin (entry 5). These conditions outperform state-of-the-art photochemical methods (entry 6), and as illustrated below, this functional electrochemical protocol enables unprecedented scalability (up to 1 kg) in the context of decarboxylative olefination of unactivated carboxylic acids. Remarkably, these optimized conditions did not work (on preparative scale) if direct current (DC) electrolysis was applied instead of alternating polarity. In fact, no conversion was observed under DC conditions (entry 7, mechanistic studies vide infra). A catalytic amount of base is necessary to ensure conversion of carboxylic acid 2 (entry 8). Use of tetramethylammonium salts can be circumvented by replacement with KOH (entry 9). Furthermore, the addition of pivalic acid is necessary to avoid partial oxidative degradation of the olefin product (entry 10). 2-Methyl-2-butene can also be used in place of pivalic acid (entry 11), further confirming the role of PivOH as a sacrificial additive rather than a pH buffering agent due to its acidic nature. With optimized conditions in hand, the scope of this electrochemical decarboxylative olefination was explored. In general, this practical method is applicable to a variety of unactivated secondary, tertiary, and even primary carboxylic acids as illustrated in Table . Compatible functional groups include: esters (3, 10, 11, 20), ketals (4), sulfones (5), carbamates (6, 7), amides (10), epoxides (11), arenes (12, 16), alcohols (14), alkenes (14, 19) and ketones (15). As demonstrated in 3-7, this method offers rapid access to unsaturated six-membered rings with a variety of functional groups.
|
646e91bd4f8b1884b7396688
| 2 |
Although such a motif is a prime target for Diels-Alder approaches, accessibility and stability of the reactants, as well as polarity matching, is not always trivial. For example, access to olefinic sulfone 5 can be laborious (previously reported in 7 steps), whereas electrochemical decarboxylation allows its preparation in a single step from the commercially available acid ($5.4/g). Similarly, the synthesis of 4-aminocyclohexene derivatives 6 and 7 from a readily available carboxylic acid ($4.8/g, free amine) also exemplifies the potential of this protocol to circumvent a cumbersome Diels-Alder/Curtius rearrangement sequence. Medicinally relevant gemfibrozil, used as a lipid regulator to treat high cholesterol, offers an intriguing case study on chemoselectivity. The electron-rich alkyl aryl ether prone to electrochemical oxidation was tolerated in the decarboxylative olefination, highlighting the chemoselective nature of the oxidation process. The method also allows facile derivatization of naturally occurring carboxylic acids such as ursolic acid, isosteviol, and dehydroabietic acid to deliver unique olefins with complex polycyclic scaffolds (14-16). Interestingly, in the case of camphoric acid, the tertiary acid was selectively decarboxylated to afford g,d-unsaturated acid 13. Simple primary carboxylic acids-the most abundant yet least prone to undergo Hofer-Moest-are efficiently decarboxylated under this protocol. Oleic acid, a common component of plant oil, can be turned into an exotic diene 19, while suberic acid monomethyl ester can be converted into a valuable unsaturated ester 20 ($160/g). In addition to olefin product formation, this protocol demonstrates different reaction outcomes based on the nature of the carbocation intermediate. b-Hydroxy acid 8 can readily undergo decarboxylation followed by pinacol-type rearrangement to form cyclic ketone 9, while adamantanol derivative 17 efficiently leads to Grob fragmentation product 18. Another unique evidence of carbocation intermediacy was obtained in the decarboxylation of a primary acid, where the formation of alkene 22 was accompanied by cyclopropane 21. This byproduct formation invokes a non-classical carbocation stabilization followed by a new C-C s-bond formation via gdeprotonation. Since oxidative decarboxylation is balanced by reduction of acetone used as solvent, confirmed by the detection of ~1.0 equiv. of isopropanol in crude H NMR (see SI), functional groups susceptible to electron-transfer within the redox window of carboxylate and acetone might not be tolerated (Table , bottom left). Therefore, reductively labile functionalities such as alkyl halides (23), aryl sulfonamides (24), electron-poor heterocycles (25), or enones (26) were not amenable to the current reaction conditions. Similarly, tertiary alkyl amines (27) or electron-rich heterocycles (28) interfered with oxidative decarboxylation. In certain cases, olefin formation provided an inseparable mixture of isomers. In this scenario, the mixture could be converted into a single product by enlisting alkene isomerization