id
stringlengths 24
24
| idx
int64 0
402
| paragraph
stringlengths 106
17.2k
|
|---|---|---|
6690e02801103d79c5cb42f9 | 19 | Pysoftk v1.0 adds a complete standalone module for the analysis of soft matter systems. The new module described in this paper provides a set of interconnected tools that are useful for determining the physical properties of soft matter self-assembled aggregates as well as the molecular-scale interactions that underlie such emergent behavior. One of the key features of PySoftK v1.0 is that it properly accounts for periodic boundary conditions when determining the positions of atoms within the molecules that make up a soft matter aggregate, particularly if the aggregate is larger than half the size of the simulation box in one or multiple dimensions. Other software tools are not designed to account for molecular assemblies of such size. A thorough set of tests have been created to cover all code to ensure its correct functionality. Furthermore, PySoftK v1.0 is designed to provide maximum flexibility to the user, so most functions output the data per outputted configuration of the trajectory, so that the user can decide how to represent or further process the data. Although the initial version of PySoftK has a particular focus on polymers, the analysis module has been created such that it is fully chemically agnostic. The goal of this module is to create an open-source platform that allows users to analyse complex structures, dynamics and interactions in their simulations with minimal user input. PySoftK v1.0 contributes to the standardisation of molecular-scale simulation analysis, which will promote accurate comparisons across different simulations to support the rational in silico design of new soft materials. |
6705093712ff75c3a1d7ee30 | 0 | Enamides, known for their versatility in organic synthesis, represent important motifs in complex natural products (Figure ), pharmaceuticals, as well as intermediates for successive functionalization. Common transition metal catalytic syntheses utilize palladium catalyzed decarbonylative elimination, reductive acylation of oximes by iron, coppercatalyzed amidation of vinyl halides, nickel-mediated decarbonylation of amino acid thioester, and rutheniumpromoted anti-Markovnikow hydroamination of terminal alkynes (Figure 1B, left). However, a broadly applicable methods, favoring the (Z)-products, operating at ambient conditions and starting from renewable resources and naturally abundant transition metals would be highly desirable. Metallaphotoredox catalysis, which has greatly expanded the capacities of traditional cross-coupling reactions, enables the conversion of naturally abundant fatty or amino acids into target molecules under mild blue light conditions. The combination of an initial decarboxylation event with subsequent metal-catalyzed elimination or dehydrogenation would be an effective strategy achieving the desired enamide products. Seminal photocatalytic studies by Glorius and Ritter employed the decarboxylative elimination or dehydrogenation of amino or aliphatic carboxylic acids. On the one hand, Glorius' group reported a photocatalytic dehydro-decarboxylation of redox-active esters utilizing Cu catalysis merged with an organic dihydrophenazine photocatalyst. On the other hand, Ritter's group reported the combination of Co and Ir photoredox catalysis to obtain terminal olefins. In related studies, Tunge et al. employed Cu or Co catalysis for hydrogen atom transfer (HAT) merged with organic (acridinium) photocatalysts and achieved enamides from amino acid (Figure , right). Enamide products were obtained in good yields with moderate (E):(Z) ratios. Additionally, the groups of Larionov, König, as well as Wang and Chen developed strategies to obtain enamides or enamines with the help of photocatalysis. Most noteworthy, König's group introduced a Ni and Ir catalytic system capable of facilitating a formal β-hydride elimination, which was promoted by bromine radical formation and subsequent HAT, producing cyclic products in high yields. Despite these seminal contributions, the highly (Z)-selective synthesis of enamides remains challenging and arylated motifs as well as trifunctionalized olefins remain inaccessible. To this end, we were wondering whether (Z)-enamines can be obtained by an initial Ir-photocatalyzed decarboxylation utilizing natural or non-natural amino acids as abundant starting materials. An Ir photocatalysts with an appropriate redox window for efficient initial decarboxylation will lead to the formation of carboncentered radicals, which readily engage with nickel catalysts (Figure ). In the presence of aryl halides, nickel(III) intermediates are formed, which could be reduced if long-lived, facilitating the β-H elimination step. Kochi observed that orthosubstituents on aryl halides hamper the reductive elimination in Ni-catalyzed cross-coupling reactions. Likewise, Percec studied Ni-catalyzed homo-polymerization and found a decrease in yield when ortho-substituents were present and more recently Doyle detected sluggish oxidative addition kinetics of 2bromotoluene. Based on these seminal reports we argue that such an ortho-effect would be a powerful design criteria in our hands for redox-modulation of Ni(III) intermediates and thus outcompeting reductive elimination with β-H elimination. Subsequently triplet energy transfer from the very same Ir photocatalyst leads to distinct (Z)-selectivity of enamide products based on highly efficient photo-isomerization. In this report we successfully achieved the (Z)-selective synthesis of enamides via a new nickel photocatalytic strategy merged with single electron transfer (SET) and energy transfer catalysis simultaneously. Moreover, we could expand this reactivity to the synthesis of aliphatic and aromatic carbonyl compounds. |
6705093712ff75c3a1d7ee30 | 1 | Starting from plausible nickel photoredox conditions we developed the decarboxylative dehydrogenation reactivity in order to obtain (Z)-enamides as pure product (Table ). To facilitate this, we employed Cbz-protected phenylalanine (1a) as model substrate. The entire optimization process can be found in the SI (Scheme S1-3 and Figure ). No dependence was observed for the employed nickel source thus the nickel(II)chloride glyme complex was used (Scheme S3). Using electron-rich and ortho-functionalized arenes like ortho-methoxy bromobenzene (X1) or 2-bromothiophene (X2) (Table and and and) both gave high yields with moderate (Z)-selectivity utilizing benchmark Ir(dF(CF3)ppy)2(dtbbpy)PF6 (Ir-1) as photocatalyst (PC). We anticipated that structural modifications on the Ir(III) complex would enable more efficient photocatalytic selectivity. Hence, we prepared derivatives with modifications on the phenylpyridine (ppy) and bipyridine (bpy) respectively (Table ) and were pleased to see that removal of the fluorine atoms on the ppy led to high yield and promising