id
stringlengths 24
24
| idx
int64 0
402
| paragraph
stringlengths 106
17.2k
|
|---|---|---|
66b23c3f01103d79c5e84bbe | 24 | Graph encoder: Prior to capturing the complex interaction between the surface graphs and the adsorbate features, geometric GNNs are used to encode the surface graphs into atom-wise embedding, which contains chemical and structural information. Formally, given a surface graph G c = (H, E) for the combination c, the atom embedding matrix H ′ is computed according to: |
66b23c3f01103d79c5e84bbe | 25 | where g c is the surface graph embedding, W S is a learnable weight matrix, s i is the depth embedding of surface atom i, and S is the surface atom depth embedding matrix similar to the position encoding of AdsGT encoder. The depth embedding s i describes the relative position |
610cf54f4cb479f729299249 | 0 | Curcumin (CURC) is a natural compound, which is extracted from turmeric (curcuma longa) and is widely used as spice or food additive. The latter is also due to its intense orange colour, which makes curcumin particularly attractive for the food processing industry. More recently the interest in CURC also concerns its remarkable properties as antinflammatory and antibacterial drugs. Furthermore, it has been recently confirmed that curcumin exerts an important anticancer activity, which has further revived the pharmacological interest towards this natural compound. In addition, curcumin absorbing in the visible light, the possible exploitation of its excited state manifold for photodynamic (PDT) or light assisted chemo-therapy (LAC) have also been investigated. In particular, exploiting CURC, a natural occurring non-toxic spice, for PDT aimed at food processing and its antibacterial treatment is particularly attractive. Other analogue components of turmeric such as cyclocurcumin, have also been highlighted for their possible photophysical and photochemical properties. However, to ensure the efficiency of PDT, which is based on the activation of molecular oxygen to its reactive singlet state 1 O2, crucial photophysical conditions should be reunited. In particular, a high intersystem crossing (ISC), leading to a fast and efficient population of the photosensitizer triplet state manifold is required. Furthermore, the accessible triplet state should be energetic enough to allow sensitization of the molecular oxygen. Obviously, intense absorption in the visible, or even near infrared portion of the electromagnetic spectrum, together with a high bioavailability are also necessary to increase the global efficiency. Unfortunately, the last two properties are usually conflicting, since red-shifting the absorption spectrum usually requires the inclusion of large pconjugated moieties, which in turn diminish solubility in aqueous media and biological conditions. As a matter of fact, the use of specific drug-delivering techniques for PDT agents, also based on their encapsulation in supramolecular aggregates or liposomes, has been proven successful to increase bioavailability and in vivo efficiency. A related advantage of tailored-drug delivery protocols relies on the possibility to functionalize the vector to enhance selectivity for cancer cell over healthy ones, hence adding up to the spatial selectivity brought by the use of light and hence minimize side-effects. Moreover, to maximize the efficiency of PDT or LAC, and in general light-activated phenomena, the influence of nonradiative decay channels, such as thermal deactivation or photochemical isomerization, should be minimized to avoid quenching of the active excited state. Hence, molecular modelling and simulation occupies a place of choice in the development of efficient PDT and LAC agents. Indeed, it allows to disentangle the complex interplay between opposite phenomena as well as the crucial influence of the inhomogeneous surrounding on the photophysical properties. In this sense molecular modelling and simulation are also invaluable to foster a more successful rational molecular design. |
610cf54f4cb479f729299249 | 1 | Despite its non-negligible 1 O2 quantum yield, the complexity evoked is also clearly present for CURC. Indeed, natural CURC may exist in two tautomeric forms, namely keto and enol, which coexist in nature including upon administration into the human body. Recently, we have also shown that the inclusion in liposome or membranes, is susceptible to displace the equilibrium between the two forms, hence posing important burdens on the possibility of drug delivering strategies. As shown in Scheme 1, the structural properties of keto and enol form are significantly different, which in turn leads to different optical and photophysical properties. Indeed, the enol tautomer is stabilized by an intermolecular hydrogen-bond (H-bond) between the OH and C=O groups. Such interaction confers rigidity to the global structure, and more importantly leads to a highly planar scaffold, which favours delocalization and p-conjugation. As a consequence, enol CURC presents a considerable red-shifted absorption spectrum, which is highly beneficial for biological applications. Conversely, and also due to the C=O lone pairs repulsion the keto tautomers shows a broken conjugation pattern and a twisted geometry, which correlates with a blue-shifted absorption. We have previously demonstrated, also through the proper exploration of the involved potential energy surfaces (PES), that both tautomers offer suitable, and energetically favourable, routes for ISC as well as possible fluorescence. However, the enol tautomer is usually regarded as unsuitable for PDT applications and ISC because of the opening of a relaxation pathways, based on the excited state transfer of the hydrogen involved in the intramolecular H-bond. Indeed, the lowest excited states of CURC having np* and pp* character they may lead to the weakening of the OH bond and hence favour the transfer. In this contribution, going beyond our previous modelling, we resort to non-adiabatic molecular dynamics (NAMD) in the state-hopping (SH) formalism, to assess the competition between H-transfer and ISC during the relaxation of the initially populated excited states. This is possible thanks to the inclusion of general coupling in the SH Hamiltonian, including spin orbit coupling (SOC) elements. Our SH simulations clearly confirm the occurrence of the Htransfer, already at a sub-ps timescale. However, we also show that ISC is not quenched and a significant, albeit non-dominant and non-ultrafast, population of the triplet state manifold is achieved. |
610cf54f4cb479f729299249 | 2 | As previously stated, the equilibrium ground state of CURC is strongly dependent of its tautomer and indeed the enol form is highly rigid and planar. As shown in Figure an evident intramolecular H-bond takes places stabilizing the planar arrangement of the chromophore. As a comparison the tilted structure of the keto tautomer is also reported in Figure , even if in the following the latter will not be considered anymore. |
610cf54f4cb479f729299249 | 3 | presence of a lower laying pp* state (S1), for which the symmetry allowed transition yields a high oscillator strength, and a darker np* state (S2). Note that, coherently with what reported elsewhere the two states appears quasi-degenerate and have an excitation energy of about 3.7 and 3.9 eV, respectively. Note also that because of the high difference in the relative oscillator strength 2.048 vs 0.002 one may safely assume that the photophysics of enol CURC can be inferred considering the initial population of S1, only. An extended benchmark reporting the excitation energies calculated with different functionals and basis set can also be found in Electronic Supplementary Information (ESI). Note that, as shown in a previous contribution, the manifold of the singlet excited states for keto CURC will be more complicated due to the breaking of symmetry leading to an ensemble of mixed quasi-degenerate states having similar oscillator strength. As shown in Figure the simulated absorption spectrum of enol CURC, calculated at TD-DFT level, is only moderately sensitive to the size and type of the basis set, hence justifying the use, especially for the subsequent SH dynamics the smaller 6-31G base. On the contrary, a more marked dependency on the functional nature and type can be observed with PBE0 providing a highly red-shifted transition, while long-range corrected functionals behave coherently among them. In particular hybrid functional are also artificially lowering the energy of the np* state, due to its partial charge-transfer character which can be appreciated from the low value of the fS descriptor (~0.3) indicating a small overlap between hole and electron densities. |
610cf54f4cb479f729299249 | 4 | To further assess the effect of the functional and of the basis set on the general features of the PES we calculated the energies of triplet and singlet states at different critical point of the PES, namely Franck-Condon and S1 and T1 minima. The results are collected in ESI. Once again while the effect of the basis set appears negligible, the functional effects are more crucial. In any case the use of 6-31G basis in combination with the wB97XD exchange-correlation functional provide a coherent description. |
610cf54f4cb479f729299249 | 5 | As shown in Figure we may indeed evidence that not only the S1 state PES may cross one of the triplet (T4), but at its equilibrium region it also lays close in energy, and more importantly evolves in parallel, with the PES of T2 and T3. Such an arrangement, which is coherent with the one observed previously, may be indicative of a possible ISC, which however should not be ultrafast due to the necessity to overcome some barriers or to allow for the vibrational induced inversion between the energy of the quasi-degenerate singlet and triplet states. Note also that favourable ISC channel may exist also from S2, however due to the very low oscillator strength it is unlikely that they could play a major role in CURC photophysics. |
610cf54f4cb479f729299249 | 6 | To complete the static analysis in Figure we also report the relaxed scan profile detailing the PES along the Hydrogen transfer coordinate. Interestingly, while most of the triplet states and S2 presents a barrier, which can reach up to 0.2/0.3 eV, the process is barrierless on the S1 surface. Hence, the population of the pp* excited state should certainly be connected with the weakening of the OH bond, and the strengthening of the complementary H-bond, leading to the effective transfer. |
610cf54f4cb479f729299249 | 7 | If the H-transfer possibility is confirmed by the analysis of the PES, it is also interesting to point out that the transfer itself does not seem to alter a favourable energetic order of states potentially leading to ISC crossing. Indeed, S1 PES is crossing twice a triplet state along the transfer coordinate and the general quasi-degeneracy with the triplet manifold is maintained. Hence, while H-transfer should be considered as a clearly competitive process in the relaxation and de-excitation of enol CURC its final and global effect on the ISC rate and quantum yield appears less evident and necessitates a dynamic treatment to be properly taken into account. |
610cf54f4cb479f729299249 | 8 | The main results of the SH dynamic are summarized in Figure and in ESI and show a rather complex behaviour confirming the interplay between ISC and H-transfer. The evolution of the population of the different states (Figure ) allows to underline some rather evident effects. Due to the difference in the oscillator strength we have considered that only the S1 state will be populated after the initial excitation. However, after a few fs we may observe a redistribution of the population between S1 and S2 states, which is also due to the breaking of the symmetry and to a partial mixing of the np* and pp* character, as also seen by the strong oscillation in the respective populations. |
610cf54f4cb479f729299249 | 9 | The transfer of population to the triplet state may be observed, at very modest rates, at around 50-100 fs. Interestingly, in this early stage the population seems to be driven mostly to the higher excited T4 state, which however are never showing a substantial population, also due to the very fast internal conversion (IC) to the lower triplet states. At larger time scales the population of T1 becomes more important, especially after 500/600 fs. Interestingly the increase of the population of T1 happens at similar time scales as the sharp decrease of the S2 and a stabilization of S1 population. As a matter of fact, at around 600 fs T1 bypasses S2 and at the end of the SH dynamics (200 fs) it reaches a population of more than 20%. Note that the other triplet states never achieve significant population and that a marginal participation of the higher singlet state S3 can also be observed. This model is also confirmed by the analysis of the hops during the SH dynamics which points to a population flow from S1 to both S2 (as well as higher excited singlet states) and T4 (see ESI). Interestingly, while almost no transfer to the triplet population is observed from S2 a complex equilibrium involving singlet states up to S4 can be highlighted. On the other hand, after the initial population of T4 a very rapid interval conversion on the triplet manifold ensues leading to the sequential ultrafast population of T3, T2, and finally T1. On these bases we may also sketch out a kinetic model (see Figure ) which yields out a rate-limiting S1àT4 ISC characteristic time tISC of 3.74 ps. On the contrary the cascade of IC on the triplet manifold takes place in the sub ps-range with characteristic time tIC comprised between 100 and 200 fs. As far as the H-transfer is involved we analyse in Figure the distribution of the O-H distance d highlighted in Scheme 1 on the ensemble of the SH trajectories and its time evolution. While for all the initial conditions d peaks at around 1.8 Å, which is coherent with a classical intramolecular H-bond, even at the very first steps of the dynamic we may observe the branching of the distribution with the establishment of a cluster of trajectories presenting shorter distances of 0.9 Å, which are coherent with the formation of a novel covalent oxygen hydrogen bond. The flow of population between the two branches continues regularly along the SH population leading to an increase of the importance of the shorter distance structures which become dominant at around 600 ps. Interestingly, a continuous and bidirectional flux between the two populations can be observed all along the dynamic indicating the establishment of an equilibrium between the two tautomeric forms. Furthermore, in some cases we may observe a sudden increase of d which may reach up to 4 Å and show important oscillations. This fact can be associate to the activation of the out-of-plane rotation of the enol OH group. Finally, and as reported pictorially in Figure we may observe H-transfer both for trajectories leading to ISC and for trajectories remaining on the singlet state. Conversely, a significant breaking of the planarity can be observed for the T1 state as shown in Figure . At the 2 ps limit, around 75% of the trajectories have transferred the involved hydrogen, and the phenomenon appeared for both trajectories leading to ISC or continuing exploring the singlet manifold. |