conditions (Table , bottom right), using the recently reported electrocatalytic generation of Co-H to access the internal isomer 12b. The ultimate proving ground for this simple, metal-free protocol is in its applicability to be conducted on a large scale. Thus, we partnered with AbbVie process chemists to scale the electrochemical decarboxylative olefination of carboxylic acid 2 (Figure ). Translation of the method to the continuous stirred-tank reactor (CSTR), employing a cylindrical electrode array (Figure , top right), was straightforward. Equivalents of base and sacrificial additive were modified, and the electrolysis mode was switched from constant current to constant potential, to improve the current flow and minimize reaction time. Since the AbbVie reactor (see SI) allows facile optimization of reaction temperature, it was fine-tuned to 35 ºC. Overall, in less than 10 lab experiments on 7 g scale, suitable conditions were identified for scale-up. Notably, the reaction stalled when DC electrolysis was employed (Figure , top left), confirming that the importance of alternating polarity is universal across various reactions scales. Increasing the scale of the reaction from 7 to 28 g in a larger reactor led to nearly identical reaction performance (Figure , bottom left). This preliminary scale-up effort successfully demonstrated that scaling the reaction based on electrode surface area would lead to predictable results. Thus, further scale-up to 1 kg was undertaken by simply using a larger electrode array (Figure , top right). The surface area of this electrode supposes a ten-fold increase relative to the array used for the 28 g reaction and fits into an 8 L reactor body. To achieve a batch reaction size of 1 kg in this reactor, the reaction was further concentrated from ~0.25 M under the nominal conditions to 0.8 M without consequence.
|
646e91bd4f8b1884b7396688
| 3 |
The next hurdle to be addressed was identifying equipment capable of providing alternating polarity current. The expected current flow for a 1 kg reaction would be in excess of 100 A; however, no power source capable of supplying this amperage with polarity alternation was readily available on the market. Therefore, a lab potentiostat capable of a 20 A maximum output was used as a compromise. The elevated capacitance of the graphite electrode material presented an additional challenge, which was addressed by designing a custom step-down reversal of polarity (see SI for further details). Despite the equipment limitation resulting in a longer reaction time (90 hours), the kg-scale reaction proceeded smoothly, providing a comparable 80% isolated yield of olefin 1 (Figure , bottom right). To place these results in context, a hypothetical cost comparison of the analogous photochemical variant 20 is put forth, wherein the electrochemical reaction would require $89 for all the reagents to provide 675 g of 1 ($26/mol of 1), while the Ir catalyst alone would cost >$1,000 ($302/mol of 1) on the same reaction scale. Alternating polarity (AP) every 5 seconds and DC electrolysis usually share similar reactivity profiles; 30 therefore, it is of great surprise that merely switching from AP to DC results in little conversion under otherwise identical conditions (Table , entry 7). Accordingly, preliminary mechanistic investigations were conducted to provide a deeper understanding on the effect of AP on this transformation (Figure ). The most notable visual difference between the reactions with and without AP was the appearance of electrodes after completion of the electrolysis (Figure ).
|
646e91bd4f8b1884b7396688
| 4 |
While the appearance of the electrodes under AP conditions remained unaltered, the anode surface of the DC experiment electrodes had clearly been compromised. Moreover, electrode weight was monitored before and after electrolysis, observing non-negligible mass gain only on the DC anode, further supporting irreversible anode fouling. It was reasoned that such anode fouling occurs due to the depletion of oxidizable substance on the anode. This situation is comparable to the rAP-Kolbe reaction, where smooth decarboxylation only under rAP conditions could be explained by a local pH difference around the electrodes. Namely, DC electrolysis generates locally acidic areas around the anode through the formation of electrogenerated acids, which suppresses deprotonation of carboxylic acid, and therefore, decarboxylation. In contrast, polarity switching can partially avert this phenomenon by reversing electrode polarity, thus diminishing the accumulation of acid around the electrode. In order to support this hypothesis, probe 30 bearing an acid-labile silyl ether functionality was introduced into the reactions (Figure , top panel).