selectivity. Ir((CF3)ppy)2(dtbbpy)PF6 (Ir-3), a previously unreported photocatalyst, exhibited a higher (Z)-selectivity (4:1) combined with high yield of 76% of the β-phenyl enamide product 2a ). A detailed investigation of the physical properties of Ir-3, including UV/Vis spectroscopy, fluorescence spectroscopy, and cyclic voltammetry (CV), can be found in the SI. Strongly reducing PC like Ir(ppy)3 or Ru(bpy)3(PF6)2 performed poorly since the initial decarboxylation is unfeasible (Scheme S1) and common organic PC, like 4CzIPN, gave both low yield and selectivity (Table ), which is remarkable compared to the photochemical synthesis of (Z)-alkenes in the presence of this PC. Subsequent solvent screening revealed that lower polarity solvents like DME displayed significantly improved (Z)selectivity (18:1) combined with a high yield of 93% (Table , Entries 7-10). Functionalized thiophene X3 gave a high yield of 86% for 2a combined with exclusive (Z)-selectivity (Table , Entry 11). Similar results were obtained by using 2chlorobenzotrifluoride (X4) as another ortho-substituted additive. However, a preference for Boc-protected phenylalanine was observed (Table and). This result indicates that the steric interaction of the substrate's protecting group and the orthoeffect of the aryl halide plays an important design role for nickel photocatalytic dehydrogenation. Nonetheless, further research will shed light on the nature of sterically demanding orthosubstituents as handles for stereoselectivity in nickel catalysis. From the optimal conditions we conclude a plausible mechanism for the decarboxylative dehydrogenation reaction, which is depicted in Figure . The process initiates with the photoexcitation of Ir-3 by blue light, followed by intersystem crossing, resulting in the formation of a long-lived triplet excited state *Ir(III). Deprotonation of substrate 1a and excited state oxidation (E*red(*Ir-3 III /Ir-3 II ) = +1.24 V vs. SCE) enables rapid decarboxylation, generating the carbon-centered radical intermediate 2a'. Low-valent and in situ generated Ni(0) species I readily traps radical intermediate 2a' yielding Ni(I) species II. Oxidative addition of an aryl halide (e.g., X4) occurs, forming the key Ni(III) intermediate III. We propose that III can undergo reductive halide dissociation when reduced by Ir-3 (Ered(Ir-3 II /Ir-3 III ) = -1.40 V vs. SCE), a process previously demonstrated by Molander. Subsequently, β-H abstraction is likely to proceed from Ni(II) complex IV due to more favorable agostic interactions and its relatively high basicity, resulting in the formation of Ni(II)-H species V and dissociation of the olefin 2a. For full optimization see SI, Scheme S1-3 and Figure . |
6705093712ff75c3a1d7ee30 | 2 | Finally, reductive elimination from Ni(II)-H species V is considered thermodynamically feasible, yielding the dehalogenation by-product ArH, which was observed quantitatively by both F NMR and GCMS, and regenerating Ni(0), thereby closing the catalytic cycle. We assume that 2a already features high (Z):(E) ratios at this stage (vide infra). Nevertheless, we examined the subsequent possibility of an energy transfer process, where remaining (E)-2a is converted efficiently to (Z)-2a in the presence of photocatalyst Ir-3 (Figure 2B, Table). In the absence of light, no thermal background isomerization is observed (Entry 1). Likewise, without PC present poor selectivity is observed for UV-light or blue light irradiation, respectively (Entries 2-3). Only under standard reaction conditions high (Z)-selectivity is obtained (Entry 4). To further confirm the (E) to (Z) photo-isomerization, we followed the kinetics by photo-probe 1 H NMR spectroscopy. The results indicated full conversion within 5 min with a rate constant for the isomerization reaction (E → Z) of 1.5 s -1 (Figure ). We were surprised to observe a strong difference in reaction performance for various protected amino acids, which is in contrast to previous reports. Carbamate amine protecting groups like Boc performed exceptionally well (Figure , Table ). However, blocking the amine proton with a methyl group completely suppressed the reactivity and product 3 was not formed (Entry 2). Amides, like acetate and benzoate, gave slightly lower yield with maintained distinct (Z)-selectivity (Entries 3-4). Unfortunately, phthalimide or tosyl protection groups hampered the reaction and products 2e and 2f could be obtained (Entries 5-6). DFT calculations (B3LYP D3/def 2 level of theory, see SI) indicate steric demand of Ni(II) IV intermediates and distortion toward the (Z)-elimination product (Figure , gray box), similarly as reported by Newhouse for a cobalt catalytic system. Unfavorable geometries are obtained for 3 and 2e thus pointing toward a steric mismatch. Steric bulk at the final double bond was tolerated giving access to trifunctionalized olefins (Figure , bottom right). β-Methyl andphenyl phenylalanines were both tolerated, but showed a strong aryl halide dependence leading to 49% analytical yield for 4 with X4 and 85% yield for 5 with X3, respectively. α-Methyl phenylalanine 6 was well tolerated and a smaller dependence toward either of the conditions was observed. Further research is required to better understand the origin of steric and electronic influences on the reactivity. |
6705093712ff75c3a1d7ee30 | 3 | With the established optimized conditions for decarboxylative enamide formation in hand, we proceeded to explore the scope of various natural and non-natural amino acids (Scheme 1). Both Cbz-and Boc-protected phenylalanine (2a and 2b), notably, exhibited excellent performance with exclusive (Z)-selectivity in up to 73% isolated yield. This methodology demonstrated good tolerance towards a variety of phenylalanine derivatives and both electron-donating groups (7 and 8) and electron-withdrawing groups (9-13) were applicable in up to 78% isolated yield with high (Z)-selectivity. Gratifyingly, this approach is compatible with Bpin (4,4,5,5-tetramethyl-1,3,2-dioxaboryl) substituted-phenylalanine 12, enabling future Suzuki cross-coupling reactions to construct new molecular entities. Additionally, nitrile substituted phenylalanine 13 exhibited excellent yield, however with moderate stereoselectivity, which hints to the electronic dependence of the photoisomerization. As described above, diphenylalanine 5 exhibits excellent yield. Delightfully, our method can be extended to (non-)natural amino acids with a γ-carbonyl group resulting in α,β-unsaturated products 14 and 15, although with exclusive (E)-selectivity for the acyclic one. Heteroaromatic and fused compounds showed either lower yield or lower selectivity. Pyridyl-substituted product 16 gave only 30% isolated yield, but with high stereoselectivity. Fused-aromatic systems like 2-naphthyl-alanin (Nal)-and tryptophan (Trp)-derived products 17 and 18 feature excellent yields with moderate (Z)-selectivity. Most