610cf54f4cb479f729299249 | 10 | The initial equilibrium geometries for the GS, as well as S1 and T1 states of enol CURC have been obtained at DFT level using the wb97XD 27 functional together with the Pople double-z 6-31+G(d,p) basis set, coherently with what has been presented in Girardon et al. Water solvent has been implicitly considered using polarizable continuum method (PCM). In the case of the first singlet excited state TD-DFT has been used. All the calculations have been performed using the Gaussian09 suite of codes, harmonic vibrational frequencies have been obtained to assure the minimum nature of the stationary point. For benchmark purposes, the absorption spectrum has been calculated as vertical transitions from the GS using different functional namely, CAM-B3LYP, The same levels of theory have also been used to assess the energetic order of the excited states, of both singlet and triplet multiplicity, at S1 and T1 geometries. The excited states have been calculated at TD-DFT level with the Tamm-Dancoff Approximation (TDA), which allows to avoid problems due ot triplet instability and hence provides a more balmanced description of the two multifold. As discussed in the Result section it appears that the combination of 6-31G and wb97XD provides a reliable description of the PES, especially avoiding a computational burden, which will make render the subsequent SH dynamic sampling untreatable. The nature of the excited states has been identified in terms of Natural Transition Orbitals (NTO) obtained via a postprocessing of the Gaussian output with the Nancy_Ex code. To better quantify the nature of the excited states we also calculated the topological fS index which provides the spatial overlap between hole and density matrices. To preliminary assess the possibility of H-transfer and ISC crossing we also performed a relaxed scan at the 6-31G/wb97XD level along the O-H distance d (see Scheme 1 and Figure ) and we also explored the global degrees of freedom potentially leading to ISC by performing a linear interpolation of the internal coordinates to bridge T1 and S1 equilibrium geometries, using the same level of theory as for the relaxed scan. These calculations have been performed with Gaussian 09 code as well. Finally, to assess the competition between the different relaxation routes and their timescales we performed in vacuo SH dynamics using the Sharc code coupled with Orca. This strategy has already been used successfully by some of us on similar biological systems or photosensitizers. To this aim, 200 initial conditions around the Franck-Condon regions have been generated via a Wigner distribution, and the trajectories have been propagated at TD-DFT level, using the 6-31G/wb97XD level of theory, coherently with the previous benchmark. Due to their possible influence in the forthcoming dynamics we have calculated, in addition to the GS, 4 singlet excited states and 4 triplet states. SOC elements have been directly computed at each step of the dynamics with Orca, while the non-adiabatic coupling is approximated via the wavefunction overlap strategy. Considering the high difference in the oscillator strength, more than two order of magnitude, only the S1 state was initially populated. Each SH trajectory has been propagated with a time step of 0.5 fs to assure energy conservation for a total of 2 ps. The analysis of the population and distances have been performed on the ensemble of the trajectories, and the characteristic times have been obtained fitting the resulting evolution of the population to the established kinetic model. The individual trajectories have been analysed and visualized also using the VMD code. |
610cf54f4cb479f729299249 | 11 | CURC is a natural compound which can be promising for photosensitization purposes and more specifically for PDT and LAC applications. However, its complex, and environmental driven, conformational equilibrium may alter its native photophysical properties. Here we report a non-adiabatic molecular dynamic study of the behaviour of enol CURC after excitation to the bright S1 state. After having properly benchmarked the chosen level of theory, achieving a good compromise between accuracy and the need of an extended sampling, we have shown that enol CURC may lead to ISC, albeit not in an ultrafast process and is characterized by a 3.4 ps rate determining step. ISC proceeds via the initial population of the high-laying T4 state and is followed by an ultrafast IC cascade in the triplet manifold. An equilibrium between S1 and the higher excited states, especially S2, and resulting in a population redistribution can also be evidenced. |
610cf54f4cb479f729299249 | 12 | In enol CURC ISC is also in competition between the excited state Hydrogen transfer that can happen via the intramolecular H-bond relying the keto and enol group. This process is usually regarded as susceptible to quench ISC severely limiting its efficiency. Our SH dynamics have clearly shown the occurrence of the H-transfer via a dynamic and ultrafast equilibrium. However, it has also been shown that triplet population is still possible, even in presence of this competitive mechanism. On the other hand, triplet population leads to a significant deformation of CURC skeleton, in particular breaking its planarity. |
66d93ab151558a15ef1727f7 | 0 | The human Respiratory Syncytial Virus (RSV), is an enveloped negative-sense RNA virus causing worldwide highly dangerous respiratory infections in infants and elderly adults. After host-cell recognition by the RSV G surface protein, cellular penetration and infection are dependent of host-cell-viral membrane fusion mediated by RSV F homotrimer virion surface proteins . |
66d93ab151558a15ef1727f7 | 1 | Preventive new RSV vaccines and/or therapeutic anti-RSV monoclonal antibodies are available . However, no complementary anti-RSV small molecules for therapeutic use are yet clinically available . Despite the identification of numerous and potent in vitro anti-RSV fusion inhibitors, the emergence of resistance mutations together with pharmacological issues, has prevented their clinical use . New small-drugs are required to therapeutically overcome most of the resistant mutations implicated in RSV fusion . |
66d93ab151558a15ef1727f7 | 2 | The F homotrimer proteins that coat the membrane surface of RSV, start infection of host cells by promoting pH-independent fusion of cellular and virion membranes. In the cytoplasm the F homotrimers are synthesized from inactive precursor monomers of 574 amino acids. Each prefusion F monomer contains signal peptide, heptad-repeat A (HRA), heptad repeat B (HRB), transmembrane α-helix, and cytoplasmic tail (Figure ). Three monomers fold around a central top-to-bottom cavity, as shown by crystallographic 3D models (Figure ). The F homotrimers are then transported to the hostcell membrane driven by its signal peptides . |
66d93ab151558a15ef1727f7 | 3 | Furin-like endoproteases form the host cell cleave F at two sites per monomer, producing p27 peptides , and generating disulfide-linked F (F1F2) prefusion homotrimers . The efficiency of furin-cleavage is Golgi-, cell line-, RSV subtype-and time-dependent. Therefore, uncleaved, partial or total cleavage F homotrimers may be present at the membranes of RSV virions and/or of infected host-cells . Additionally, prefusion homotrimers may exist as dynamic interchanges among ± partially linked p27 and/or monomer/trimer conformations 20,17, 18 . Most crystallographic F prefusion models, however, lack p27 because of its variable conformations and absence at the postfusion conformation . |
66d93ab151558a15ef1727f7 | 4 | An extensive RSV F conformational change brings together HRA and HRB, to form one six-helices bundle stable postfusion structure (Figure ). Such conformational change precedes viral / host membrane fusion by inserting extended hydrophobic F fusion peptides into the host cell membranes. To preserve antibody epitopes for vaccine development, the F prefusion-conformation have been stabilized by including S155C and S290C mutations and eliminating p27, transmembrane and cytoplasmic domains . Such prefusion-stabilized F models have been intensively explored not only for antibody-binding studies but also for small drugs inhibiting RSV fusion and infection . |
66d93ab151558a15ef1727f7 | 5 | Co-crystallization of prefusion-stabilized RSV F with JNJ2408068, JNJ49153390, TMC353121, BTA9881, or BMS433771 have shown they all bind to similar central cavities formed by their homotrimer inner α-helices (Figure ABCD). Such bindings further stabilized the prefusion F state inhibiting the postfusion stable six-helix bundle formation and therefore subsequent viral penetration and infection. However, despite the high number of reports describing in vitro inhibitions of RSV-cell fusion (i.e., for instance by cell-cell syncytium tests), at low-nanoMolar ranges, only the GS5806 has been further explored clinically. GS5806 blocks the prefusion RSV F conformational changes and shows inhibition of infection in vitro and in animals. Nevertheless, several RSV-resistant mutations emerged during GS5806 treatment tests of RSV controlled infections in human volunteers . The need for alternative small molecular RSV inhibitor drugs including those to F resistant mutations remains. |
66d93ab151558a15ef1727f7 | 6 | To generate thousands of new F-docking candidates, the java-based DataWarrior Build Evolutionary Library (DWBEL) 2-5 fast algorithms were employed. Provided with a drug parent and its crystallographic 3D binding cavity , DWBEL co-evolutions randomly add/insert atom/bond variations into the parent molecule to generate tens of thousands of raw-children. Those children best-fitting the F cavity were then automatically selected . The DWBEL predicted affinities and cavities were then confirmed by consensus with those form the highly accurate AutoDockVina (ADV) docking program . Thousands of children targeting F with low toxicities, high specificities and lower than sub-nanoMolar affinities were predicted. Hundreds of top-children also docked to inhibitor resistant F mutant cavities. |
66d93ab151558a15ef1727f7 | 7 | Some of the limitations of the top-children predictions described here are: a) Crystallographic 3D-models contained prefusion-stabilizing mutations and excluded p27 sequences (both of which may interfere with docking), b) only single mutations could be studied, due to the limited number of mutants coding for several mutations each , and c) the high-docking affinity selection criteria has not yet known in vitro / in vivo fusion inhibition properties. Nevertheles, the topchildren identified here may be employed for in vitro experimental evaluations. |
66d93ab151558a15ef1727f7 | 8 | Drugs previously inhibiting RSV fusion in vitro (anti-F drugs) which were co-crystallized with prefusion-stabilized RSV F models (anti-F drugs) were employed for co-evolution. For that, anti-F drug *.sdf files and their corresponding crystallographic F complex *.pdb files corresponding to JNJ2408068, JNJ49153390, TMC353121, BTA9881, and BMS433771 fusion-inhibitors , were supplied to DWBEL co-evolution as 2D parents-3D cavities. The program randomly generated tens of thousands of raw-children and selected those non-toxic and best-fitting their corresponding F cavities. ADV docking of TMC353121, BTA9881, and BMS433771 confirmed thousands of non-toxic fitted-children predicting higher affinities (lower docking-scores) than their corresponding drug-parents (Figure large-open circles, inside the drug rank profiles). |
66d93ab151558a15ef1727f7 | 9 | The BTA9881-derived (BTA) top-children predicted maximal ADV affinities of ~ 0.00001 nM (Figure , red), 3-4-logs higher than those predicted from any of the other drug-parents (Figure Representative F nearby amino acids of selected BTA top-children confirmed the contacts with the 3 α-helix chains of BTA but targeting more amino acids and predicting hydrogen bonds, all of which may explain their higher affinities (Table ). Similar crossdocking of central α-helix cavities were also previously described in other viral trimers, such as those from SARS-CoV2 33 and VHSV, a fish rhabdovirus . |
66d93ab151558a15ef1727f7 | 10 | Individual amino acid mutations previously reported in RSV resistant to fusion inhibitors (anti-F fusion inhibitors) were identified from those: i) fusion inhibitors co-crystallized with RSV fusion-stabilized recombinant F, ii) sequencing of RSV isolated after in vitro co-culture with host-cells and fusion-inhibitors, iii) sequencing of F from RSV isolated from naturally-infected RSV patients, and iv) emerging on F from GS5806-inhibitor-treated RSV-infected human volunteers. |
66d93ab151558a15ef1727f7 | 11 | First, all mutations mentioned above were ADV blind-docked (60x60x60 Å grids) to ~ 200 BTA-derived top-children. The resulting ranking profiles (n= 2-3) identified two groups of mutations. The first group included most of the fusion-resistant mutations predicting sub-nanoMolar affinity ranges (Figure A, different colors), such as those targeting the top-surface, bottom-surface and some of the bottom-drug sites. Most probably their distances to the drug site could not detect by docking any of their possible allosteric interferences (Figure ). The second group included only F140I, F488I, and D489Y predicting ~100-1000fold lower affinities than those of the first group (Figure A, larger circles green, red, cyan, respectively). These 3 mutations belonged to those surrounding the drug site (Figure B, green). |
66d93ab151558a15ef1727f7 | 12 | BTA-derived ~ 1000 top-children (Figure , red) were finally selected to ADV target the bottom-site mutations with highly restricted grids (13x13x13 Å) centered at the drug-site mutations. Results were tabulated to compare X-axis with top-children affinities versus one mutation per Y-axis column (Supporting Materials / 979BTAmutations.dwar). Top-children predicting affinities < 50 nM to most resistant mutations, identified compact star-like molecules with 3 short arms extending from a central carbon (Figure , red circle) containing 6-8 ring structures in 2 slightly different scaffolds (Figure ). Further DWBEL may be required to explore for possible F140I and F488I higher docking affinities using the new DWvs6 (java's BellSoft Liberica), which included more efficient handling of larger heap-memories compared to DWvs5 (java's 8) (Figure ). Nevertheless, since some amino acid substitutions to those mutations reduced RSV replicative fitness during co-infections with wildtype RSV , such RSV variants may also have difficulties to spread into a natural infection 30 . |