|
646e91bd4f8b1884b7396688
| 5 |
Furthermore, cyclic voltammetry (CV) studies clearly indicate the lack of an oxidation event without adding a base (Figure , bottom panel), which could lead to high potential at the electrode and result in anode fouling. On the other hand, CV of carboxylic acid 29 in the presence of base exhibits an oxidation peak around 1.5 V, which is in good agreement with the reported oxidation potential of alkyl carboxylates. This oxidation event could be acting as overcharge protection, thus maintaining the integrity of the electrodes throughout the reaction. Following the acidity study highlighted in Figure , further CV studies were undertaken to provide a more detailed analysis on the impact of AP on this decarboxylative process (Figure ). In order to replicate the reaction process as closely as possible, a graphite electrode was used as working electrode, and potential was swept for 10 cycles between +3 V and different low ends. Three values were chosen for the low end: 0 V (barely reductive), -1.0 V (reductive, but no solvent reduction), and -2.5 V (sufficiently reductive for solvent reduction to occur). Notably, in the +3 -0 V window, the oxidative current intensity dropped considerably after the first cycle, possibly indicating electrode fouling. Lowering the reductive end of the window to -1.0 V attenuated this current drop. Upon reaching -2.5 V, a new reductive event associated with acetone reduction was observed, and the oxidative current intensity was maintained during ten cycles. This result indicates that the reduction of acetone, which consumes protons, is crucial to support smooth decarboxylation over time, thus resupplying carboxylate species for further oxidation. Based on the observation of isopropanol in crude 1 H NMR on preparative scale, this scenario likely explains the role of AP.
|
646e91bd4f8b1884b7396688
| 6 |
To summarize, a simple protocol for decarboxylative olefination is presented which precludes the need for expensive catalysts, ligands, additives, or any metals. The closest alternative method to this transformation is a photochemical decarboxylative olefination, which is cost-prohibitive on scale. Indeed, the scalability of this electrochemical reaction is vividly illustrated by the facile
|
675826d5085116a133ee45fd
| 0 |
Rare-earth metal ions (Sc, Y, lanthanides) have attracted significant attention, mainly due to their Lewis acidity and oxophilicity. In particular, Sc(OTf)3 has been introduced in recent decades as a promising Lewis acid for C-H functionalisation and C-C bond formation reactions. In the past 20 years, scandium salts have also effectively assumed a role in photoredox catalysis-a process that has emerged as a full-fledged alternative in organic synthesis, often unlocking transformations that are not attainable in the dark. Photoredox catalysis relies on an electron transfer process between a highly reactive, light-excited catalyst and a redox-labile substrate (S). The result is a reactive radical intermediate that undergoes further transformations, ultimately resulting in the formation of new bonds under mild conditions. Photoredox catalysis requires a catalyst that is both photochemically and redox active-typically a compound derived from an organic dye or a complex of a redox-active metal, such as ruthenium or iridium (see Fig. ). Conversely, redoxinactive/innocent metals, such as scandium, which have a predominantly trivalent redox-stable state, are not directly involved in photoredox reactions. Nevertheless, these metals, and particularly Sc(OTf)3, can serve as Lewis acids (LA) that activate substrates/organophotocatalysts by altering their redox characteristics, thereby facilitating their participation in photoredox and photosensitisation reactions (Figs. and). For instance, the redox properties of a flavin derivative (Fl) were significantly enhanced via their coordination to Sc(OTf)3. This alteration enabled the oxidation of some electron-deficient substrates via a SET process between the substrate and flavin in excited Fl-2Sc 3+ complex (Fig. , right); such reaction cannot be achieved without Sc(OTf)3. Notably, the oxophilicity of some M n+ cations facilitates their binding to superoxide radical anion (O2 •-) to form superoxide-metal complexes (M n+ -O2 •-) that have been recognised within the reductive activation of dioxygen by metalloenzymes in numerous biological redox reactions that employ O2 as an oxidant. This process has been mimicked in the dioxygen activation