likely, products with expanded π-character exhibit facile backisomerization from the (Z)-isomer and thus result in poor selectivity. While cyclic unfunctionalized substrates showed decreased reactivity (Scheme S6, vide infra), dihydroisoquinoline 19 exhibits lower yield compared to indane 20 and tetralin 21, indicating that ring-strain is no limitation for the β-H elimination step. Among them, tetralin 21 underwent decarboxylative elimination followed by dehydrogenation to yield naphthalene, indicating a subsequent aromatization mechanism. Control experiments demonstrate that all reaction components are crucially required for the successful decarboxylative dehydrogenation reactivity, since without light, both catalysts, or without aryl halide no product was obtained (Scheme S4). In the absence of X4, 15% yield of product 2b may be attributed to enhanced substrate solubility in DME and high basicity of BTMG. Likewise, initial decarboxylation is key, since N-Boc 2phenyl ethylamine itself does not convert to 2b under the optimized conditions (Scheme S5). |
6705093712ff75c3a1d7ee30 | 4 | In order to apply this strategy to more complex biomolecules in the future, we probed dipeptide substrates (Scheme 2). Phenylalanine-containing dipeptides exhibited moderate yield but maintained excellent (Z)-selectivity for 23 and 24 respectively, in line with the effect of protection groups observed before (vide supra). Primary carboxylic acids are not yet compatible and dehydroalanine product 22 was not obtained. |
6705093712ff75c3a1d7ee30 | 5 | Scheme 1: Scope of the photochemical nickel-catalyzed decarboxylative dehydrogenation reaction. Reactions performed on a 0.5 mmol scale. Isolated yields given and 1 H NMR yields vs. mesitylene as internal standard in parentheses. Isolated product and major isomer shown. (a) ArBr X3 was applied as aryl halide with Cs2CO3 as base. ArCl X4 was applied as aryl halide with BTMG as base. (c) No aryl halide was added. |
6705093712ff75c3a1d7ee30 | 6 | In conclusion, we report a novel photolytic Nickel-catalyzed decarboxylative dehydrogenation toward (Z)-enamides using natural and non-natural amino acid starting materials. This transformation is reinforced by modulating the electronic and steric properties of nickel intermediates with aryl halides additives. Additionally, the introduction of ortho-substituents significantly enhances β-H abstraction due to diminished reductive elimination kinetics. A novel photoredox catalyst (Ir-3) has been carefully designed, incorporating the capability of performing SET and EnT in the same reaction mixture for obtaining stereoselective enamides. Ultimately, we have successfully performed the synthesis of diversely functionalized arylated (Z)-enamides in up to excellent yields and selectivity. The mechanistic details, in particular with regards to the steric effects of this methodology on the nickel catalytic cycle were elucidated and are currently under further investigation in our laboratory. The careful modulation of redox states of nickel catalysts will enable the design and development of new catalytic reactivities in the future. |
67be0f8b81d2151a020bcc40 | 0 | Heteroaryl groups are functional groups in organic molecules that contain a heteroaromatic ring with at least one heteroatom, such as nitrogen, oxygen, and sulphur. They are naturally abundant and widely utilized in functionalized organic molecules. For example, pyridine, one of the most prevalent nitrogen-containing heteroarenes, is present in 54 small molecule drugs approved by the FDA between 2013 and 2023. Heteroaryls are also common structural motifs in both coupling partners and catalysts in transition metal catalysis and organocatalysis. The prevalence of heteroaromatic compounds is attributed to several factors. First, different heteroarene cores have distinct electronic and steric properties that effectively alter the target compound's chemical reactivity and biological function. One area where these properties have been leveraged is in the design of covalent modifier drugs with warhead reactivity modulated by heteroaryl groups; examples of this include FDA-approved drugs afatinib and selinexor, as well as other covalent modifiers such as roblitinib, nitrofuran derivative C-176, and TC9-305 (Fig. ). Second, heteroarene cores could be further functionalized with electron-donating or electron-withdrawing substituents at different sites, leading to a large number of regioisomers with a substantially expanded property space of heteroaromatic compounds. Third, structurally diverse functionalized heteroarenes could be synthesized from readily available starting materials via a number of established synthetic methods, including recently developed strategies via site-selective functionalization and skeletal editing . |
67be0f8b81d2151a020bcc40 | 1 | Quantitative description of the intrinsic steric and electronic properties of heteroaryl substituents is essential for establishing structure-activity relationships (SARs) and machine learning models for heteroaromatic compounds used in drug discovery and reaction design. DFT-computed descriptors are widely used in catalyst design , materials science , and reactivity and selectivity predictions . Descriptors for heteroaromatic compounds such as HOMO/LUMO orbital coefficients/energies and atomic charges have been applied to various reaction types, including electrophilic and nucleophilic aromatic substitution, and radical C-H functionalization. Despite these advances, a systematic and comprehensive database that integrates various physical-organic descriptors for heteroaryl substituents is still lacking, which has hindered the development of reactivity and selectivity prediction models. In contrast to the broadly used Hammett substituent constants (σp and σm) to describe electronic properties of aryl substituents, similar universal electronic descriptors for heteroaryl substituents have not been developed (Fig. ). This is in part due to the lack of experimental data (i.e., pKa values of corresponding heteroaryl carboxylic acids) as well as the inherent complexity of heteroaryl groups with different ring types, heteroatom substitutions, and regioisomers. We expect that a database of steric and electronic descriptors of heteroaryl substituents could serve as a foundation for developing robust predictive models that expand to the entire chemical space of heteroaromatic compounds. These could streamline reaction and catalyst developments by enabling the rational selection of heteroaryl substituents based on their electronic and steric features, rather than relying on trial-anderror approaches. In addition, this database for intrinsic chemical reactivity factors would complement existing cheminformatics databases for heteroarene synthetic feasibility and ADMET properties , which have been broadly used in drug discovery. Here, we present HArD, a HeteroAryl Descriptor database of >31,500 heteroaryl substituents based on 238 commercially available parent heteroarene cores (Fig. ). To capture the structural diversity of heteroaryl groups, we included both 5-and 6-membered heteroaromatic rings as well as 5,6-and 6,6-fused ring systems with carbon, nitrogen, oxygen, and sulphur as possible heavy atoms in the ring scaffold (Fig. ). Each parent heteroarene was functionalized with commonly used electron-withdrawing and electron-donating substituents to give monosubstituted heteroaryl groups (Fig. ). For each heteroaryl substituent, 49 DFT-computed electronic, steric, and geometrical descriptors and 16 fingerprint-type descriptors were included (Fig. and). This database includes computed Hammett-type substituent constants for heteroaryls (σHet), which would allow straightforward extensions of existing SAR and ML models of aryl compounds based on Hammett substituent constants (σp and σm) into previously unexplored space of heteroaryl-containing compounds. These newly developed σHet electronic parameters were computed based on pKa values of corresponding heteroaryl carboxylic acids (Fig. ), in analogy to the original definition of Hammett constants for aryl substituents to enable backward compatibility. In addition, other previously used descriptors, such as HOMO/LUMO coefficients, HOMO/LUMO energies, and partial atomic charges have also been computed for all heteroaryl groups in the database. Overall, HArD not only bridges a critical gap in the quantitative characterization of heteroaryl substituents but also provides a practical tool to design and predict the properties of these key building blocks in drug discovery, catalysis, and materials science. |
67be0f8b81d2151a020bcc40 | 2 | Establishing the Heteroaryl Library. Parent heteroarene cores were selected based on commercially available unsubstituted heteroaromatic compounds with five-and sixmembered rings, as well as 5,6-and 6,6-fused ring systems from the Reaxys® database (Fig. ). Only compounds with C, N, O, and S atoms in the heteroaromatic rings were included. A total of 238 unsubstituted parent heteroarenes were selected, including 23 five-membered heteroarenes, 9 six-membered heteroarenes, 157 5,6-fused rings, 47 6,6-fused rings, plus benzene and naphthalene. This results in 812 regioisomers of unsubstituted heteroaryl groups. Next, each unsubstituted heteroaryl group was functionalized using the RDKit 33 "ReactionFromSmarts" function to substitute a C-H bond on the heteroaromatic ring with a substituent to generate mono-substituted heteroaryl groups. The substituents used include 12 common electron-donating and electron-withdrawing groups-NMe₂, NH₂, OH, OMe, Me, TMS, F, Cl, Br, Ac, CN, and NO₂. This resulted in approximately 31,500 unique heteroaryl groups (Fig. ). To calculate the steric and electronic properties of each heteroaryl group (ArHet), SMILES strings of three compounds were used, including ArHet-H, ArHet-CO2H, and ArHet-CO2 -. The RDKit Experimental-Torsion Distance Geometry (ETDG) method was used to generate 3D structures as Gaussian 16 input files for subsequent DFT calculations. |
67be0f8b81d2151a020bcc40 | 3 | Geometries of all structures were optimized using the dispersion-corrected 36,37 B3LYP-D3(BJ) functional with the 6-31+G(d) basis set using the Gaussian 16 program 35 (Fig. ). Vibrational frequency calculations were performed at the same level of theory of the geometry optimization to confirm that each structure is a local minimum (i.e., with no imaginary frequencies). Single point energy calculations were carried out using the M06-2X functional with the 6-31+G(d) basis set. Solvation energy corrections were calculated using the SMD solvation model 41 in single point energy calculations with water as the solvent. Carboxylic acids (ArHet-CO2H) and carboxylate anions (ArHet-CO2 -) may have several conformers depending on whether the carboxylic acid or carboxylate group is co-planar with the heteroaromatic ring. The "SetDihedralDeg" function in RDKit was used to generate conformers of carboxylic acids and carboxylate anions by rotating about the Cipso-Ccarbonyl bond. Only the lowest energy conformer of each structure was used to compute the reported properties. The Automated Quantum Mechanical Environments (AQME) software was used in post-processing to check for self-consistent field (SCF) and geometry optimization convergence errors and imaginary frequencies. Calculations with convergence errors were resubmitted by using the intermediate structure during the previous geometry optimization with the lowest root-mean-square gradient as the input geometry. In cases where imaginary frequencies were present, the calculations were adjusted by slightly perturbing the geometry and resubmitted with the keyword "opt=(calcfc,maxstep=5)". This automated process was repeated twice, and any calculations still showing errors after the attempted re-calculations were not included in the final database. |
67be0f8b81d2151a020bcc40 | 4 | where R is the gas constant, T is 298.15 K, and the Gibbs free energies (∆G) of the carboxylate anion and the carboxylic acid were calculated using DFT at the M06-2X/6-31+G(d)/SMD(sSAS,H2O)//B3LYP-D3(BJ)/6-31+G(d) level of theory under standard conditions (i.e., 298.15 K, 1 mol/L). A modified scaled solvent-accessible surface (sSAS) approach was used in the SMD solvation model. The value 270.29 kcal/mol was used for the Gibbs free energy of a proton in aqueous solution (∆𝐺 , & ). This is calculated from the sum of the gasphase free energy of proton (-6.28 kcal/mol) and its hydration free energy 44 (-265.9 kcal/mol). Similar to Hammett constant for substituted aryl groups, a negative σHet value indicates a more electron-donating heteroaryl compared to phenyl, whereas a positive σHet indicates a more electron-withdrawing heteroaryl. While traditional aryl Hammett substituent constants often fall in a range of approximately -1 to +1, σHet values exhibit a broader range (Fig. and), showcasing the electronic diversity of heteroaryl groups. Regioisomers of the same parent heteroarene could offer significant variance in their σHet values (e.g., 0.91, 0.71, and 1.33 for 2-pyridyl, 3-pyridyl, and 4-pyridyl, respectively, Fig. ). Further, while 5-membered rings such as pyrrole are known to be often more electron-donating than 6-membered rings such as pyridine, these rings can be tuned using electron-withdrawing groups to alter their electronic properties (Fig. ). 2. Other Electronic Descriptors. Electronic properties such as HOMO and LUMO energies, total dipole moments, and quadrupole moments were extracted from Gaussian 16 single-point energy output files using the Python-based cclib library. From these values, the chemical potential, HOMO-LUMO gap, global electrophilicity, and global nucleophilicity were derived. The chemical potential was calculated from the average of the HOMO and LUMO energies , while the HOMO-LUMO gap is the energy difference between these orbitals. Global electrophilicity (electrophilicity index) was determined using the formula 𝜔 = -$ ./ where μ is the chemical potential and η is the HOMO-LUMO gap. Global nucleophilicity (nucleophilicity index) was computed as the inverse of global electrophilicity (𝑁 = 1 𝜔 ). The HOMO and LUMO coefficients of the ipso carbon were extracted directly from the Gaussian 16 outputs (keyword: pop=(orbitals=10,ThreshOrbitals=5)). Partial atomic charges were computed using Natural Population Analysis (NPA), Hirshfeld, and Charge Model 5 (CM5) charge schemes for the ipso carbon and the sum of the atomic charges of the heteroaryl group in the ArHet-H, ArHet-CO2H, and ArHet-CO2 -compounds. |