66d93ab151558a15ef1727f7 | 13 | New docking molecules were predicted to fit into the RSV prefusion F stabilized central cavity including those of fusion-inhibitor resistant mutations. The new top candidates predicted low picoMolar affinities to F, low nanoMolar affinities to most fusion-inhibitor resistant mutations, low toxicity risks and high specificity to target similar trimer cavities than the crystallographic inhibitor binding. The new top candidates belong to compact 3 star-like molecules fitting the 3-fold symmetric internal RSV F cavity contacting more amino acids than the inhibitory drugs. Chances are that some of the new candidates may also interfere the conformational changes required for RSV fusion. |
66d93ab151558a15ef1727f7 | 14 | The AutoDockVina (ADV) algorithm (PyRx-0.98/1.0 package) was used to: i) quantify in ~ nM the DWBEL non-toxic fitted-children affinities, ii) identify PyMol amino acid contacts on ADV docked protein-children complexes and iii) confirm the targeted cavities on F trimers. The ADV algorithm was selected for docking because of its higher accuracies when compared to other docking programs and its most recent improvements . The PyRx/Obabel/ADV algorithms were adapted to large scale docking by home-scripts 47 . As mentioned above, the DWBEL non-toxic fittedchildren sdf files required DW mmff94s+ minimization to preserve their 2D geometries during ADV . Several ADV grids were used for docking. A wide grid of 60x60x60 Å centered to the PyMol / centerofmass surrounding most of the F homotrimers was initially employed for blind-docking. Both 20x20x20 Å and 13x13x13 Å grid, centered at 19x19x19 Å (from the F centerofmass), were employed to restrict docking to the drugsite cavities. Duplicates and/or triplicates for selected mutations were used to confirm their affinities and reduce variations between runs due to their abundant random rotatable bonds (± 2.5 Kcal/mol). After ADV docking, only the best conformer of each non-toxic fitted-children was selected for additional calculations. The output ADV docking-scores in -Kcal/mol were converted to ~ nM affinities by the formula, 10 9 *(exp (Kcal/mol/0.592) ). A, F2p27F1+TM+CT homotrimer prefusion side-view . B, F2F1+TM+CT homotrimer postfusion side-view . C, F-wt truncated homotrimer crystallographic model side-view . D, Cl bottom-view of its central cavity . Yellow cylinders, Hydrophobic α-helices containing part of the HRB heptad-repeats (~444-485 residues) Green cylinders, transmembrane α-helices (485-514). Blue cylinders, Hydrophobic α-helices containing part of the HRA heptad-repeats (~173-182 residues) . Green spheres, amino terminal amino acids of F1 and/or F2 (p27 included) Red spheres, mapping of fusion-inhibitor drugs nearby residues 140, 486-489 (drug-site) ii) Red sticks, side-chains at the bottom-surface surrounding the drug-site (L138I, L141W, G143S, V144A, T323A, I379V, V384I, D388Y, D392G, K394R, M396I, T397S, S398L, K399N, T400I, D401E, I474T, V482A, S509I). |
66d93ab151558a15ef1727f7 | 15 | 3594BTA.dwar. Data from the 3549 BTA-derived non-toxic fitted-children from Figure , were tabulated in this file. The file contains all randomly generated 3D conformer order (ID), with their corresponding 2D and 3D chemical structures, molecular weights, cLogP hydrophobicities, order number by docking score (NN) and their ADV affinity in kcal/mol and calculated ~ nM. The file can be opened in freely-available DW program ( ). |
66d83c1f12ff75c3a15d07ff | 0 | The pursuit of advanced energy storage technologies has significantly featured the role of lithium metal batteries (LMBs) owing to their superior energy density and potential for enabling long-range electric vehicles. However, the widespread adoption of LMBs is hindered by safety concerns related to liquid electrolytes, which are prone to flammability and stability issues. This has promoted the exploration of solid electrolytes as a safer alternative. The pioneering work of Wright and Armand in 1970s ignited the interest and passion for research into solid polymer electrolytes (SPEs) for rechargeable solid-state batteries. Since then, polyethylene oxide (PEO) has become a frontrunner in SPEs due to its high safety, ease of processing, low cost, and good compatibility with lithium metal, making it attractive for integration into next-generation LMBs. High molecular weight PEO offers the benefits of preventing Li dendrite formation due to its superior mechanical properties and enhancing cycling stability compared to its lower molecular weight counterpart. However, ionic conductivity of PEO at room temperature is inadequate (up to 10 -6 S cm -1 ) because its semicrystalline structure which hinders chain mobility, essential for ion transport, often requiring higher temperatures for effective application. Additionally, PEO-based solid polymer electrolytes (SPEs) are limited by a narrow electrochemical stability window (ESW < 4V vs. Li + /Li) due to the low oxidative stability of ether oxygen. These drawbacks considerably constrain their practical use in high voltage solid-state lithium metal batteries (LMBs). To overcome these issues, research has focused on modifying the PEO microstructure, including the addition of inorganic fillers, plasticizing ionic liquids, 10 copolymerization, and creating crosslinked PEO matrices. Ionic liquid (IL)-based electrolytes have gained significant interest in recent years due to their superior properties, such as low vapor pressure, nonflammability, and outstanding thermal and chemical stability. Shin et al. investigated the impact of C3mpyrTFSI ionic liquid on the ionic conductivity of PEO-based ternary polymer electrolytes, and demonstrated that the inclusion of highly conductive room temperature ILs significantly enhances ionic conductivity, surpassing the limits previous observed in IL-free electrolytes. Crosslinking, as a versatile polymerization technique, is often used to improve the mechanical property and hinder polymer crystallization of plasticized SPEs. Watanabe et al. first studied the crosslinked PEO-LiClO4 systems which showed decreased PEO crystallinity by decreasing Tg from -22 to -30°C and enhanced ionic conductivity from 10 -9 to 10 -5 S cm -1 at 30°C compared to linear PEO system (EO/Li mole ratio = 20). Joost et al. further explored ionic mobility in crosslinked PEO-based ternary SPE systems that incorporate C4mpyrTFSI ionic liquid with relatively low concentration of LiTFSI (25 mol%), demonstrating a decrease of Li + -EO interactions in the ternary system due to the additional TFSI -anions from ionic liquid that contributes to the high ionic conductivity. The crosslinking process, induced by UV irradiation, helps to form a stable network that enhances the mechanical integrity of the SPE while accommodating the beneficial properties of ILs. The effectiveness of UV-induced crosslinking in the development of polymer-plasticiser-salt ternary systems was further validated by other researchers. In recent years, inspired by the concept of high concentrated electrolyte in traditional liquid electrolytes, there is growing interest in concentrated ionic liquid (CIL) electrolytes due to their advantageous properties such as expanded electrochemical stability window, 21 high transference number, high rate capability, and enhanced cycling stability. Recent work from Pal et al. identified the use of ether-aided concentrated LiFSI-C3mpyrFSI ionic liquid electrolyte as an effective approach to enhance ion transport, interfacial stability, and support fast charging. These desirable properties make them promising candidates to work with the crosslinking strategy for developing high-performance PEO-based SPEs. Thus, one promising approach to form a robust PEO-based SPE is to combine crosslinking with the incorporation of CILs, which shows great potential in further suppressing PEO crystallinity, improving ionic conductivity, expanding electrochemical stability while maintaining safety and mechanical integrity. |
66d83c1f12ff75c3a15d07ff | 1 | In this work, we report a novel UV-crosslinked polymer-in-concentrated ionic liquid system (PCIL) based on PEO, lithium bis(fluorosulfonyl)imide (LiFSI) and N-propyl-Nmethylpyrrolidinium bis(fluorosulfonyl)imide (C3mpyrFSI) ionic liquid (Li + :IL molar ratio = 1:1). By focusing on the enhancement of ion transport, electrochemical stability, and cell performance, we seek to tackle a critical gap in the current understanding of PEO-based SPEs. The obtained PCIL-SPEs achieved high ambient temperature ionic conductivity, wide electrochemical stability window, stable and reversible cycling toward Li metal electrodes and promising solid-state LMB performance. Thus, by incorporating concentrated ionic liquids with UV-crosslinking to form a robust SPE, we shed light on the potential of PEO-based SPEs to meet the rigorous demands of high-performing solid-state LMBs and pave the way for their practical application in future advanced energy storage systems. |
66d83c1f12ff75c3a15d07ff | 2 | The polymer-in-concentrated ionic liquid SPEs (PCIL-SPEs) were prepared using a solvent casting technique within an Ar-filled glovebox. The preparation steps are shown in Figure . First, predetermined amounts of LiFSI salt and C3mpyrFSI IL (Li + :IL molar ratio = 1:1) were mixed and dissolved overnight at room temperature. Then the obtained solutions were mixed with various amount of PEO in acetonitrile with EO:Li + :IL molar ratios equal to 10:1:1. The obtained solutions were then cast into silicon moulds and rested for 48 hours to ensure slow evaporation of solvent. The obtained free-standing solid membranes were then sandwiched between two Mylar foils and hot-pressed (70°C, 2 MPa for 10 mins) to obtain homogenous membranes with a thickness of 150 ± 10μm. Crosslinked PCIL-SPEs were prepared by using benzophenone (BP, 5 wt% of PEO) as photo-initiator and crosslinked in a TBK 905 UV curing box (power 200W) for 3 mins right after hot-pressing. The prepared SPEs were then transferred to the Buchi oven for a secondary vacuum drying process (50°C for 24 h) before further characterization. All prepared SPEs were kept in argon-filled glovebox. |
66d83c1f12ff75c3a15d07ff | 3 | Ionic conductivity (σ) of the designed SPEs was measured by electrochemical impedance spectroscopy (EIS) analysis (MTZ-35 impedance analyser, Biologic). The measurements were performed over the temperature range of 30-50 °C in frequency from 0.1 Hz to 1 MHz. A SS|PCIL-SPEs|SS (stainless-steel) coin cell configuration was used. The bulk resistance was calculated from the impedance curve at high frequency, and the ionic conductivity was calculated by the equation: |
66d83c1f12ff75c3a15d07ff | 4 | Raman spectra were obtained using a Renishaw Invia microscope with a laser generating 633 nm (red) light delivering 13.0 mW at the sample. The laser power was set to 10% with the exposure time of 60s and one accumulation. All measurements were done under room temperature in a sealed homebuilt sample holder with a quartz window in the range of 100 to 3200 cm -1 . Baseline corrections and normalization were applied to the spectra. The Voigt function was used to deconvolute bands into constituent peaks. |
66d83c1f12ff75c3a15d07ff | 5 | Thermal gravimetric analysis (TGA) was performed using a Mettler Toledo TGA STAR instrument to understand the thermal stability of the SPEs. with a sample loading of 10 mg. Mass loss was recorded over 30-600 °C with a 10 °C min -1 heating rate under constant N2 flow. Sample loading was kept approximately 10 mg. |
66d83c1f12ff75c3a15d07ff | 6 | Scanning electron microscope (SEM) analysis was carried out for the pristine SPEs as well as cycled cells. A coin cell disassembly unit (Hohsen) was used inside a glove-box to disassemble the cycled cells. The JSM-IT300 SEM from JEOL (Japan) was used to study the surface and cross-section morphology with a 10 kV acceleration voltage. |
66d83c1f12ff75c3a15d07ff | 7 | The LMO electrode was prepared by mixing LMO powders, Super C65 and PVDF (80:10:10 wt.%) in NMP. The casting and drying methods for the LMO cathodes were as previously described. The resulting LMO electrodes have a loading of 7 mg cm -2 and an area capacity of 1 mA cm -2 . On the other hand, the LFP electrode from Custom Cells has a similar areal capacity of 1 mA cm -2 with ~6 mg cm -2 active material loading. The electrodes were then cut into 8 mm diameter discs before coin cell assembly. |
66d83c1f12ff75c3a15d07ff | 8 | For all the battery measurements, lithium metal (100 µm, Gelon) was used as anode material. Battery cells were then assembled using a Li metal anode, a free-standing designed SPE membrane, and the prepared LFP/LMO cathodes. Electrochemical measurements were performed using a multi-channel potentiostat VMP3 (BioLogic) with all measurements conducted in CR2032 coin cells (Hohsen Corp.). Linear sweep voltammetry (LSV) experiments were performed using a Li||SS (stainless steel) cell at 50 °C. The scans ran from the open-circuit voltage (OCV) up to 6 V (vs. Li/Li + ) at a scan rate of 0.1 mV s -1 . The galvanostatic cycling for Li||Li, LFP||Li, and LMO||Li cells were also carried out. Cell cycling was performed at 50 °C, with voltage limits set to 2.5-4 V vs. Li + /Li for LFP||Li cells and 3-4.3 V vs. Li + /Li for LMO||Li cells. Two formation cycles were used before long-term cycling unless mentioned otherwise. The reproducibility of the results has been checked by multiple tests. Both linear and crosslinked PCIL-SPEs were synthesized from homogeneous mixtures of PEO, LiFSI and C3mpyrFSI (EO:Li:IL molar ratio=10:1:1) by solvent casting and UVcrosslinking processes, namely SPE10:1:1 and SPE10:1:1-CL respectively (Figure ). Thermal gravimetric analysis (TGA) of both SPE10:1:1 and SPE10:1:1-CL confirms the thermal stability up to 220 °C which is sufficiently high for the safe operation of solid-state LMB (Figure , Support Information). |