cycle during photoredox transformations, commencing with the photoinduced single electron transfer (PET) from a light-absorbing, redox-active compound to O2 and followed by the generation of M n+ -O2 •-complexes (Fig. , upper part). Binding of M n+ to oxygen enhances its ability to accept an electron. Thus, an electron transfer from donor (D) to O2, which is thermodynamically infeasible in the absence of metal ions, becomes possible in the presence of M n+ . This process is generally called metal ion-coupled electron transfer (MCET). Scandium 3+ ion, with its smallest radius among M 3+ , can bind strongly with O2 •-, making it far more effective than other M n+ ions for providing SET reactions driven by M n+ -O2 •-complex formation A review of previous reports on the role of Sc(OTf)3 in photoinduced electron transfer (PET) transformations involving the reductive activation of dioxygen confirms that a light-absorbing electron donor and additional stabilisation are necessary. For example, Fukuzumi recorded the EPR spectrum of the (HMPA)3Sc 3+ -O2 •-complex stabilised by three equivalents of hexamethylphosphoramide (HMPA) ligand, formed via PET from the singlet excited state of 1-benzyl-1,4-dihydronicotinamide dimer 1 [(BNA)2]*, to O2. This procedure was accompanied by fast cleavage of the C-C bond in the formed (BNA)2 •+ (Fig. , bottom part). Our present study is the first report to demonstrate the activity of Sc(OTf)3 functioning as a sole photoredox catalyst. Sc(OTf)3 is shown to facilitate aerobic oxidation of a benzylic C-H bond and direct oxidative cyanation of arenes as examples of C-H activation and C-C coupling reactions (Fig. ). These unprecedented procedures dispel the usual notion of scandium salts as redox-inactive species useful mostly as Lewis acids and expand their possible application as simple, readily available photocatalysts. In general, photoredox processes might be a further application of scandium, which is abundantly found on earth but remains an underutilised metal. 30
|
675826d5085116a133ee45fd
| 1 |
Exploring and optimization of Sc(OTf)3-based aerobic photocatalytic oxidation of benzylic substrates Upon the surprising observation of the photocatalytic activity of Sc(OTf)3, we directed our attention towards the aerobic oxidation of toluene (1a; Eox = +2.36 V vs SCE ) as a benchmark. The oxidation was performed with 5 mol% of Sc(OTf)3 (Ered = -1.15 V vs SCE 32 ; i.e. not strong enough in ground state to oxidize 1a) in O2-saturated acetonitrile (MeCN) under blue light irradiation (400 nm LED) for 24 h. Encouragingly, this experiment afforded benzoic acid (2a) as the sole product at an 80% yield (Fig. , entry 1). Under anaerobic conditions, or in the absence of light or Sc(OTf)3, the reaction did not proceed (entries 2-5). This system demonstrated significantly decreased effectiveness with 450 nm LED illumination (Entry 6). An assessment of Lewis acids revealed that Sc(OTf)3 was the optimal selection, probably due to its superior ability to bind strongly with O2 An investigation of catalyst loading (Fig. , entries 7 and 8 and Supplementary Section S2) also revealed that using a smaller amount of Sc(OTf)3, at 2.5 mol%, slightly reduced the yield of benzoic acid (2a), whereas employing 10 mol% of the catalyst notably achieved a full conversion to 2a. Moreover, the time-dependent yield profile showed an almost constant rate of toluene (1a) oxidation with increasing reaction time (Fig. ). Consequently, the plot of turnover frequency (TOF) values over the eaction time (4-20 h) remained stable, signifying maintenance of the photocatalytic activity of Sc(OTf)3 (see Supplementary Section S2).
|
675826d5085116a133ee45fd
| 2 |
The unexpected finding of the photocatalytic activity of Sc(OTf)3 prompted us to check our experiments thoroughly multiple times. We ensured that no organic impurities were driving the reaction. We also tested Sc(OTf)3 from different suppliers, but we found no significant variations except for reduced activity caused by traces of oxygen atom-containing solvents like MeOH, attributable to synthesis residues. This reduced activity could be improved by recrystallising Sc(OTf)3 from its solution in MeCN and using chloroform for precipitation. The photocatalytic activity could be ascribed to the weak absorption of Sc(OTf)3 in MeCN (this extended to the visible region around 400 nm), and it was also indicated by blue fluorescence with a distinct band at 505 nm observed upon excitation with 400 nm light (Fig. ).