67be0f8b81d2151a020bcc40 | 5 | All steric descriptors were computed using the MORFEUS program. These include Sterimol parameters, buried volume, distal volume, Cipso-H bond length in ArHet-H, and Cipso-Ccarbonyl bond lengths in ArHet-CO2H and ArHet-CO2 -. Sterimol parameters, developed by Verloop, describe substituent size. The Sterimol length (L) is the vector length from the hydrogen on the ipso carbon of ArHet-H through the carbon to the tangent of the van der Waals (vdW) surface. B1 and B5 are the minimum and maximum widths, defined by the shortest and longest vectors from the ipso carbon to the vdW surface and perpendicular to L. Buried volume was originally developed to quantify the steric hindrance caused by ligands in transition metal complexes. Here, the fraction buried volume (VBur) was calculated from the percentage of space occupied by the ArHet substituent within a sphere of 3.5 Å radius centred on the ipso carbon (Fig. ). The distal volume describes the volume occupied by ArHet outside of this sphere. Using this approach, the computed steric descriptors could distinguish between different regioisomers (e.g., 3-bromo-2-pyridyl vs. 3-bromo-4pyridyl, Fig. ). The universal quantitative dispersion descriptor (Pint) was derived by constructing a molecular vdW surface and calculating dispersion coefficients using Grimme's D3 dispersion correction method. The solvent-accessible surface area (SASA) of the ipso carbon of ArHet was determined using the double cubic lattice method (DCLM) algorithm, which applies a constant surface density of points using a 1.4 Å probe. The volume enclosed by this solvent-accessible surface is also computed. 57 |
67be0f8b81d2151a020bcc40 | 6 | where Ropt is the optimal bond length for a reference aromatic bond, n the total number of bonds within the ring, and ⍺ serves as a normalization factor 60 to scale the result to a value between 1 and 0, where 1 indicates perfectly aromatic benzene and 0 represents the hypothetical Kekulé structure of a nonaromatic 1,3,5-cyclohexatriene ring. The Ropt and ⍺ values for CC, CN, CO, CS, NN, and NO bonds were taken from the literature. For the NS bond, the optimal bond length Ropt = 1.61 Å was also taken from the literature, 61 whereas the 𝛼 (71.875 Å -2 ) was calculated according to standard procedures. 58 |
67be0f8b81d2151a020bcc40 | 7 | The HeteroAryl Descriptor (HArD) Database, along with a Python script for data processing, is available on a publicly accessible GitHub repository (github.com/turkiAlturaifi/HArD). The repository includes the scripts used to generate the database, as well as an Excel file listing the SMILES representations of mono-substituted heteroaryls along with their associated descriptors. The repository is organized into two main folders: (i) the database processing |
67be0f8b81d2151a020bcc40 | 8 | folder, which contains a script for end-users to perform SMILES-based searches of the database and extract descriptor data from the Excel file; and (ii) the database generation folder, which includes scripts and files for developers to further extend the existing data points, including scripts to filter Reaxys query search, generate substituted heteroaryls, perform high-throughput calculations, analyse the results, and extract descriptors. Additionally, a copy of the code, an Excel file, and the Cartesian coordinates of all optimized geometries (provided as XYZ files) are available on FigShare. Finally, we provide a simple website (hard.pengliugroup.com) to search for the descriptors by SMILES strings or via a graphical interface. |
67be0f8b81d2151a020bcc40 | 9 | In our automated DFT-calculation/descriptor extraction workflow, several validation checks were performed, including post-processing using the AQME software to address convergence issues (vide supra) and connectivity validation to exclude intrinsically unstable compounds. Similar to traditional Hammett substituent constants, which only included metaand parasubstituted benzenes to avoid influences of steric effects of ortho substituents, we excluded σHet values for all heteroaryl groups with another substituent at an ortho position and those with A 1,3 -type interactions (e.g., 4-benzothiophenyl with another substituent at the 3position). These procedures ensure that the reported σHet values describe electronic properties only. |
67be0f8b81d2151a020bcc40 | 10 | Next, we analysed all DFT-computed descriptors in the database using both linear and nonlinear dimensionality reduction techniques, specifically Principal Component Analysis (PCA) and Uniform Manifold Approximation and Projection (UMAP) (Fig. ). In the PCA analysis, the principal components with the greatest variances consist of atomic charge descriptors (PC1, explaining 28% of variance) and quadrupole moments, dispersion, and steric (PC2, 19% of variance). The PCA plots of PC1 versus PC2 show broader distributions of properties across each heteroaryl class (purple, red, yellow, and green for 5-, 6-membered rings, and 5,6-and 6,6-fused cycles, respectively) than those of aryl groups (black colour) (Fig. ). Similarly, the UMAP projection (Fig. ) revealed that many heteroaryl groups occupy distinct property space not accessible by aryl substituents. Heteroaromatic compounds are known to have distinct electronic and steric properties affected by their heteroarene cores. We selected a subset of common heteroaryl groups and plotted their average steric and electronic properties based on the heteroarene core (Fig. ). We chose a commonly used steric descriptor, fraction buried volume (VBur), and our newly developed Hammett-type substituent constants (σHet) as an electronic descriptor to illustrate the properties of the heteroaryl groups. Each point represents the average value of all heteroaryls with the same heteroarene core shown in the top of Fig. . This plot revealed several general trends that could be qualitatively validated by previous experimental observations. In terms of steric effects, five-membered heteroaryls are generally less hindered due to the smaller size of the ring, and sulphur-containing heteroaryls often have larger fraction buried volumes due to the longer C-S bond (e.g., compare thiophenyl 10 with