66d83c1f12ff75c3a15d07ff | 9 | The effect of UV-crosslinking on the phase behaviour, ionic conductivity, and morphology of the prepared PCIL-SPEs are shown in Figure and Figure . As illustrated in Figure , both the melting point and enthalpy decrease dramatically after introducing CIL (Li:IL molar ratio = 1:1) to PEO, with the melting transition temperature drops from 65 °C of pure PEO to 38 °C of SPE10:1:1, and the Tg broadens significantly, suggesting a more complex heterogenous system. Notably, a small peak at approximately -20°C appears in SPE10:1:1, attributed to a phase transition in the C3mpyrFSI/LiFSI mixture, suggesting limits in IL solubility and potential phase separation. Crosslinking SPE10:1:1 to create SPE10:1:1-CL results in the disappearance of both the melting and phase transition peaks, with a new, pronounced Tg appearing at -74 °C. This suggests that crosslinking leads to a highly amorphous structure and enhances the uptake of IL in PEO matrix. In addition, the presence of a single Tg in the crosslinked materials indicates a good miscibility between IL and the PEO matrices. To summarize, crosslinking effectively hinders crystallization of PEO and PEO-LiFSI complexes and increases the IL uptake in the crosslinked network, which is crucial for achieving high ionic conductivity due to the resulting high amorphicity of the SPEs. |
66d83c1f12ff75c3a15d07ff | 10 | illustrates the effect of IL content and UV-crosslinking on the ionic conductivity of the prepared SPEs. Compared to systems without IL (SPE10:1), adding IL significantly increases the ionic conductivity, with more than three orders of magnitude improvement from 1.5 x 10 -7 S cm -1 for IL-free SPE10:1 to 4 x10 -4 S cm -1 for SPE10:1:1 at 30°C as reported in our previous study, suggesting the substantial role of IL in enhancing ionic conductivity. After crosslinking, SPE10:1:1-CL shows a slight decrease in conductivity, consistent with an overall high salt content system. Crosslinking generally increases the formation of amorphous phases that facilitate conductivity in systems with high polymer content. However, it can also restrict polymer chain movement, hindering ion transport. This effect is particularly noticeable in systems with low polymer content that are already highly amorphous and conductive, such as SPE10:1:1. Overall, while crosslinking affects ionic conductivity by restricting chain mobility, its downside impact is mitigated by the reduced SPE crystallinity and increased IL uptake. Furthermore, the impact of crosslinking on ionic conductivity is much less pronounced compared to the influence of IL concentration. |
66d83c1f12ff75c3a15d07ff | 11 | The surface morphology of the prepared SPEs was characterized by SEM analysis as shown in Figure . Compared to SPE10:1:1, the crosslinked SPE10:1:1-CL shows a dense, smooth surface with a high degree of amorphous nature. The nonuniform and hardly homogeneous texture of SPE10:1:1 transformed into a finer wrinkled feature after crosslinking, which might be due to the increased encompassing of lithium salt and IL, leading to a more homogenous and mechanically robust network. 13,36 Raman spectroscopy was performed on linear SPE10:1:1 and crosslinked SPE10:1:1-CL to investigate and reveal the effect of crosslinking on cation-anion interactions by identifying the states of FSI -in the Raman shift between 700 and 800 cm -1 (full Raman spectra is given in Figure , Supporting Information). The -SNS stretching mode of FSI -in the PCIL- SPEs was deconvoluted into three constituent peaks: free FSI -at 721-723 cm -1 , Li + -FSI -ion pairs at 725 cm -1 , and complexed FSI -anions at 730-735 cm -1 (aggregated ion clusters). As shown in Figure , the SPE10:1:1 (Fig. ) has a low proportion of the free FSI -state (28%) and a high proportion of Li + -FSI -ion pairs (46%). In contrast, the crosslinked SPE10:1:1-CL (Fig. ) is rich in free FSI -anions (53%) and contains very low proportion of Li + -FSI -ion pairs (10%), likely ascribed to the crosslinking of the EO chains that form a crosslinked network with high dissociation ability. The resulting high solvation environment may lead to an increased Li + -EO coordination that can lower the highest occupied molecular orbital (HOMO) level of PEO and therefore benefit the electrochemical stability of the SPE. The SPE10:1:1-CL also exhibited higher proportion of complexed FSI - anions, indicating the presence of aggregated ion clusters. The crosslinked polymer network may have helped to segregate the large, aggregated domains into more uniformly distributed ion clusters to facilitate Li + transport. Therefore, we believe that crosslinking results in enhanced Li + solvation and more uniformly distributed ion clusters which would account for the high ionic conductivity and the potential high electrochemical stability of the crosslinked SPE10:1:1-CL. The electrochemical stability window (ESW) of the linear SPE10:1:1 and crosslinked SPE10:1:1-CL was evaluated in a Li||SS cell by LSV measurements. As shown in Figure , the linear SPE10:1:1 begins to oxidize slowly above 4V vs. Li + /Li, a common behaviour in PEO-based SPEs. In contrast, the crosslinked SPE10:1:1-CL showed high electrochemical stability up to 4.9V vs. Li + /Li (5µA cm -2 current density was used as the cut-off of a current rise ). This indicates its excellent oxidation stability and suitability for use with highenergy 4V-classs electrodes. The enhanced electrochemical stability of SPE10:1:1-CL can be attributed to its crosslinked network, which limits the mobility of linear EO chains and therefore reduces their likelihood of oxidation at the cathode surface as illustrated in the inserts in Figure . Furthermore, the crosslinking changes the coordination environment enhancing more Li + -EO coordination and promoting the formation of FSI --rich ion clusters, as evidenced by Raman analysis. These coordinated ions and surrounding ion clusters likely form a protective layer that prevents further oxidation of EO, synergistically contributing to the enhanced oxidation stability. |
66d83c1f12ff75c3a15d07ff | 12 | The ability of the crosslinked SPE10:1:1-CL to support lithium plating-stripping at various current densities was assessed using Li||Li symmetrical cells at 50 °C, with current densities ranging from 0.05 up to 0.4 mA cm -2 . The cells showed overpotentials of 20 mV and 125 mV at 0.05 and 0.4 mA cm -2 respectively. As illustrated in Figure , the SPE can sustain plating-stripping polarization steps throughout the entire polarization period up to 0.4 mA cm -2 , demonstrating stable interfacial properties, thus compatibility with the lithium metal electrodes, and good cycling performance for a solid-state polymer electrolyte. |
66d83c1f12ff75c3a15d07ff | 13 | As shown in Figure , the long-term cycling of the Li|SPE10:1:1-CL|Li cell cycled at 0.1 mA cm -2 for 1 hour polarisation demonstrated very stable performance, maintaining an overpotential of 60 mV over 130 cycles at 50°C, with no evidence of short circuit. This indicates that crosslinked SPE10:1:1-CL is highly compatible with the Li metal electrode. The interfacial resistance decreases after 130 cycles (shown in Figure ), suggesting the formation of an ion-conductive and stable SEI layer at the electrode/electrolyte interface. In contrast, the Li|SPE10:1:1|Li cell exhibits a steady increase in overpotential and a short circuit after just 60 cycles. The impedance spectra also showed an increase in interfacial resistance upon successive cycling as seen in Figure . |
66d83c1f12ff75c3a15d07ff | 14 | The top-and cross-section SEM images of the Li electrodes from both cells after cycling (Li|SPE10:1:1|Li cell for 60 cycles, Li|SPE10:1:1-CL|Li cell for 130 cycles) are shown in Figure -f. In the Li|SPE10:1:1|Li cell, the electrode surface displays large cracks and mossy Li formations (Fig. ), whereas the Li|SPE10:1:1-CL|Li cell shows a smooth and dense Li metal surface (Fig. ), indicating a uniform Li plating-stripping process. The cross-section images revealed a thick layer of dendritic/dead Li at the interface of the Li|SPE10:1:1|Li cell (Fig. ), in contrast to a dense and uniform interface observed in the Li|SPE10:1:1-CL|Li cell (Fig. ). This confirmed the formation of a smooth and compact interphase and SEI layer in the cell with crosslinked SPE10:1:1-CL, contributing to its good cycling performance. UVcrosslinking appears to play a crucial role in suppressing Li-dendrite formation and facilitating facile and reversible Li plating-stripping. The prepared SPE10:1:1 and SPE10:1:1-CL membranes were first incorporated into high loading (6 mg cm -2 ) LFP||Li cells to investigate their performance in solid-state LMBs. Figure presents the prolonged cycling results for both LFP|SPE10:1:1|Li and LFP|SPE10:1:1-CL|Li cells at 0.1 C (0.05 mA cm -2 current density). The LFP|SPE10:1:1-CL|Li cell incorporating a crosslinked membrane exhibits an initial discharge capacity of 160 mAh g -1 , maintaining 90% of its capacity after 100 cycles, achieving excellent coulombic efficiency of nearly 99.8% throughout cycling. In contrast, the LFP|SPE10:1:1|Li cell with non-crosslinked membrane exhibits an initial capacity of 158 mAh g -1 but experienced a short circuit after 42 cycles. Figure shows the charge-discharge profiles of the LFP|SPE10:1:1-CL|Li cell at different cycle numbers, noting that while the discharge capacity decreased, the charge and discharge curves remained very stable over 100 cycles. A rate capability test, detailed in Figure , was conducted at current densities of 0.05 to 1 mA cm -2 (1C equals 1 mA cm -2 ). Compared to the LFP|SPE10:1:1|Li cells, the LFP|SPE10:1:1-CL|Li cells exhibited improved rate capability, maintaining nearly full capacity at 0.05C. Overall, the battery delivered high discharge capacity ranging from 170-59 mAh g -1 at 0.05 to 1C. Figure presents charge-discharge profiles of the LFP|SPE10:1:1-CL|Li cell at various C-rates within a voltage range of 2.5-4.0 V vs. Li + /Li, showing stable potential plateaus and increasing overpotential with higher C rates. AC impedance profiles (Figure ) showed a steady decrease in interfacial resistance from 0.05C to 0.2C as the crosslinked membrane adapted itself with the electrode, followed by a slight increase at 0.5C and 1C rate due to interface stability. The improved rate performance and cycling stability of the LFP|SPE10:1:1-CL|Li cell demonstrate the advantages of the crosslinked SPE10:1:1-CL, providing superior electrochemical stability at 50°C. |
66d83c1f12ff75c3a15d07ff | 15 | Given the high stability at anodic voltage observed from LSV tests, LMO||Li cells were explored using the crosslinked and non-crosslinked PCIL-SPEs. Figure illustrates the cycling performance of both the LMO|SPE10:1:1|Li and LMO|SPE10:1:1-CL|Li cells at 0.2C (0.1 mA cm -2 current density) in the voltage range of 3-4.3V vs. Li + /Li. The LMO|SPE10:1:1-CL|Li cell showed an initial discharge capacity of 91.4 mAh g -1 , maintaining 93% of its capacity after 50 cycles, whereas the LMO|SPE10:1:1|Li cell experienced 25% discharge capacity loss after the same number of cycles. Figure presents the corresponding charge-discharge profiles of the LMO|SPE10:1:1-CL|Li cell at different cycle numbers, showing stable and relatively small over potential throughout cycling. To evaluate the effect of crosslinking on fast charge/discharge capability, rate-capability tests were performed and compared across current densities from 0.05 to 1 mA cm -2 (shown in Figure ). The LMO|SPE10:1:1-CL|Li cell exhibited full capacity of 152 mAh g -1 at 0.1 mA cm -2 and a high capacity of 145 mAh g -1 at 0.1 mA cm -2 . The cell maintained 45% and 32% of its capacity at 1 mA cm -2 and 2 mA cm -2 (2C), respectively. Upon reducing the current density back to 0.1 mA cm -2 , the capacity was restored to 139 mAh g -1 . The voltage profiles of the LMO|SPE10:1:1-CL|Li cell at different C rate are given in Figure . |
66d83c1f12ff75c3a15d07ff | 16 | Finally, compared with other types of PEO-based SPEs reported in the literature, the promising performances of the developed crosslinked SPE10:1:1-CL are highlighted (as shown in Table ). It not only shows excellent cycling performance with high loading LFP cathodes, but also surprisingly good compatibility with very high loading, high energy 4V-class LMO cathodes, which is a remarkable result within the PEO-based SPE family due to the low oxidative stability of PEO. Our findings collectively highlight the synergistic benefits of concentrated ionic liquid and the crosslinking approach in the designed PCIL-SPEs. The superior rate performance and prolonged cycling stability observed in the crosslinked PCIL-SPEs suggest they are a promising avenue for their deployment in high energy density solid-state LMBs. |
66d83c1f12ff75c3a15d07ff | 17 | In this study, we report a newly designed crosslinked PEO-in-concentrated ionic liquid (CIL) SPEs for use in high energy solid-state LMBs. The proper combination of using CIL and UV-crosslinking shows synergistic advantages in reducing PEO crystallinity, increasing ionic conductivity, and enhancing the electrochemical stability of the designed SPEs. These innovations yield crosslinked SPE10:1:1-CL with high ambient temperature ionic conductivity of 4 x10 -4 S cm -1 and a wide electrochemical stability window reaching up to 4.9 V vs. Li + /Li. Raman analysis confirms that these promising properties are ascribed to enhanced Li + solvation and increased Li + -EO coordination after crosslinking, which significantly enhances SPE oxidation stability. The crosslinked PCIL-SPEs demonstrated stable Li symmetric plating/stripping cycling and remarkable electrochemical performance (almost full capacity, excellent coulombic efficiency, and stable cycling even at high current regimes) with both high loading LFP and, particularly, high energy LMO cathodes, highlighting their strong potential for practical applications in high energy density solidstate LMBs. This study not only advances the development of PEO-based SPEs but also underscores their potential to impact the broader field of sustainable and high-performance energy storage solutions. |
65bd313466c1381729d3fa2c | 0 | The synthesis and investigation of the reactivity of molecules detected in interstellar environments are crucial for comprehending the formation of life-related molecules from simpler ones. Additionally, exploring their reactivity can offer valuable insights for the radio-astronomy detection of related compounds. (Hetero)cumulenes and nitrilecontaining molecules have been widely identified in interstellar space. The quest for cyanoketene (NCCHCO, Scheme 1a) has garnered significant attention due to its simplicity and its incorporation of the four fundamental elements for life-related molecules. Scheme 1: (a) Cyanoketene, and resonance structures and geometries of cyanoketenate. (b) Prior literature on formation of anionic ketenates from CO. (c) This work on the synthesis of [NCCCO] -cyanoketenate anion. |