|
675826d5085116a133ee45fd
| 3 |
Using our optimised conditions, we evaluated various substrates. Both electron-donating and electron-deficient aromatic compounds (Fig. ) revealed the promising photocatalytic activity of Sc(OTf)3 in most cases. Toluene (1a) yielded 80% benzoic acid (2a) and p-chlorotoluene (1b) (Eox = +2.43 V vs SEC ) underwent even complete conversion to its respective carboxylic acid 2b. As anticipated, the electron-deficient 4-(trifluoromethyl)toluene (1c), with the highest oxidation potential (Eox = +2.61 V vs SEC ), displayed considerable resistance to oxidation, resulting in trace product formation. Conversely, the less electron-deficient ester group-containing derivative 1d (Eox = +2.45 V vs SCE ) was effectively oxidised. Unexpectedly, 4-methylanisole 1e (Eox = +1.51 V vs SEC ), despite being the most electron-rich substrate, gave only a 4% yield. This could be attributed to the back-electron transfer (BET) process during the oxidation of 1e. Biphenyl 3 (Eox = +1.95 V vs SEC 34 ) underwent efficient oxidation, providing a complete yield of benzoic acid (2a). Oxidation of p-xylene (1f) afforded a mixture of 4-methylbenzoic acid (2f) (67%) and terephthalic acid (2g) (33%) after 24 h of irradiation. Ethylbenzene derivatives 4a-c provided reasonable yields of benzoic acid derivatives according to the electronic effect, although 4-ethyltoluene (4d) was oxidised to a mixture of 4-acetylbenzoic acid (2h) and 2f. By contrast, cumene (5) gave a poor yield of benzoic acid upon oxidation. Diphenylmethane (6) was transformed into a mixture of the corresponding ketone 7 and benzoic acid (2a). 9H-fluorene (8) was converted into fluorenone (9) as a sole product with excellent yield, while stilbene (10) yielded benzoic acid via oxidative C=C cleavage. Further investigation of the oxidation of various primary and secondary aryl alcohols, as depicted in Fig. , revealed that all the tested primary benzyl alcohols, including electron-rich and highly electrondeficient ones 11a-d and o-chlorobenzyl alcohol (11e), were oxidised to the corresponding benzoic acids 2 in quantitative yields. (13c), as anticipated. Surprisingly, the oxidation of 1-phenylethanol (12a) resulted in a low conversion to the corresponding ketone 13a, for a yield of only 23%. A mixture of benzophenone (15) and benzoic acid was obtained following the oxidation of diphenylmethanol (14). By contrast, cinnamyl alcohol (16) underwent a C=C bond cleavage during oxidation to generate a quantitative yield of benzoic acid. Notably, 4-methoxybenzyl methyl ether (17) yielded benzoic acid and not the corresponding ester. The optimised conditions were subsequently used at a preparative scale, with 1 mmol of the selected substrates (see Fig. ) to confirm the practicality and effectiveness of our aerobic photocatalytic oxidation approach employing Sc(OTf)3 (see Supplementary Section S3 for details). The reaction progression was monitored using 1 H NMR analysis and the time was optimised to achieve the maximal conversion of reactants to products. Interestingly, p-xylene (1f), a substrate with two potential oxidation sites, was oxidised chemoselectively to 2f at the preparative scale after appropriately adjusting the reaction time. b Isolated yield (in green); experiments were performed at a 1 mmol scale (see Supplementary Section S3 for details). c Only 2f as the sole product. d 48 h. e 59% of 13b and 16% of 2b, 24 h.
|
675826d5085116a133ee45fd
| 4 |
A further aim of this study was to demonstrate the utility of Sc-based photoredox catalysis. We chose arenecarbonitriles, which serve as crucial structural scaffolds in synthesising bioactive compounds, alongside their substantial role as intermediates in organic synthesis. We particularly focused on the direct photooxidative cyanation of the aromatic C-H bond, as this reaction remains a formidable challenge for the chemical community because of the need for prefunctionalised starting material and the lack of regioselectivity.
|
675826d5085116a133ee45fd
| 5 |