other five-membered heteroaryls). As expected, substituting a C-H moiety in phenyl or naphthyl with a nitrogen atom decreases the fraction buried volume (e.g., benzene 14 > pyridine 15 > pyridazine 16 > triazines 18 and 19). In terms of electronic properties, five-membered heteroaryls exhibited more diverse electronic effects than other ring sizes. Based on σHet values, pyrrole ( ) is more electron-rich than benzene (14), consistent with its higher reactivity in electrophilic aromatic substitution reactions. On the other hand, 1,3,4-thiadiazole (13) and 1,2,4-oxadiazole (9) are among the most electron-deficient heteroaryls, which are consistent with previous experimental observations. With a wide array of DFT-computed electronic, steric, and geometrical descriptors for diverse heteroaryl groups in the HArD database, we set out to explore whether the computed descriptors correlate with previously reported experimental reactivity and selectivity data. In 2022, the Leitch group published an extensive set of experimentally determined free energies of activation (ΔG ‡ SNAr) of nucleophilic aromatic substitution (SNAr) of benzyl alcohol reacting with 2-chloropyridines, 4-chloropyridines, and other chloro-substituted (hetero)aryl compounds (Fig. ). In another recent work, the Baud group performed a systematic kinetics study of 2-sulphonylpyrimidines as thiol-reactive covalent warheads that can undergo SNAr reactions with biologically relevant thiols in selective protein arylation (Fig. ). The second-order rate constants for each warhead indicated a wide range of reactivity with the glutathione (GSH) nucleophile by altering the substituents at the 4-and 5-positions of the pyrimidinyl group and exchanging the pyrimidine ring for different parent heteroaryl groups. Both sets of experimental reactivity data strongly correlate with the DFT-computed CM5 charge of heteroaryl group, q(Het)-CM5 (Fig. and), illustrating the capability of the q(Het)-CM5 descriptor for quantitative SNAr reactivity prediction. Next, we examined whether the computed electronic descriptors correlate with experimental site-selectivity trends in radical-mediated heteroaryl C-H functionalization from Baran et al. (Fig. ). The computed ipso carbon LUMO coefficients in the HArD database could qualitatively predict the preferred |
67be0f8b81d2151a020bcc40 | 11 | The Excel file containing the descriptors can be utilized with various data analysis software libraries. We provide a simple Python script to perform searches by SMILES and by similarity. All scripts are accompanied by a README file, which provides guidance on reproducing and expanding this database. Finally, example bash and slurm scripts for high-throughput calculations on HPC systems are also included. |
659bbb7be9ebbb4db9ada167 | 0 | The growing demand for batteries, due to the widespread use of portable electronics and electric transportation, has led to extensive research efforts to develop more efficient, low-cost, and sustainable energy storage technologies. Since their inception in the early 1990s, lithium-ion (Li-ion) batteries have dominated the market for portable electronics, primarily due to their higher energy density compared to other battery types. As a result, they are persistently being explored and further developed as power sources for electric vehicles and grid energy storage. Over the past two decades, several insertion-based cathode materials such as LiCoO2, LiMn2O4, and LiFePO4 have been developed, and are commonly used in modern Li-ion batteries. However, in order to meet the energy requirements for all-electric vehicles (EVs) and grid energy storage, there is an urgent need to develop alternative cathode materials which can offer a significant increase in energy density and capacity, while also reducing costs and environmental detrimental footprints of current cathode materials. Sulfur, which is the 5 th most abundant element on Earth, has gained special attention as a promising cathode material owing to its theoretical specific capacity of 1675 mA h g -1 , which is the highest among solid elements, and around tenfold the capacity traditional cathodes offer. In addition to the high specific capacity of sulfur, transitioning from traditional insertion cathodes to sulfur-based counterparts offers several additional benefits, namely, low operating voltage of 2.15 V vs Li/Li + , which enhances overall safety of the cell, along with a great difference in price and environmental impact between sulfur and costly toxic transition metals such as cobalt, nickel and manganese. The aforementioned increase in capacity over conventional insertion compound cathodes can enable the assembly of Lithium-Sulfur batteries (LSBs) with a high energy density of 400-600 Wh kg , which is two to three times greater than that of current Li-ion batteries, and can potentially achieve the coveted target range of 500 km between charges for EVs. The low cost and high energy density also make sulfur cathodes appealing for grid energy storage for renewable energies such as solar and wind, provided that high system efficiency and long cycle life can be achieved. |
659bbb7be9ebbb4db9ada167 | 1 | Despite the massive advantages LSBs offer, significant challenges must be still overcome to ensure their seamless industrial integration. Both S8 and Li2S exhibit intrinsically poor ionic and electronic conductivities, which lead to poor discharge/charge performances. Just as importantlythe discharge products of LSBs include highly soluble lithium polysulfides (Li2Sn, 2 ≤ n ≤ 8), which, besides being lost during battery cycling operation with a concomitant capacity loss, can diffuse through the porous separator and detrimentally react with the lithium anode. This process, known as the shuttle effect, results in the loss of active sulfur and leads to rapid capacity fading and poor cycling stability. Furthermore, the high reactivity of sulfur and the polysulfide intermediates can cause severe electrode corrosion, leading to electrode dissolution, and a decrease in the battery's overall capacity and lifespan. Moreover, the large volume expansion and contraction of the sulfur cathode during charging and discharging cycles can result in electrode pulverization, and a loss of electrical contact between the active material and the current collector. To overcome these challenges, various strategies have been proposed and investigated, including the use of advanced electrolytes, the development of nanostructured cathodes, and the use of protective coatings and binders for both the sulfur-based cathodes, the separator, as well as for the Li anode. |