65bd313466c1381729d3fa2c | 1 | The deprotonated cyanoketenate [NCCCO] - is isoelectronic with carbon suboxide (O=C=C=C=O), which has been widely studied as the structure and bonding situation in this molecule is ambiguous between a linear heterocumulene structure or a bent motif. Despite numerous attempts by chemists to generate cyanoketene and the cyanoketenate anion using various techniques, the unusual connectivity and reactive nature of these molecules have posed challenges. Consequently, scientists have only succeeded in trapping them in low-temperature matrices or detecting them through mass spectrometry, achieved by subjecting precursors to heat and irradiation. Therefore, a synthetically approachable methodology for the generation of cyanoketene and cyanoketenate is highly desired. |
65bd313466c1381729d3fa2c | 2 | The utilization of carbon monoxide as a C1 feedstock dates back to 1834 when Liebig demonstrated the reductive homologation of CO with molten potassium. Several industrial processes exploit the functionalization of CO on very large scales, such as the Fischer-Tropsch and Oxo process for the conversion of synthesis gas (syngas, a mixture of CO and H2) into hydrocarbons and aldehydes; and the Monsanto and Cativa processes that convert methanol into acetic acid with CO using rhodium and iridium catalysts, respectively. Although this field of chemistry was once dominated by transition metalcatalyzed reactions, there has been growing interest in the functionalization of CO using main-group compounds. In our own recent work, we have explored the activation of CO with alkali metal salts of reactive nucleophiles. Potassium di-tertbutylphosphide and benzyl potassium both react with CO to yield intermediates with carbene character. Lithium dicyclohexylamide enabled transition metalfree Fischer-Tropsch chemistry, and various alkali metal silylamides could readily convert CO into cyanide (CN -) and isocyanides (RNC). Most pertinently to the current work, we showed that a dilithiomethane derivative featuring two phosphine sulfide substituents could react with CO, and this intermediate was oxidized with N2O to afford an anionic ketenate (A, Scheme 1b). This ketenate could be heated to promote cyclo-trimerization and the formation of a hexa-functionalized benzene. The Gessner group subsequently developed an elegant synthesis of anionic ketenates by reacting CO with an alkali metal ylid derivative (B), and demonstrated that this ketenate moiety could be transferred to a range of electrophiles. This process was explored in more detail to enable selectivity control of the ketenate formation. The Liu group also developed a synthesis of a (phosphino)ketenate by reaction of the corresponding (phosphino)diazomethyl anion and CO (C). In the present study, we explore the functionalization of CO targeting the first large-scale synthesis of the fundamental [NCCCO] -anion as the [K(18-crown-6)] + salt (Scheme 1c). We probe its solidstate and electronic structure, explore its reactivity with respect to electrophiles, and synthesize the parent cyanoketene stabilized by a carbene. We note that shortly before our submission of this manuscript, a pre-print from the Gessner group appeared on ChemRxiv detailing the synthesis of the [NCCCO] - anion and reactivity towards E-H bonds (E = O, N, S), CO2 and a SO2 adduct. |
65bd313466c1381729d3fa2c | 3 | Freshly prepared potassium bis(trimethylsilyl)amide (K[N(SiMes3)2], KHMDS) was added to a cloudy suspension of (triphenylphosphoranylidene) acetonitrile to generate the anionic ylid that was first prepared as the sodium salt by Bestmann and Schmidt in 1987. Addition of 18-crown-6 and CO to this mixture, by analogy with the pioneering work from the Gessner group, led to loss of triphenylphosphine (detected by 31 P NMR spectroscopy) and clean formation of the target compound [K(18-crown-6)][NCCCO] |
65bd313466c1381729d3fa2c | 4 | (1) after purification in an isolated yield of 93%. The 13 C NMR spectrum of 1 shows highly distinctive resonances at 120.1, 115.9 and -13.2 ppm. This latter resonance with a negative chemical shift corresponds to the central carbon in the [NCCCO] -anion, and is significantly upfield shifted from the analogous signals in the aforementioned phosphorus-bound ketenate species (A, B, C, Scheme 1b) that resonate at 5.8, 2.4 and 36.2 ppm, respectively. The IR spectrum of 1 showed an absorption at 2176 cm -1 that corresponds to nitrile stretchν(CN), the carbonyl absorption band ν(CO) is found at 1462 cm -1 , while the ν(CCO) andν(CCN) are found at 2120 and 2242 cm -1 , respectively. However, it should be noted that due to the highly delocalized nature of the anion in 1, all the vibrational modes are highly concerted and not localized on one particular bond. In comparison, the ν(CCO) absorption band is found at 2086 and 2047 cm -1 in B and C, respectively. The observed red shifting of ν(CCO) in 1 presumably results from the reduction of bond order due to delocalization. |
65bd313466c1381729d3fa2c | 5 | Single crystals of 1 suitable for single crystal X-ray diffraction were grown by slow diffusion of either pentane or diethylether into an acetonitrile solution of 1 at -20 °C and confirmed the target structure (Figure ). The solid-state structure shows onedimensional coordination polymers of [K(18-crown-6)] + cations capping each end of the [NCCCO] -anion (Figure ). 1 crystallizes in the monoclinic space group P21/c, and the central carbon C2 sits on an inversion centre, so only half an equivalent of 1 makes up the asymmetric unit. This feature enforces linearity on the C1-C2-C3 moiety although the thermal ellipsoids for the NCCCO moiety suggest that the linearity may be an artifact of crystal packing. We note that similar deviation from linearity was also suggested for the isoelectronic carbon suboxide, where irregular thermal ellipsoids for the OCCCO unit were also seen. The end-for-end disorder precludes a detailed discussion of the bond metrics, and thus the structure of 1 has been further probed using computational methods. ), N1-C1-C2: 171.5( ), C1-C2-C3: 180.0 (C2 on inversion centre), C2-C3-O3: 173.9(5). Structures generated through -x,-y,-z symmetry element. DFT calculations performed at the M06-2X-D3(benzene)/Def2-TZVP//M06-2X-D3/Def2-SVPP level of theory were used to probe the structure of 1 further (see SI for full details). The optimized structure of the free anion was shown to have a linear geometry (Figure ). This is in agreement with observations for the isoelectronic system of carbon suboxide, in which only high levels of computational theory could replicate the non-linear structure that was observed in the solid state. Inclusion of the K counterion in the calculations result in a contraction of the bond angles around the carbon atoms. Due to the polymeric network observed in the solid state, structures with the K bound through the C-N bond (1') and the C-O bond (1'') separately were optimized (Figure ). 1'' is marginally higher in energy than 1', but within the error of the computational method (ΔE = +1.1 kcal/mol), in accord with the observation of the polymeric network. In both cases the coordination of the K ion results in a considerable contraction of the central C1-C2-C3 angle to 149.2° (1') and 166.3° (1''), and of the angle around the carbon atom involved in K coordination to 166.9° (N-C1-C2 in 1') and 167.8° (C2-C3-O in 1''). This produces a W-shaped geometry, reminiscent of that observed with carbon suboxide, again highlighting the similarity between these isoelectronic species. These deviations from linearity are more pronounced than those observed in the solid-state structure; however, interactions with the cation and packing forces in the solid state are thought to override the small barriers between these two computed bonding geometries. The electronic structure of 1 was explored using NBO analysis performed at the M06-2X/Def2-TZVP level of theory (see SI for full details). In all the calculated structures the NPA charges alternate across the structure in the format N δ--C δ+ -C δ--C δ+ -O δ-. In the free anion the most negative charge is calculated to reside on the oxygen atom, but this value is close to that of the central carbon atom. Inclusion of the K counterion changes this, as the NPA charges in 1' for C2 and O are -0.80 and -0.54, respectively, and in 1'' they are -0.61 and -0.71, respectively. As these two structures represent the two extremes of the K coordination an average of these charges could represent the experimental observations that would occur due to the polymeric framework. The average NPA charges on C2 and O are therefore -0.71 and -0.63, respectively, which suggests that C2 could be the site of nucleophilic reactivity in 1. |
65bd313466c1381729d3fa2c | 6 | The stability of this unusual unsaturated molecule is undoubtedly due to its negative charge, which confers stability with respect to dimerization or oligomerization by nature of electrostatic repulsion. This draws parallels with other fundamental unsaturated main-group molecules such as [PCO] -, where the reactive P-C multiple bond is stabilized with respect to dimerization by the same electrostatic repulsion, instead of requiring kinetic stabilization from bulky substituents such as those seen in neutral phosphaalkynes. Compound 1 is stable under inert conditions and can be obtained on a multigram scale in excellent yield. This prompted us to probe various aspects of its reactivity. Our first target was the parent cyanoketene. Diffusion of HCl gas into a benzene solution of 1 at 10 °C resulted in formation of a cloudy yellow suspension, and after work-up the product was obtained as 2 (Scheme 2). This species was confirmed by single crystal X-ray diffraction to be a dimer (Figure ) comprising a four-membered core presumed to form from a head-to-tail [2+2] cycloaddition reaction of the parent cyanoketene and another equivalent of 1. In the corresponding reaction of 1 with a more sterically demanding electrophile Ph3SiCl, the product (3) also contained a monoanionic four-membered core (Figure ). Interestingly, a similar dimeric product was obtained in the reaction of 1 with CO2 at 10 °C. The resulting pale yellow cloudy suspension afforded the product 4 in a 76% isolated yield (Figure ). This species contains two [K(18-crown-6)] + cations and a dianionic component, where the negative charges are delocalized in OCCCO moiety across the fourmembered core (as in 2 and 3), and the carboxylate fragment. |
65bd313466c1381729d3fa2c | 7 | The observed dimerization of the parent cyanoketene suggests that this species is amphoteric. Noting that strong donors such as N-heterocyclic carbenes (NHC) have been exploited to stabilize and isolate a wide range of highly reactive main-group compounds, we probed the corresponding reactivity of 1. To this end, 1 was treated with 1,3bis (2,4,6-trimethylphenyl) imidazolium chloride (IMes•HCl), which acted as a proton source and also generated the free NHC 1,3-bis-(2,4,6trimethylphenyl) imidazol-2-ylidene (IMes) that could trap the cyanoketene in situ and generate the carbene-stabilized cyanoketene (5) in a 46% yield. The yield could be improved by employing KHMDS as a base and proton-shuttle. In this case, the reaction is thought to proceed via initial deprotonation of IMes•HCl to afford IMes and HN(SiMe3)2. NHC attack of the electrophilic carbonyl position of 1 increases the basicity of C2 of the ketenate, enabling deprotonation of HN(SiMe3)2. The enhancement of the Bronsted basicity of the anion by the NHC is clear as 1 is formed in the presence of HN(SiMe3)2 and does not effect deprotonation on its own. Thermal ellipsoids at 50%, most hydrogen atoms are omitted for clarity. Atom colours: C (grey), O (red), N (blue), K (cyan), Si (orange), H (white). Selected bond metrics can be found in the SI. The formulation of 5 as an IMes carbene adduct with cyanoketene was confirmed through an X-ray diffraction experiment. Orange crystals suitable for X-ray diffraction were grown by slow diffusion of Et2O into an acetonitrile solution of 5 (Figure ). In the imidazolium-2-enolate, the C4-C3 bond distance was determined to be 1.512(2) Å, while the C3-O1, C3-C2, and C2-C1 distances were found to be 1.2552 (19), 1.381(2), and 1.410(2) Å, respectively. Notably, C2 and C3 were observed to be planar, with the sums of surrounding bond angles totaling 360°. These observations are indicative of charge delocalization from IMes to the cyanoketene C(O)-C-C fragment. In this arrangement, the oxygen atom and nitrile group are positioned in a cis-orientation on the C=C double bond. These structural parameters closely resemble those observed in the SIMes adduct with diphenyl ketene, as reported by Delaude in their mechanistic study of the NHC-catalyzed Staudinger reaction involving diphenyl ketene and imines. This work shows that the σ-donating NHC enables the interception of the parent cyanoketene and precludes dimerization. |
65bd313466c1381729d3fa2c | 8 | In conclusion, the cyanoketenate anion, [NCCCO] -has been synthesized and characterized. The compound can be isolated as the [K(18-crown-6)] + salt 1 on a multigram scale in over 90% yield from commercially available reagents. Reactions of 1 with the electrophiles HCl, Ph3SiCl or CO2 give related structures in which the cyanoketenate fragment dimerizes to give a four-membered ring (products 2, 3 and 4, respectively). On the other, the parent cyanoketene is stabilized by reaction of 1 with the nucleophilic NHC IMes (5) allowing the inception of the parent cyanoketene. Work is ongoing in our laboratories to explore the reactivity of the fundamental cyanoketenate anion as a building block to more complex molecular architectures. |
67b722916dde43c908b286d4 | 0 | Ligand-engineered nanoparticles (NPs) have emerged as promising platforms for many applications, including biomedical imaging, sensing, and drug delivery. The surface chemistry of NPs is of great importance, as it directly dictates not only the colloidal stability of NPs but also their interaction with the environment. Multivalency (multiple receptors per particle) has been proven to be a powerful strategy for targeting. However, counter-intuitively, more targeting ligands are not always beneficial. For example, overfunctionalization of the surface can result in a high consumption of receptors per NP, causing lower cellular uptake since there are fewer receptors for other NPs to bind. Overfunctionalization can also give rise to steric clashes between the targeting groups, reducing their ability to bind to receptors. Therefore, there is often an optimal targeting ligand density to achieve the highest cell binding and cellular uptake. Gold NPs with one antibody per particle have even been reported to have better tumor accumulation than those with two antibodies, demonstrating the need for precise control of surface functionalization. Despite its clear importance for nanomedicine, fine-tuning of surface chemistry and quantification of surface functional groups have rarely been reported. |