Moreover, photocatalytic methodologies for direct C-H cyanation are limited and still require further exploration. We first used diphenyl ether (18a) (Eox = 1.88 V, vs. SCE) as a benchmark substrate and performed the cyanation using TMSCN as a nucleophilic cyanating agent (12 equiv.) in oxygen-saturated MeCN under blue light irradiation (400 nm) for 24 h. Our preliminary investigations (Fig. , entry 1) showed that a mixture of 4-phenoxybenzonitrile (19a) and 2-phenoxybenzonitrile (19b) could be obtained in a good yield of 74%, with notable regioselectivity towards the para-substituted isomer. Control experiments demonstrated that the cyanation of 18a ceased in the absence of light or Sc(OTf)3, or under anaerobic conditions (entries 2-5). Raising the wavelength of the irradiation source from 400 nm to 450 nm resulted in only trace yields of the desired product (entry 6). For explanation, the cyanating agent TMSCN was used in a significant excess (12 equiv.) to avoid competitive photooxidation reactions (see Supplementary Section S4). Similarly, as was observed for the benzylic oxidations, switching Sc(OTf)3 to Mg(OTf)2, Zn(OTf)2, La(OTf)3, or Ba(OTf)2 resulted in the cessation of cyanation under otherwise identical experimental conditions (see Supplementary Section S4). Switching from MeCN to MeOH or DMSO also completely inhibited the reaction, whereas the reaction in MeNO2 occurred with remarkable conversion. Reaction in dioxane gave small amount of product (Fig. , entries 7-10). Reducing the amount of the Sc(OTf)3 catalyst (2.5 mol%) lowered the yield, whereas doubling the amount (10 mol%) gave a slightly higher yield (entries 11 and 12). Using the optimised conditions, we next investigated the direct cyanation of various arenes, including both electron-rich and electron-poor substrates, as depicted in Fig. . Cyanation of 1,3,5-trimethoxybenzene (18b) and 1,3-dimethoxybenzene (18c) furnished the corresponding carbonitriles 19c and 19d, respectively, as sole products in quantitative yields. Cyanation of 3-chloroanisole (18d) resulted in a mixture of 4-chloro-2-methoxybenzonitrile (19e) and 2-chloro-4-methoxybenzonitrile (19f) in excellent yields, with 3.9:1 regioselectivity. Efficient cyanation of biphenyl (18e) occurred, yielding 4-phenylbenzonitrile (19g). Cyanation of butoxybenzene (18f) led to the formation of 4-butoxybenzonitrile (19h) in a quantitative yield; however, the efficiency of this system in the cyanation of electron-poor substrates was only modest. Cyanation of methyl 2-methoxybenzoate (18g) and 1-methoxy-2-(trifluoromethyl)benzene (18h) gave poor yields of the desired products 19i and 19j, respectively. The highly electron-poor substrates, such as chlorobenzene (18i), methyl benzoate (18j), and trifluorotoluene (18k), did not yield any cyanation products. Subsequent application of the optimised conditions using 1 mmol of the selected substrates 18a and 18d led to cyanation products in good yields, thereby confirming the practicality of our cyanation approach on a preparative scale.
|
675826d5085116a133ee45fd
| 6 |
Additional control experiments (see Figs. and (Fig. ), indicating the involvement of a radical pathway in this oxidation process. Similarly, the cyanation of diphenyl ether was almost completely halted in the presence of TEMPO (see Supplementary Section S7 for details). Stern-Volmer quenching experiments revealed that toluene, diphenyl ether and 1,3,5-trimethoxybenzene quenched the fluorescence of a Sc(OTf)3 solution in acetonitrile with Stern-Volmer constants Ks of 3, 22 and 42 L mol -1 , respectively (see example in Fig. ). This signified SET from the tested substrates to an excited scandium complex (see Supplementary Section S7 for details). As mentioned earlier, the UV-Vis absorption spectrum of Sc(OTf)3 in acetonitrile showed an elongated tail in the visible region around 400 nm and a fluorescence emission spectrum with a sharp band at 505 nm (Fig. ). ESI-MS measurements aimed at elucidating the structure of the scandium complex responsible for this absorption in a solution of Sc(OTf)3 in MeCN. It suggested the formation of a cluster of complexes, including [Sc(OTf)2(MeCN)2] + , as the primary species. However, these particles do not absorb in the visible region but below 260 nm, as indicated by the theoretical spectra (see Supplementary Section S10). By contrast, the [Sc(MeCN)3((η 2 -O2)] 3+ particle in S0 state, a complex with three MeCN molecules and coordinated oxygen molecule, seems to absorb in the visible region (about 415 nm) as documented by the theoretical spectrum, consistent with our experimental results (Fig. ). Indeed, the coordination of oxygen molecule is ascribed to the oxophilicity of Sc 3+ ions and is further supported by our EPR measurements (see below). It should be noted that the isomeric [Sc(MeCN)3(η 1 -O2)] 3+ complex (in T1 state) is more stable by 0.950 eV (91.67 kJ mol -1 ) at DLNPO-CCSD(T)/aug-cc-pVTZ level of theory (see Supplementary Section S10 for further information) compared to the S0 state [Sc(MeCN)3((η 2 -O2)] 3+ , nevertheless, [Sc(MeCN)3(η 1 -O2)] 3+ is proposed to not absorb in the visible region (absorption below 300 nm, see Supplementary section S12). We believe the small barrier can be overcome thermally.