659bbb7be9ebbb4db9ada167 | 2 | Another approach for tackling the current major limitations of LSBs, has been the preparation of carbonsulfur composites and their implementation as a cathode material. Carbon-sulfur composites can mitigate the issue of poor electrical conductivity of elemental sulfur by providing a conductive network of carbon, which can also act as a buffer to accommodate the volume changes during cycling. Additionally, the use of carbon-sulfur composites can minimize the dissolution of polysulfide intermediates and improve the overall cycling stability and capacity retention of the battery, through several proposed mechnisms. While these composites offer several advantages, other difficulties related to the stability of the carbon-sulfur interface have shown to be quite significant, and can lead to the detachment of sulfur particles and loss of active material during cycling. This can result in a decrease in the overall capacity and cycling stability of the battery. Moreover, the incorporation of batterygrade carbon and binder materials also leads to a massive increase in the overall preparation complexity and cost of the resulting cathodes, which can limit their practical applications. Furthermore, the syntheses of such composites frequently employ complex multiple-step energy-intensive methods that are economically viable. |
659bbb7be9ebbb4db9ada167 | 3 | In this work, a novel approach is presented that effectively addresses the persistent detrimental issues associated with conventional sulfur and carbon-sulfur-based cathodes. Unlike many previous studies that have focused on solvent or separator engineering or chemical modification, this method directly targets the creation of a remarkable sulfur-graphenic composite material that overcomes these problems, in a single preparation step. The cornerstone of this approach lies in the precise control of molecular sulfur distribution within the in-situ formed graphene structure. Extreme unprecedented sulfur homogeneity is achieved through a unique, low-power laser treatment of a readily prepared carbon precursor. This innovative process results in the formation of a highly conductive sulfur-graphenic composite materials, where sulfur is molecularly-dispersed in the formed 3D graphenic matrix as individual sulfur molecules. This dispersion strongly anchors and sterically confines sulfur molecular adducts within the formed graphenic composite, chemically immobilizing them to act as molecular traps. As a result, the formation and diffusion of polysulfides are efficiently inhibited. |
659bbb7be9ebbb4db9ada167 | 4 | The experimental results clearly showcase the remarkable improvements in electrochemical performance achieved through this approach. These include enhanced capacity, sulfur loadings and exceptional cycling stability, all of which underscore the efficacy of this system in conquering the persistent challenge of polysulfide-related performance degradation. Crucially, this work stands apart from previous research as the sulfur-graphenic active matrix directly overcomes longstanding challenging issues related to sulfur as a cathode material. This innovative approach not only addresses these issues, but also illuminates a promising pathway for the advancement of Li-S battery technology. |
659bbb7be9ebbb4db9ada167 | 5 | To ensure the preparation of a uniform mixture, a slurry comprising 10 mL of phenolic resin and 3 wt/wt % of sulfur powder was combined in an ice bath using an ultrasonicator operating at a frequency of 15 kHz for a duration of 5 minutes. Subsequently, thin layers of this slurry were applied onto a standard stainless steel (316L/304) current collector with a diameter of 15 mm and thickness of 0.5 mm using a spin-coating technique, followed by drying on a hot plate. The resulting dried films were subsequently subjected to near-UV laser irradiation (450 nm, 2.8 W, Zmorph Fab) using various lasing parameters such as power, speed, defocusing, and raster spacing to fabricate the graphene/sulfur cathodes. |
659bbb7be9ebbb4db9ada167 | 6 | Microscopy and energy-dispersive X-ray spectroscopy (EDS) images were captured using a scanning electron microscopy (Model: FEI Quanta 200 FEG, Make: Joel Co.). Raman Spectroscopy tests of the sample were performed using the InVia TM confocal Raman Microscopy. Transmission Electron Microscopy (Fei Themis Z G3) was performed by extracting a lamella from the sample using a ThermoFisher Helios 5 UC focused ion beam system (FIB), further applied for the analysis of the crystallographic structure and lattice spacing. X-ray photoelectron spectroscopy (XPS) studies were carried out utilizing a Scanning 5600 AES/XPS multi-technique system (PHI, USA). |
659bbb7be9ebbb4db9ada167 | 7 | The electrochemical performance of the graphene-sulfur composite cathodes was evaluated using a CR2025 coin-cell configuration. Batteries were assembled in an inert glovebox (<0.1 ppm O2) using the pre-made cathodes, separator, and lithium metal as the counter-electrode. The electrolyte used in all experiments was 80 μL of 1 M LiTFSI in 1:1 DOL:DME. The separator used was cut from Whatman® glass microfiber filters (Grade GF/B) and the lithium metal discs (15 mm in diameter) were purchased from S4R, France. All the assembled cells were kept at 30 °C throughout their cycling. Electrochemical measurements were conducted using a Bio-Logic BCS battery cycler. |
659bbb7be9ebbb4db9ada167 | 8 | In recent years, there has been considerable focus on the advancement of carbon-sulfur composites as cathode materials for Lithium-Sulfur Batteries (LSBs). This attention stems from their notable advantages over elemental sulfur cathodes, primarily attributed to their enhanced conductivity, impediment of the shuttle effect, and capacity to accommodate volume expansion. However, the successful integration of these composites into LSBs, as well as LSBs in general, necessitates meticulous optimization throughout the synthesis and manufacturing processes to address numerous challenges that detrimentally impact the final product, thereby leading to subpar performances. |
659bbb7be9ebbb4db9ada167 | 9 | Moreover, minimizing the loss of sulfur content during the light-induced heating process is a critical aspect that demands thorough attention. Hence, during the optimization process of our approach, both factors were methodically taken into account to ensure the desired outcomes. Through rigorous testing, a comprehensive set of parameters, encompassing lasing power, rastering speed, and laser beam defocusing, were identified as key factors in achieving optimal conditions for the transformation of the precursors blend into a sulfur-graphenic matrix, while substantially preserving the original sulfur content within the final solid matrix. These parameters can be consolidated into a single metric known as laser fluence (H, J/cm 2 ). The results of these experiments, depicted in Figure , demonstrate the cross-sectional view of the synthesized cathodes at various selected parameters, as observed through high-resolution scanning electron microscopy (HRSEM). This analysis offers valuable insights into the transformation process itself, mainly through the achieved porosity and sulfur loading concentration post-processing. The impact of laser fluence on the pore structure complexity becomes evident as it gradually increases. The lowest fluence (Figure ) results in a poor pore structure, while significant improvements are observed at the highest fluence values depicted (Figure ). These improvements encompass both pore size and overall porosity, indicating a positive correlation between laser fluence and the development of a more intricate pore structure. An additional factor that can be observed with the increased fluence is the depth of the carbon precursor's layer transformation into a conductive graphenic network, which is imperative to ensure optimal electrical contact with the current collector. |