67b722916dde43c908b286d4 | 1 | The azide-alkyne cycloaddition is increasingly popular as an effective strategy for stoichiometric bioconjugation. It offers several advantages over other bioconjugation techniques due to its high orthogonality, biocompatibility, high yields, and rapid kinetics. Metalfree strain-promoted alkyne-azide cycloaddition (SPAAC) has been shown to be an efficient and safe alternative to Cu(I)-catalyzed alkyne-azide cycloaddition (CuAAC), where the final conjugates often contained cytotoxic copper. Both CuAAC and SPAAC have been already exploited on various NPs systems such as quantum dots, as well as polymeric, metallic and metal oxide NPs. The quantification of the dibenzocyclooctyne (DBCO) moiety, widely used as the alkyne in SPAAC, is rather straightforward because of the presence of an absorbance band in the UV range. Nevertheless, a quantitative approach to surface functionalization with azides is often missing. A large portion of the published work still opts for full surface coverage with azide-carrying groups or, if not, rarely reports the actual number of azides per particle. |
67b722916dde43c908b286d4 | 2 | However, 100% coverage of the surface with functional ligands for click reactions could add additional costs to the production of the NPs, and also influences the colloidal stability of the NPs. For example, it was shown that the fraction of azide-carrying ligands must be kept below 15% to ensure good dispersibility in water over a long time. Low reactivity of surface groups due to steric hindrance comes as another limitation. For instance, gold nanoparticles could only accommodate 8 molecules of fluorescent dye despite having 117 azide groups on the surface. Of the few reports that do quantify the azide ligand density, the azide quantification is typically based on photo-or fluorometric detection of a dye label. Particles with a certain number of azide groups are left to react with the DBCO derivative of a fluorescent dye, followed by the purification of the labeled nanoparticles and separate analysis of the nanoparticles and/or supernatant to determine the amount of bound or unbound dye. More sophisticated variations of this method include the use of cleavable reagents that are first clicked to the surface. After purification, the dye is released into the solvent by cleaving another weak point in the molecule (e.g., hydrolysis of an amide). The obvious disadvantage of relying on dye labels is the need for extensive purification of the sample to ensure the removal of all unreacted labels, as there is no optical difference between bound and free dye. The Förster Resonance Energy Transfer (FRET) effect can be used to circumvent this issue, but it is limited to the quantification of optically active nanoparticles, such as quantum dots. In addition, direct quantification is only possible when the light scattering by NPs is negligible and dye-dye interactions can be excluded because of the low functional group density on the surface. Another option is quantitative nuclear magnetic resonance (qNMR), which is typically used to analyze the organic ligand shell of nanoparticles and to determine the corresponding ligand densities. For example, this method was used for the analysis of gold nanoparticles stabilized with an azide-carrying peptide. However, as the NMR signals get significantly broadened when the ligands are bound to the nanoparticle surface, the analysis becomes complicated in the case of mixed ligand shells, especially, in the case of a low ratio of azidecarrying ligands to stabilizing ligands. qNMR can also be performed on ligands released to the solution following the addition of stronger ligands or decomposition of the NPs, but the complete removal and recovery into solution of the ligands can be problematic, especially in the case of strongly binding ligands. The limitations of NMR also include the relatively high amount of material needed for analysis compared to optical methods, as well as the inability to study NPs with paramagnetic properties, such as iron oxide, widely used in MRI imaging, or lanthanide-doped NPs. |
67b722916dde43c908b286d4 | 3 | There is thus clearly a need for precise azide quantification methods that can be applied to various NPs. In this study, we developed two methods that require no purification, use a low amount of functionalized NPs, and are independent of the inorganic core of the NPs. As a model system, we take sub-5 nm hafnium oxide NPs (HfO 2 NPs), which have been recently explored as radiosensitizers for radiotherapy and as contrasts agents for X-ray computed tomography. We capped the HfO 2 surface with catechols, which were previously shown to provide the strongest binding and therefore the best colloidal stability for industrially and biomedically relevant metal oxide NPs such as hafnium oxide, 37 titanium oxide, and iron oxide NPs. We varied the amount of azide-carrying catechols in the ligand shell from 0 to 10 per NP to test the possibility of quantification of low amounts of functional groups necessary for precise bioconjugation. The first method leverages the fluorescence quenching effect of the nitrocatechol ligands on covalently bound dye molecules by photoinduced charge transfer (PCT). The second method is more general and is based on the UV-tracking of the DBCO triple bond absorbance upon click-reaction. Both methods gave highly comparable results and were able to accurately determine the number of azide groups on the surface of the NPs. The second method was successfully generalized to NPs with polyphosphonate ligands. |
67b722916dde43c908b286d4 | 4 | To develop a quantification method for azide functional groups, we first designed a model system of NMR-compatible metal oxide nanoparticles stabilized with PEGylated catechol-based ligands. Hafnium oxide nanoparticles were synthesized from hafnium tert-butoxide in benzyl alcohol, TEM images show ellipsoidal-shaped nanoparticles with a major axis of (3.8 ± 0.6) nm and the minor axis of (2.4 ± 0.3) nm, see Figure . The colloidal stability of the nanoparticles after ligand exchange in water was confirmed by DLS (Figure ), yielding a mean hydrodynamic diameter of 5.8 nm, which is in agreement with TEM considering the additional ligand shell. The colloidal stability of the nanoparticles is not influenced by the azide content (Figure ). Finally, the organic content of the nanoparticle was determined to be 24% by TGA (Figure ) and further used to calculate the number of nanoparticles for stoichiometric reactions (see Supporting Information for calculation). The organic content is similar for particles with different azide content (see Figure ). reference NMR spectra (Figure ). Because of the broadening and the low content of bound NDA-(EG) 2 -N 3 , it is impossible to distinguish it from bound NDA-(EG) 2 -OCH 3 . Therefore, all three samples appear identical by H NMR, see Figure . Clearly, the characteristic peak of the -CH 2 -N 3 protons (labeled as f') of NDA-(EG) 2 -N 3 is too broad and too low in intensity to allow quantification of the shell composition using the qNMR method, which is often used in surface ligand quantification. To avoid the effects associated with line broadening, ligands can be released ("stripped") from the surface of previously functionalized particles for routine NMR analysis. Strong acids such as trifluoroacetic acid or HBF 4 can be used to release bound ligands. The NPs lose their stabilizing ligands and precipitate, allowing them to be separated from the ligand solution. This method has been used primarily for a qualitative analysis of the surface. We attempted to quantify the surface ligands of our hafnia NPs by stripping them with HBF 4 (Figure and Table ). All weakly bound carboxylate ligands were removed from the surface but only 30% of catechol ligands were effectively removed. Although some NDA-(EG) 2 -N 3 was detected, its quantification becomes unreliable due to the incomplete stripping. FTIR spectroscopy qualitatively confirmed the presence of azide groups (Figure ) on the surface of the NPs. The characteristic band of the azide group at 2100 cm -1 is visible for nanoparticles with targeted NDA-(EG) 2 -N 3 content, as opposed to nanoparticles with NDA-(EG) 2 -OCH 3 ligands alone. However, the IR band is relatively weak due to the low concentration of azide groups on the surface of the NPs, highlighting the need for sensitive methods of quantification. |
67b722916dde43c908b286d4 | 5 | Labeling with fluorescent dyes is a potential strategy for quantification but is usually limited by the need for purification steps to remove unbound dye, unless the fluorescence of the bound dye differs from the free dye. In this respect, catechols can act as a donor or acceptor in photoinduced charge transfer (PCT) -depending on its structure -thus affecting the fluorescence intensity of bound dyes. Nitrodopamine-based ligands were shown to act as electron acceptors in donor-excited PCT due to the electron-withdrawing effect of the nitro group. Upon excitation of a conjugated dye, an electron is promoted from the dye's HOMO to its LUMO and then moves into the vacant LUMO of the nitrocatechol group via a non-radiative transition (thus quenching the fluorescence). Unlike FRET, the quenching effect only occurs when the dye is covalently bound to the nitrodopamine-based ligands. |
67b722916dde43c908b286d4 | 6 | For a given sample, we prepare several solutions with the same number of NPs and add increasing amounts of AF488-DBCO. Upon conjugation to NDA-(EG) 2 -N 3 , the fluorescence is quenched. After around 72 h at room temperature no further changes in fluorescence intensity were observed. Although "click" reactions are known for their high reaction rates, the relatively long reaction time is due to the very low concentrations used. The latter have the advantage that the samples are immediately ready for spectrocopic characterization and minimize the cost of consumed dye. The fluorescence intensity at 518 nm was measured and plotted against the initial dye:NP ratio, see Figure -f for NPs with 0, 2, 5 and 10 targeted NDA-(EG) 2 -N 3 ligands per NP (raw PL spectra can be found in Figure ). |
67b722916dde43c908b286d4 | 7 | In the absence of NPs, the fluorescence intensity increases linearly with the dye concentration (green data points in Figure ). In case NPs have only NDA-(EG) 2 -OCH 3 ligands (Figure ), the dependence is also linear and differs only slightly from the dye alone. For the samples where NDA-(EG) 2 -N 3 ligands are present, two linear regions are observed. The first linear region has a lower slope than the second linear region. Furthermore, the latter has a slope comparable to that of the dye alone (see also Table ). We conclude that the fluorescence of AF488 is quenched until the dye:NP ratio is equal to the valency of the NPs, i.e., the number of azide groups on the NP surface. Above that concentration, there is additional free dye, and its fluorescence is not quenched by PCT. Hence, PL quenching by nitrodopamine is only observed when the dye is covalently bound to nitrodopamine-based ligands. This was confirmed by control experiments in which AF488 was added to either free NDA-(EG) 2 -OCH 3 , free NDA-(EG) 2 -N 3 (see Figure ). Fluorescence is quenched by NDA-(EG) 2 -N 3 (until the stoichiometry reaches 1:1), but not by NDA-(EG) 2 -OCH 3 . |
67b722916dde43c908b286d4 | 8 | The intersection point of the two linear curves in Figures , represents the valence of the NPs. It is the amount of dye that was conjugated to the NDA-(EG) 2 -N 3 on the nanoparticle surface. To avoid any bias in the determination of the intersection, we opted for an automatic fitting using a piecewise linear function (see Methods and Supporting Information for further details). As can be seen in Table , the intersection found between two linear parts is indeed close to the target value of N 3 /NP and corresponds to the number of azide groups per particle. To test the repeatability and reproducibility of this method, we performed the azide quantification three times on the same sample (twice by the same operator and once by another, see Figures and Table ). All three independent experiments yielded the same result (within the experimental error). The above method based on fluorescence quenching can be applied only to particles with electron-accepting ligands. In this current section, we demonstrate a more general method that exploits the UV-Vis absorption band of the DBCO triple bond (alkyne) with a maximum at 308 nm (ε = 12000 L mol -1 cm -1 ), which disappears during cycloaddition (Figure ). To calculate the concentration of azide groups in the sample, we calculate the concentration of reacted DBCO using the Beer-Lambert law via: |
67b722916dde43c908b286d4 | 9 | The quantification via UV-Vis yielded identical results to the fluorescence quenching method, see Figure . Both methods are capable of detecting low amounts of surface azides (as low as 0.8 nmol for fluorescence quenching and 13.8 nmol for UV-Vis), do not involve any purification, and are independent of the magnetic properties of the nanoparticles. An important requirement is colloidal stability throughout the duration of the quantification measurements. Aggregation would result in scattering that would impact both fluorescence and UV-Vis spectroscopy. The methods could be potentially applied to other aqueous buffers because the click reaction is largely insensitive to pH. One limitation of the fluorescence quenching method is the need for electron-accepting ligands, but it can be an interesting alternative for NPs with high absorbance around 308 nm. To demonstrate the broad utility of our azide quantification strategy based on UV-Vis, we prepared hafnium oxide nanoparticles with a different ligand composition. We chose a (sulfobetaine-azide) phosphonic acid block copolymer as the capping ligand (Figure ), which has been shown to stabilize several different NPs (e.g, iron oxide) in physiological media while reducing protein corona formation. We first stabilized hafnia nanoparticles in water using (6-{2-[2-(2-methoxy-ethoxy)-ethoxy]-ethoxy}-hexyl)phosphonic acid and subsequently exchanged the phosphonate ligands for the polyphosphonate (Figure ). The nanoparticles were purified by size exclusion chromatography (SEC). The final NPs were characterized by DLS, TGA and NMR to confirm, respectively, the colloidal stability, the inorganic fraction, and the success of the ligand exchange, see Figure . P NMR spectroscopy is particularly useful for detecting bound polyphosphonate, in addition to a small amount of residual phosphonate (Figure ). The UV-Vis quantification of azides was carried out as described above, and the corresponding absorbance spectra and kinetic curves are plotted in Figure .The raw spectra can be found in Figure . Interestingly, the reactions were completed much later than for catechol-stabilized particles, which can be attributed to the lower concentration of particles used for this experiment and also to the sterical hindrance and accessibility of azide groups surrounded by polymer chains. The change of ∆A after 12 h of the reaction was used to calculate the number of azide groups per NP: 4.23 ± 0.21 N 3 /NP. This is in agreement with previously published estimates for the number of accessible azide groups in NPs with sim-ilar polymer ligands determined using fluorophore labels. Therefore, this UV-tracking method is a promising alternative for determining the amount of derivatizable azide groups without the need for labelling with dyes and the associated purification steps. |