|
675826d5085116a133ee45fd
| 7 |
In subsequent experiments, we investigated the mechanism of Sc-based photooxidative catalysis using toluene oxidation as the representative example. Irradiation of an oxygen-saturated acetonitrile solution containing Sc(OTf)3 (0.1 M) and toluene (2 M) with a 400 nm LED at room temperature resulted a distinct isotropic EPR signal (Fig. ). The pronounced eight-line isotropic spectrum was attributed to the formation of a complex of superoxide radical anion with scandium ion, Sc 3+ -O2 •-. This resulted in superhyperfine splitting (aSc = 4.28 G) due to the 7/2 scandium nuclear spin and featured an isotropic g value, giso = 2.0160. This observation is consistent with reported EPR spectra in the literature. Interestingly, while previous studies required a low temperature to observe Sc 3+ -O2 •-by EPR, we were able to obtain well-defined spectra at room temperature. The kinetic recordings showed stability of the Sc 3+ -O2 •-complex for about 2 min after the cessation of irradiation (Fig. , insight). Importantly, no significant EPR signal appeared upon irradiation of a blank solution containing no toluene. Irradiation of the acetonitrile solution of Sc(OTf)3 and toluene under anaerobic conditions also did not reveal any EPR signals. Replacing Sc(OTf)3 with La(OTf)3 (which showed no catalytic activity in photooxidative procedures) led to a very faint EPR signal, even after an extended period of irradiation (see Supplementary Section S8). These observations provide evidence that all species -toluene as a substrate and electron donor, oxygen as a terminal electron acceptor and scandium as mediator -are necessary for efficient photoinduced electron transfer and stabilised Sc 3+ -O2 •-formation (see Fig. for a proposed mechanism).
|
675826d5085116a133ee45fd
| 8 |
We utilised ultrafast transient absorption (TA) spectroscopy to unravel the nature of the excited state species involved. We performed measurements of Sc(OTf)3 in MeCN under both O2 and N2 atmospheres (purging for 30 minutes, followed by gassing with N2 in a purity of 99.99%). Initially, measuring a solution of Sc(OTf)3 in MeCN under an inert N2 atmosphere resulted in a strong bleach signal at approximately 500 nm, accompanied by weak excited-state absorption (ESA) features (Fig. ). This bleach rapidly converted into a positive signal over a slower time scale of a few nanoseconds. The positive signal then continuously grew over the microsecond time scale until reaching a maximum of around 1 microsecond, after which it faded slowly, as depicted in Figs. . In the presence of O2, similar kinetic behaviour was found at the early stage, but with a feeble TA signal, and only during the long timescale (microseconds, the decay was somewhat faster than this in the N2 atmosphere) (see Fig. ). The initial bleach signal at 500 nm was assigned to the initial fluorescence of the Franck-Condon (FC) state, which quickly relaxed to some nonemissive state (S1 relaxed); this took 300 fs. This relaxed state populated another state, the triplet state, at a slow rate (10 ns). The triplet state was then slowly quenched over a microsecond lifetime. The weak invariance between the N2 and O2 atmosphere was attributed to the O2 that was strongly bound to the Sc 3+ complex and present even under the N2 atmosphere because of strong Sc oxophilicity. A role for the triplet state of O2 in the population of the Sc 3+ complex triplet state was highly expected. When measured in toluene/MeCN (Figs. ), the transient absorption data were changed to a greater extent than those in pure MeCN, but remained similar under O2 and N2 atmospheres. For instance, the initial fluorescence from FC was missing, but a fast lifetime decay of 500 fs (due to the relaxation state) was present, followed by a lifetime rise of 20 ps that was assigned to the early formation of the triplet state (see Fig. ). Within 20 ns, the bleach signal around ca. 500 nm started to evolve again, followed by an increase in the overall ESA features across the visible region over 90 ns. The appearance of the bleach signal was assigned to the delayed fluorescence of the Sc complex, which can happen due to the reversible ISC (r-ISC) process from the triplet to the singlet state, as depicted in Fig. . Due to the presence of toluene molecules, an ET was expected to occur from toluene to the Sc complex on a time scale of 90 ns, leading to the disappearance of the fluorescence peak and formation of S1 -state, as evident from Fig. . The recombination of this reduced state to the S0 state occurred at a slow rate > 1.2 microseconds. The absence of toluene remarkably affected the Sc complex behaviour by delaying the triplet state formation in MeCN vs in the MeCN/toluene mixture. The insensitivity of the Sc complex towards the O2 atmosphere also highlighted the role of strongly bound O2 molecules in extending the lifetime of the observed excited states. For instance, the slow decay of the S1 - state was expected to be due to structural rearrangements within the Sc complex, leading to the desorption of the reduced O2 into the reaction mixture for further use in product formation pathways.