67b722916dde43c908b286d4 | 10 | We demonstrated that the NMR quantification of azide-carrying ligands, either in the bound state or stripped from the NP surface, comes with several unsolved challenges. We presented two convenient alternatives based on fluorescence or UV-Vis spectroscopy, both having similar accuracy and precision. The first quantification method is based on fluorescence quenching of a fluorophore by photoinduced charge transfer to nitrocatechol ligands upon bioconjugation. The method is limited by the requirement for electron-accepting ligands. |
67b722916dde43c908b286d4 | 11 | It does not depend on the chemical nature of the stabilizing ligand, as we demonstrated for both catechol and polyphosphonate ligands. In the future, both methods could be partially automated using a plate reader, which would significantly speed up the quantification process. Both developed methods are capable of detecting low amounts of surface azides (as low as 0.8 nmol for fluorescence quenching and 13.8 nmol for UV-Vis). Compared with other optical methods like dye labeling, our methods do not require any purification. Furthermore, these methods are orthogonal to the magnetic properties of NPs and could be used for the quantification of surface azides on, e.g., iron oxide NPs. We believe that this quantitative analysis of surface functional groups will play an important role in precise bioconjugation, which is crucial for biomedical applications of NPs. |
67b722916dde43c908b286d4 | 12 | High-resolution transmission electron microscopy (HR-TEM) imaging was performed on a JEOL JEM-F200 operated in TEM mode at a beam energy of 200 kV. Samples were prepared by drop-casting diluted solutions of NPs onto ultra-thin carbon film-coated 400 mesh lacey carbon copper grids. TEM images were analyzed by ImageJ software. For the core size measurements, 138 particles were manually fitted with the "Elliptical Selections" tool, and for each particle, the lengths of the major and minor axes were derived. |
67b722916dde43c908b286d4 | 13 | Thermogravimetric analysis (TGA) was performed on a TGA5500 (TA Instruments). The samples were heated to 800 • C at a ramping rate of 5 • C/min in air. The samples were held for 15 min at the final temperature to ensure that all organic substances had burned out. The TGA curves were analyzed with the TRIOS software. |
67b722916dde43c908b286d4 | 14 | Photoluminescence (PL) spectra were recorded in quartz cuvettes on a Spectrofluorometer SF5 (Edinburgh Instruments) with a standard cuvette holder SC-05. A xenon lamp was used as an excitation source, and data were processed with the Fluoracle software. Absorbance spectra were measured from 200-600 nm with a 1 nm bandwidth and a step of 1 nm. The absorbance data were used to calculate a real concentration of dye in the samples. |
67b722916dde43c908b286d4 | 15 | HfO 2 nanoparticles HfO 2 NCs were synthesized using a solvothermal method according to Lauria et al. In a nitrogen-filled glovebox, 2.26 g of hafnium(IV) tert-butoxide (4.8 mmol, 1.94 mL) and 40 mL of anhydrous benzyl alcohol were mixed in a Teflon-lined 125 mL Parr bomb, sealed and taken out of the glovebox. The Parr bomb was then placed in a muffle furnace for 96 h at 220 • C. After synthesis, the nanoparticles were collected by addition of 20 mL of diethyl ether and centrifugating for 3 min at 8000 rpm. The supernatant was discarded and the precipitate was washed once more with 20 mL of diethyl ether. For functionalization, the nanoparticles were dispersed in 5 mL of toluene, resulting in a milky turbid liquid, to which 381 uL of 2-[2-(2-methoxyethoxy)ethoxy]acetic acid (MEEAA, 0.328 g, 1.82 mmol) |
67b722916dde43c908b286d4 | 16 | was added followed by 20 min of sonication. The solution became transparent immediately upon the addition of the ligand. For purification, the nanoparticles were precipitated by adding 10 mL of hexane (a mixture of isomers) and centrifuged at 5000 rcf for 5 min. The supernatant was then discarded, and the solid residue was redispersed in 1 mL of toluene. |
67b722916dde43c908b286d4 | 17 | Alternatively, (6-{2-[2-(2-methoxy-ethoxy)-ethoxy]-ethoxy}-hexyl)phosphonic acid was used as a stabilizing ligand. In this case, the particles were dispersed in 5 mL of chloroform, and, instead of MEEAA, 1.2 mL of 0.5 M ligand solution in chloroform (0.197 g, 0.6 mmol) was added to the synthesized NPs and further purification was carried out in the same way as described above. The final yield was 44%. |
67b722916dde43c908b286d4 | 18 | NDA-(EG) 2 -OCH 3 synthesis: The ligand was synthesized in three steps according to Deblock et al. 1.028 mmol, 3.4 eq) of N-methylmorpholine was added dropwise to the mixture using an air-free technique. The mixture was left on stirring for 72 h at room temperature. For purification, the solvent was removed under the reduced pressure. The remaining dark brown liquid was dissolved in 3 mL of deionized water and 2 mL of 1M HCl was added to adjust the pH to 1. The product was extracted three times with 5 mL of chloroform. The organic phase was collected, dried under reduced pressure, and dissolved in MQ water. Finally, after freeze-drying a fluffy yellow product was obtained with a yield of 55%. (Sulfobetaine-azide) phosphonic acid copolymer: The synthesis of the methacrylate-PEG-N 3 monomer and the (sulfobetaine-azide)-phosphonic acid block copolymer was carried out as previously reported in and 57 respectively with the exception that only 2 eq of azide monomer was used for the synthesis of the sulfobetaine-azide polymer block. The monomer was isolated with a yield of 20% and the final polymer was obtained with an overall yield of 38%. |
67b722916dde43c908b286d4 | 19 | As both the nitrodopamine-based and polyphosphonate ligands are not soluble in toluene, we first disperse the nanoparticles in water before ligand exchange. A portion of the nanoparticle solution with the amount needed for the ligand exchange (typically, 20 mg) was dried from toluene on the Schlenk line. The solid residue was dissolved in 1 mL of ethanol, followed by 15 min of sonication. Then, the nanoparticles were dried again and dissolved in 1 mL of methanol in the same way as described above. Finally, dried nanoparticles were dissolved in the amount of MQ water required for the ligand exchange. For the NPs stabilized with MEEAA, slight turbidity was observed in the final dispersion caused by the partial desorption of carboxylic acid. |
67b722916dde43c908b286d4 | 20 | Ligand exchange to catecholate ligands For the preparation of nanoparticles with 5 or 10 N 3 groups per particle, the ratio of the N 3 carrying ligand is changed accordingly while keeping the total number of ligands as 1.1 eq w.r.t. surface ligands, and the rest of the procedure remained the same. |
67b722916dde43c908b286d4 | 21 | Sample preparation for qNMR Fluorescence quenching upon click-reaction with N 3 groups 10 solutions of 0.0164 mg/mL of nanoparticles (2.4 • 10 14 particles) with increasing amounts of fluorescent dye (AF488-DBCO) were prepared by diluting 32.8 µL of 1 mg/mL of nanoparticles in PBS (Dulbecco PBS, pH = 7.4) followed by addition of 0 -56.2 µL of a 0.01 mg/mL stock solution of the dye prepared in MQ water corresponding to 0 -15 eq equivalents of the dye w.r.t. to the number of NPs. The amount of PBS buffer was varied to keep the concentration of the particles constant (total volume = 2 mL) and the concentration of the dye in the stock solution was confirmed from the absorbance at 492 nm. All solutions were prepared in brown glass vials and kept in the dark after preparation to prevent photodegradation of the dye. The reactions were allowed to stir at RT for 72 hours in the dark before the PL measurements to ensure that the click reaction had reached completion. In parallel, reference solutions with the same concentration of the DBCO-carrying dye, but without nanoparticles, were prepared and kept under the same conditions. The measurement parameters and dye concentrations were chosen such that the emission intensity of the sample with maximum concentration did not exceed 10 6 counts to avoid saturation of the PL detector. |
67b722916dde43c908b286d4 | 22 | The maximum PL intensity at 518 nm was plotted against the number of dye equivalents (dye:NP ratio) and the obtained dataset was fitted with the built-in piecewise linear function with two segments "PWL2" in OriginPro (See more detailed explanation in Supplementary Information) from which the intersection and its error were calculated. The intersection found corresponds to the number of azide groups per NP. |
67b722916dde43c908b286d4 | 23 | For a typical experiment, five quartz FUV range spectrophotometer cuvettes (R-3010-T, Spectrocell Inc, USA) were equipped with a magnetic stirring bar and used for reactions and UV-vis absorbance measurements. In one of the cuvettes, a DBCO control was prepared: 600 µL of 0.35 µmol/mL DBCO-PEG 4 -alcohol solution in PBS was diluted with 2400 µL of PBS. The NPs control was prepared in a second cuvette: 600 µL of 1 mg/mL nanoparticles solution (N N P = 4.14 • 10 15 ) in PBS was diluted with 2400 µL of PBS. In three sample cuvettes: 600 µL of 0.35 µmol/mL DBCO-PEG 4 -alcohol solution in PBS, 600 µL of 1 mg/mL nanoparticles solution in PBS, and 1800 µL of PBS were mixed. There is at least 30x excess of DBCO w.r.t. number of particles in each sample. This means that for NPs with 2, 5 or 10 azide groups per particle, there is at least a 15, 6, or 3x excess of DBCO, respectively. |
67b722916dde43c908b286d4 | 24 | To take into account only the contribution of DBCO in the samples, we subtracted the spectra of the NPs control at different times from the sample spectra. Using the extinction coefficient of DBCO at 308 nm (ε = 12000 L • cm -1 • mol -1 ) and the optical path of cuvette l, the absorbance A of the control samples and the corrected sample were recalculated to concentration: |
67b722916dde43c908b286d4 | 25 | The concentration difference of DBCO at 8 h ∆C 8h between the control and the sample is recalculated to the number of DBCO molecules N DBCO . As DBCO and N 3 react in a 1:1 ratio, the number obtained is equal to the number of azide molecules N N 3 in the reaction mixture. Using the volume of the reaction mixture V reaction and Avogadro's number N A : |
64e4c031dd1a73847f4cbe23 | 0 | Austin reported that nitrofurazone can be reduced to a glyoxylopropionitrile by means of a reductive fission: a similar but different reductive fission of nitrofurans has also been reported elsewhere. This led Cisak et al., to study the reactivity of 5-nitrofurfural in alkaline and acidic solutions. In stark contrast to the aforementioned reported reactivity of 5-nitrofufural in alkaline solutions, almost invariably all the previously reported procedures for synthesis of hydrazone 1 have been conducted in highly alkaline hydrazine hydrate in methanolic solutions (Scheme 1). For though Li et al. reported that a purum sample of 1 was obtained via a modified literature procedure (loc. cit.) the particulars of which was never disclosed, it was not feasible to reproduce such findings in this work. To this end, and in accordance with the aforementioned inferential rationale for highly impure and somewhat yellow in colour (vide infra) polymeric samples of 1 which were phenomenologically obtained via the previously reported procedures, a highly versatile and relatively anhydrous method for synthesis of 1 and its derivatives was devised. |
64e4c031dd1a73847f4cbe23 | 1 | Anhydrous hydrazine itself is toxic and a powerful reducing agent capable of bringing about colossal explosions. In point of fact, a mixture of methanol 57%, hydrazine 30% and water 13% w/w, which is similar to that of the literature procedures (Scheme 1) for synthesis of 1 is also referred to as 'C-Stoff' and was made use of by the Germans during World War II as a rocket fuel, and by others elsewhere. 9 Guiding principles the practising synthetic chemist encouraged thereunder, preclude the use of unsafe and risky procedures, no matter how common they may be, as such there is a pressing need for reasonably safe procedures for synthesis of 1 and its derivatives. |
64e4c031dd1a73847f4cbe23 | 2 | Given the aforementioned hazards associated with the use of anhydrous hydrazine, hydrazine monohydrate which is much safer to handle than anhydrous hydrazine in combination with a desiccant was deemed a suitable substitute. As such, initially anhydrous Na 2 SO 4 was placed in a flame-dried vessel; thereto was added 5nitrofurfural, the vessel was sealed and degassed under a partial flow of a dry inert gas, and was subsequently cooled in an ice-bath, thereto was added anhydrous THF, followed by dropwise addition of hydrazine monohydrate (1.1 equiv.). |
64e4c031dd1a73847f4cbe23 | 3 | The desiccant in the mixture was able to remove most of the water so as to provide a suitable anhydrous condition under which the reaction proceeded to completion, and the desired crude hydrazone 1 was obtained in a pleasing 90% yield (Scheme 2). The crude was then purified using flash column chromatography (80% EtOAc in cyclohexane; R f 0.25) to obtain 1 in high purity ~ p.a. ≥98%, as evidenced by H NMR (Fig. ), LC-MS and HRMS (ESI S2 †), and intriguingly in marked contrast to that of the literature as an orange solid (Fig. ) in 62% yield. Ultra-pure samples of 6 and 7 were also obtained using the newly devised method (ESI S5 & S6 †), alongside 2 (ESI S3 †) for comparison with 1. Hydrazone 1 is quite insoluble in most common organic solvents such as ether, dichloromethane, toluene, chloroform etc., but sparingly soluble in acetonitrile and fairly soluble in DMSO. To this end, a relatively clean 1 H NMR spectrum of 1 in acetonitrile-d 3 was first obtained; however, the exchangeable protons, in the absence of deuterons, complicate the overall trace integration of the thus obtained spectrum: this was rectified by means of a 'D 2 O shake', nevertheless, hardly surprisingly, even traces of residual water from the added D 2 O reacted with 1 so as to give rise to minor inexplicable new peaks at 8.55, 7.53, 7.52, 7.27 and 7.26 (Fig. ). It cannot be too firmly emphasised that unlike clean 1 H NMR spectra of 1 in acetonitrile-d 3 , the 1 H NMR spectrum of 1 in methanol-d 4 was a complete mess (Fig. ), this clearly shed light on reactivity of methanol towards 1; thus any attempt at synthesis of this compound in methanolic solutions could lead to highly impure products. Knowing that methanol reacts with 1, it was then investigated as to whether methanol in lieu of the nucleophile i.e. hydrazine would also react with the starting material, namely, 5-nitrofurfural, in such syntheses. Unlike 1, 5-nitrofurfural is sufficiently soluble in chloroform; to this end, its 1 H NMR spectrum in CDCl 3 was obtained and compared with that in methanol-d 4 . It did not come as a surprise to discover that similar but different to 1, 5-nitrofurfural is equally susceptible to attack by methanol as its 1 H NMR spectrum in methanold 4 contains extra signals in the region 7.5 -5.5, which were absent in its obtained spectrum in CDCl 3 (Fig. ). |
64e4c031dd1a73847f4cbe23 | 4 | Analysis of the 1 H NMR spectrum of 5-nitrofurfural in methanol-d 4 (Fig. ), on looking more closely, together with electron deficiency of the furan ring in 5-nitrofurfural despises any S E Ar reaction, and suggests formation of hemiacetal 3 and/or acetal 4. It is unlikely that the reaction of methanol-d x with 5-nitrofurfural resulting in acetal 4 such that even though the -OCD 3 moieties in 4 are chemically equivalent, they are, in point of fact, anisochronous by virtue of the furan ring current. Thus, probably, almost certainly, two signals from the -OCD 3 moieties in 4 should be observed in its C NMR spectrum which is not the case, so it is very likely that the product be hemiacetal 3 (Fig. ). As depicted in Fig. Intriguingly, 1 H NMR spectra of 5-nitrofurfural in CDCl 3 and DMSO-d 6 did not differ much in terms of the overall pattern of signals (Fig. ). |
64e4c031dd1a73847f4cbe23 | 5 | As illustrated above, methanol can react with 1 to yield unwanted by-products, which perturb its colour; in effect, solutions of have a yellow tinge. As depicted in Fig. , there are a total of eight delocalised electrons in 1 which set up a conjugated system, leading to bonds elongation and lower energy gaps between the pertinent ungeraden and geraden molecular orbitals to such an extent that the 18 non-bonding electrons in this molecule can also participate in the delocalisation; as a the fundamental electronic excitation of 1 absorbs greenish-blue light ( max ~ 480 -490 nm), and hence this compound appears as an orange solid. Henceforth, whereas energy is variant in all privileged frames of reference by all observers, the change in energy is invariant in all privileged frames of reference for the same observers. By the same token, the fundamental electronic excitation between the pertinent ungeraden and geraden molecular orbitals of inherently coloured organic compounds is invariant in a given polyene system in any privileged frame of reference, irrespective of variant energies of different molecules upon which the conjugated system is conferred, in that privileged frame of reference by all observers. Taking as an example, albeit lycopene and β-carotene are energetically variant in a privileged frame of reference, both exhibit the same characteristic colour to all observers in that frame of reference, in virtue of their pretty much identical conjugated systems (Fig. ). It is well established that the more extended a conjugated polyene, the lower the energy gap between its HOMO u and LUMO g . Thus, as a rule of thumb, the more electrons delocalise in a conjugated polyene system, the lower the energy gaps between its ungeraden and geraden molecular orbitals, which in turn this effect stabilises the system relative to that with equal but isolated electrons. It follows that increasing the length of conjugation leads to a bathochromic fundamental electronic excitation. For this reason, n-hepta-1,3,5-triene with a six electron conjugated system absorbs blue light and appears as a deep-yellow solid. 12 On the other hand, azulene with ten delocalised electrons in its conjugated system absorbs green light, and exhibits an intrinsic purplish 13 colour. |
64e4c031dd1a73847f4cbe23 | 6 | Inasmuch as the fundamental electronic excitation is invariant in all privileged frames of reference by all observers, the number of delocalised electrons in a conjugated system should in the absence of exceptions e.g. tunnelling etc., dictate the intrinsic colour of its corresponding polyene, provided that the conjugated system is not affected by other factors such as auxochromic effects, Kuhn's 15 and Woodward's rules etc. To this end, empirically six to twelve delocalised electrons in an uncomplicated linear conjugated polyene bring about a fundamental electronic excitation with max ~ 450 -550 nm, giving rise to coloured compounds from yellow such as n-hepta-1,3,5-triene to purplish-pink azulene (loc. cit.). |
64e4c031dd1a73847f4cbe23 | 7 | In view of the above, 1 with eight delocalised electrons (Fig. ) similar to 1-methylindene should at a bare minimum, in theory, though not of necessity, absorb blue light ( max ~ 430 -480 nm). However, as reasoned above, the implied conjugative effect lowers the energy gaps between its ungeraden and geraden molecular orbitals such that the orbitals would be in a much closer proximity to the non-bonding orbitals that accommodate more than twice as much electrons as that in the bonding orbitals. At this point, in practice, the fundamental electronic excitation of 1 responsible for its orangish colour ( max ~ 480 -490 nm) 10 must be probably, almost entirely due to one of the allowed n u → g excitations with the highest transition probability, which would not be the case in the absence of its extended conjugation, and as purported in the literature (loc. cit.) would have been of a yellow colour. |
64e4c031dd1a73847f4cbe23 | 8 | The belief that the position of the absorption maximum is directly proportional to the length of the conjugated system implies that lycopene and β-carotene, both having 11 conjugated bonds in which 22 electrons shuffle (Fig. ) ought to have a much longer max than say, 1 with only four conjugated bonds (Fig. ). Brushing the effects of non-bonding orbitals in 1 aside, the max of lycopene/β-carotene seems to be close to that of 1, as all exhibit orangish-red colour. This discrepancy stems from the fact that for a 1-dimensional observable 'particle in a box', its invariant symmetrical potential V(x) with respect to space inversion is as follows: |
64e4c031dd1a73847f4cbe23 | 9 | Eq. 4 states that 𝑃 ̂|ψ E (x)⟩ is an eigenfunction with the same eigenvalue (E), which implies that E would be a 'degenerate' eigenvalue; however, that is not true in virtue of linear dependency of the eigenvectors |ψ⟩ and 𝑃 ̂|ψ⟩. On the other hand, for the 'particle in an n-dimensional box' i.e. 'Hilbert space' 21 its centrosymmetric potential V(r) is only dependent on the distance the centre, and its eigenvector angular momentum 𝐿 ̂2 is as follows: |l, m⟩ |
64e4c031dd1a73847f4cbe23 | 10 | In recognition of the above, even though both lycopene and β-carotene possess an 11 linear conjugated bond system, but because they are centrosymmetric, half of their molecular orbitals 'degenerate'. To this end, lycopene and β-carotene behave as if they only have a five or six conjugated bonds which is well in agreement with the empirical observations i.e. they appear as an orangish-red pigments. In contrast, a non-centrosymmetric polyene such as flavoxanthin with an eight linear conjugated bond system (Fig. appears exclusively as a bright yellow solid. This might seem a bit puzzling as the longer a conjugated system, the longer its max . However, even though the gaps between the ungeraden and geraden molecular orbitals of flavoxanthin should be smaller than say, that of lycopene, but because the energy levels are getting much closer, then the fundamental electronic excitation is no longer S 0 → S 1 and higher energy excitations will have a much higher transition probability to an extent that they preponderate over the lowest energy excitation, and hence flavoxanthin absorbs blue light, but lycopene absorbs the lower energy blue-green light. Equally, whatever concoction in the literature 6 has been ascribed to 1 must have absorbed the higher energy blue light, hence the resulting yellow compounds; nevertheless, in contrast to the literature precedents, as delineated above, the true puriss. or even purum samples of 1 are intrinsically orangish in colour (Fig. ). |
64e4c031dd1a73847f4cbe23 | 11 | Intriguingly, similar but different to accounts of the literature for the appearance of 1, in particular, its colour, 6 also exhibits a yellow colour (Fig. ), and inasmuch as this furan technically has only one extra auxochromic oxygen than say, 1; this auxochromic impact on the conjugated system of 6, renders this molecule to fit into the literature descriptions of 1 (loc. cit.), which all seem to sit right with 6 both in 'structure & colour' (Fig. ). On another development, carbohydrazide 7 was serendipitously obtained in lieu of 6 via the action of hydrazine monohydrate on 5-nitro-2-furonyl chloride (ESI S6 †): the colour of 7 is again of particular importance, and reference has already been made to the correlation between purity of conjugated nitrofuranyl derivatives and their intrinsic colours (vide supra). With that borne in mind, we saw that the delocalisation of eight electrons in 1 was responsible for its orangish colour; it also brought home to us that symmetry would influence the colour of conjugated molecules cf. lycopene and β-carotene: here again, 7 is not fully conjugated, but indeed symmetric. On the other hand, 7 possesses several nitrogen and oxygen auxochromes, which can donate their accessible lone pairs to the conjugated system so as to extend the electron delocalisation in the already established conjugated system. This effect together with the partial double bond character of semicarbohydrazide in the dimer facilitates its tautomerisation to give a more stable fully conjugated dimer 8 with 16 delocalised electrons in its conjugated system (Scheme 4). |
64e4c031dd1a73847f4cbe23 | 12 | Although the thus obtained tautomer 8 is centrosymmetric, and in theory, its fundamental electronic excitation would be no different from that in 1 with only eight electrons in its conjugated system; however, in practice, it is unclear as to what extent this tautomerisation does take place in solid phase for dimer 7; moreover, both 7 and its tautomer 8 possess more auxochromic oxygens in their conjugated systems than that of 1, which should not be overlooked. Thus, even though the above complications do add a certain element of confusion when we come to consider the fundamental electronic excitation for this type of conjugated nitrofuranyl species, but what matters is that the combined conjugative effect of 7 together with its more stable tautomer 8 in solid phase gives its crystal lattice an intrinsic coral pink colour (Fig. ); otherwise, the pertinent dimer itself similar but different to 1 would only exhibit a dark orangish colour, and that this should be so is a corollary of its centrosymmetric nature. |
64e4c031dd1a73847f4cbe23 | 13 | During the course of synthesis of nitrofuranyl derivatives described herein using the already established procedures, highly impure and polymeric samples were obtained. To this end, and in line with the previously reported reactivity of nitrofuranyl derivatives, in particular, 5-nitrofurfural in alkaline media, herein this work offers a versatile and high yielding new synthetic procedure for synthesis of nitrofuranyl derivatives which has potential to be extended to a large array of similar unstable heterocycles in protic solvents. Furthermore, 1 H together with 13 C NMR analysis shed light on the nature of some of the resulting products of the aforementioned side reactions, all of which differ in appearance, in particular, their intrinsic colours to that of the desired products. Inasmuch as the much remarked erroneous physical properties of 1 in the literature, in particular, its intrinsic colour had been overlooked for more than half a century, due, in part, to the highly impure samples of 1 elsewhere, it compelled the author to lay a secure foundation of sound theory of observable phenomena in question for the outsider such that considerable detail had to be given in parts of this manuscript which are concerned with the intrinsic colour of nitrofuranyl derivatives, and may, though it need not of necessity, be of interests to the synthetic and medicinal chemist when occasion demands. |
61bad290d6dcc2026c3a4b10 | 0 | Unconventional strategies for expanding the use of solar energy have attracted significant attention in recent years. Using photon upconversion, in which low-energy photons are combined to form high-energy light, it is expected that the conventional limits in photovoltaics can be shifted upwards. This process has also been utilized in contexts of e.g., optogenetics, targeted drug-delivery, photocatalysis, and photochemistry. For solar applications the mechanism called triplet-triplet annihilation photon upconversion (TTA-UC) is of specific interest, as this process functions under low-intensity, non-coherent light. By using a donor, or sensitizer, species in conjunction with a fluorescent annihilator, triplets generated by the sensitizer from incident long-wavelength light may be converted into a highly energetic singlet state within the annihilator species in a spin-allowed TTA process. This scheme has been demonstrated for many different spectral ranges and with a variety of compounds, spanning purely organic systems, nanocrystals, metallic complexes, and metal-organic frameworks, to name a few. |
61bad290d6dcc2026c3a4b10 | 1 | The most success in terms of upconversion (UC) efficiencies has been obtained in the visible region. In particular, red-to-blue TTA-UC have been reported with UC quantum yields (FUC) as high as 42% (out of a theoretical maximum of 50% owing to the two-to-one nature of the UC process), while other spectral regions have proven more challenging. Upconverting near-infrared or infrared light to the visible region, which is especially important for biological and photovoltaic applications, have seen much lower efficiencies with a FUC of 8% at best. Similarly, the performance of visible-to-ultraviolet (vis-to-UV) TTA-UC systems suffer from limited efficiencies. Significant progress has however been made recently, with reports on FUC of around 10% for three different systems. Still, there is no fundamental reason as to why much higher efficiencies would not be possible. |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.