|
675826d5085116a133ee45fd
| 9 |
Drawing from the outcomes of control experiments, mechanistic studies using EPR and transient absorption measurements (TA), and considering previous reports, a plausible mechanism for the Sc(OTf)3-catalysed photooxidation reactions is postulated (Fig. ). Initially, the Sc 3+ complex can be excited upon coordination with molecular oxygen, as demonstrated by the quantum chemical calculations (see Fig. and Supplementary Section S10). This type of complex is proposed to involve acetonitrile ligands and the participation of triflates either as counterions or ligands. Upon excitation, the scandium complex forms a singlet excited state, which can subsequently undergo an intersystem crossing process, as elucidated by TA, to generate a triplet excited state of the scandium complex.
|
675826d5085116a133ee45fd
| 10 |
Nevertheless, according to TA measurements, in the presence of toluene, a delayedsinglet excited state of the Sc 3+ complex (with oxygen involved) seems to be responsible for the productive reaction pathway that undergoes SET from toluene during the production of a toluene radical and the Sc 3+ -O2 •-complex, as verified by the eight-line EPR spectrum arising from the I = 7/2 45 Sc nucleus. Subsequently, the superoxide ion, O2 •-abstracts a proton from the benzyl radical cation to produce a hydroperoxy radical, HOO • and a benzyl radical. Subsequently, the interaction between the benzyl radical and O2 produces the peroxyl radical, followed by benzaldehyde formation. Further oxidation of benzaldehyde yields the desired product, benzoic acid (2a), as well as H2O2, as confirmed by iodometry (see Supporting Section S9 for details). Alternatively, the aryl radical cation can react with a cyanide source (TMSCN) to generate a cyclopentadienyl radical and produce a cyanoarene by hydrogen transfer. The final transformation using the hydroperoxy radical is likely the one proposed by Nicewicz but an alternative participation of oxygen is not excluded.
|
675826d5085116a133ee45fd
| 11 |
All the results obtained during this study confirm that Sc(OTf)3 possesses noteworthy photocatalytic activity. Hence, we introduce a new aerobic oxidation approach for benzylic substrates with Sc(OTf)3 as the sole photocatalyst. This system revealed considerable efficiency with a wide range of benzylic substrates, including alcohols, toluene derivatives, and methylenecontaining substrates. A straightforward Sc(OTf)3based photocatalytic approach for the oxidative cyanation of arenes under aerobic conditions was also developed as an example of direct aromatic C-H derivatization resulting in C-C bond formation. The photooxidative cyanation method, despite some limitations (e.g. the cyanation of electron-poor substrates), demonstrated an interesting efficiency with a notable regioselectivity across the tested scope of arenes.
|
675826d5085116a133ee45fd
| 12 |
Our results on benzylic photooxidation and oxidative cyanation mediated by Sc(OTf)3 as sole photocatalyst break the long-held belief that Sc 3+ salts serves only as a potent Lewis acids for activating organophotocatalysts or substrates. We expect that Sc 3+ salts can find applications as photocatalysts also in other oxidative organic reactions, some of which are now under investigation in our laboratory. From a general perspective, our examples could pave the way for the exploration of other redox-innocent metals in photoredox catalysis. was monitored by 1 H NMR. When the reaction was completed, the solvent was evaporated, and the crude product was purified either by column chromatography on silica gel or by extraction (see Supplementary Section S2 for details). General procedure for photocatalytic oxidative cyanation of arenes Small scale experiments (C). A vial was charged with a mixture of the substrate (140 μmol), Sc(OTf)3 (5 mol%, 7 µmol) and TMSCN (12 equiv., 1.68 mmol) in MeCN (250 μL). The reaction mixture was bubbled with oxygen (2 min) and then stirred at 45 °C under irradiation with 400 nm LEDs. After selected time, the reaction mixture was diluted with DMSO-d6 and the yield was determined by 1 H NMR.
|
675826d5085116a133ee45fd
| 13 |
A mixture of the substrate (1 mmol), Sc(OTf)3 (5 mol%) and TMSCN (12 equiv., 12 mmol, 1501 μL) in MeCN (5 mL) was bubbled with oxygen (2 min). It was then stirred at ambient temperature in a 50 mL Schlenk tube under irradiation by 400 nm Luxeon LEDs under aerobic conditions (oxygen balloon). The reaction progress was monitored by TLC and 1 H NMR. When the reaction was completed, the solvent was evaporated, and the crude product was purified by column chromatography on silica gel (see Supplementary Section S6 for details).
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.