659bbb7be9ebbb4db9ada167 | 10 | The pore size in the synthesized cathodes plays a crucial role in inhibiting the shuttle effect, which, as mentioned previously, is a significant challenge this work intends to overcome. Large pores can allow for the diffusion and loss of polysulfide species, facilitating their migration and subsequent reactions with the lithium anode. By reducing the pore size, it is possible to confine the polysulfide intermediates within the cathode structure, hindering their diffusion and migration. This confinement can effectively inhibit the shuttle effect, leading to improved cycling stability and higher energy efficiency. |
659bbb7be9ebbb4db9ada167 | 11 | Additionally, a fine-tuned pore size distribution can also promote better electrolyte infiltration, facilitating ion transport and ensuring more uniform sulfur utilization throughout the cathode. As such, the fine pore structure achieved via this method is highly desirable, and the keyset of parameters used to synthesize the cathode depicted in Figure were used throughout the work. A top view of the synthesized cathode's surface is depicted in Figure , as well as a close-up view in Figure . |
659bbb7be9ebbb4db9ada167 | 12 | Exploring the scalability of the process was imperative to validate applicability in the market, and as such, large cathodes were prepared on aluminum foil current collector using identical processes as shown in Figure . The following figures (Figures ) present mechanical testing such as twisting and rolling, which had no residual effects on the prepared cathodes, further cementing the promise of the technology. To study the distribution of sulfur within the carbon matrix, Transmission Electron Microscopy (TEM) coupled with Energy-Dispersive X-ray Spectroscopy (EDS) was employed. The analysis began with the preparation of a sample, involving the extraction of a lamella from the surface of the prepared sulfurgraphene cathode using Focused Ion Beam (FIB). Figure showcases the resulting lamella, which was thinned to under 100 nm via a Ga-ion beam. To shed light on the elemental composition of the lamella, EDS maps of the corresponding elements was created as shown in Figure and, revealing the uniform distribution of carbon and sulfur elements, respectively, in which a perfect overlap between the two is exhibited. The lack of detectable domains of sulfur points to the fine distribution of highly dispersed sulfur molecular moieties within the carbonaceous matrix, which is further supported by the lack of crystallinity (shown via the FFT pattern in Figure ), as well as by higher magnification images and EDS analyses showcased in Figure through 2h, in which the uniform distribution is further demonstrated. This hints to the complete encapsulation of the sulfur molecular species content in the cathode by the graphenic carbon matrix, which has been vastly reported as an excellent inhibition method for the dissolution of polysulfides species in the electrolyte, further corroborating the efficacy of the cathode synthesis method. The resulting cathode surface was further analyzed using Raman spectroscopy, to verify the presence of sulfur and characterize the carbon species formed on the current collector, and is displayed in Figure . The Raman spectra revealed two prominent peaks located between 1340 cm -1 and 1582 cm -1 , corresponding to the D and G bands, respectively, and a peak at 478.52 cm -1 indicating the presence of sulfur. The observed D band is attributed to the presence of defects and amorphous-like domains in the carbon network, while the G band is indicative of the degree of graphitization. The ratio of the intensity of the D and G bands (ID/IG) reflects the level of structural defects within the carbon material, with increasing ratios indicating a greater degree of defects. Analysis of the ID/IG ratio for the present sample revealed a value of 0.859, indicative of graphene growth with high amount of grain boundaries and significant structural defects. This can be explained by the high temperatures caused by the laser irradiation-induced graphene formation, as well as the high sulfur content molecularly dissolved in the resulting S/C matrix. |
659bbb7be9ebbb4db9ada167 | 13 | The cathodes were also subjected to X-ray photoelectron spectroscopy (XPS) analysis, to shed further light on the surface chemical states and composition of the various species. The XPS survey spectrum (Figure ) revealed the presence of S, C, N, and O at binding energies of 163 eV, 285 eV, 400 eV, and 533 eV, respectively. Curve fitting of the C 1s spectrum (Figure ) revealed the presence of two chemical species at binding energies of 284.6 and 286.2 eV, corresponding to C-C/C-H and C-O functional groups, respectively. Based on their binding energy, the formed material was identified as graphene oxide-type product, as already reported in the literature. The deconvoluted S 2p spectrum (Figure ) showed the presence of multiple peaks at different binding energies. The doublet at 163.6 and 165.0 eV can be attributed to S8 molecules, whereas the peaks appearing at 168.9 and 171.2 eV can be associated with highly oxidized SO4 groups. |
659bbb7be9ebbb4db9ada167 | 14 | To ensure homogeneous mixture preparationa slurry consisting of 10mL of phenol formaldehyde resin and Sulfur at a desired wt/wt % content were mixed in an ice-bath using an ultrasonicator for 5 minutes at a frequency of 15 kHz. Thin layers of said slurry were spin-coated at 3000 RPM for 30 seconds onto a standard stainless steel (316L/304) current collector (15 mm diameter, 0.5 mm thickness. |
659bbb7be9ebbb4db9ada167 | 15 | The current collectors were then subjected to soft baking at 100℃ for 1 minute, in order to remove any residual solvent. The dried-films were later exposed to near-UV laser irradiation (450nm, 2.8W, Zmorph Fab) at various lasing parameters of power, speed, defocusing and raster spacing to create the sulfur-graphenic cathodes. Please note, that higher, up to 70 wt/wt % of Sulfur/Resin, content ratios can be currently applied for further increasing cell capacity if so required, along the preparation of one-step slurry-based single layer deposition of higher thicknesses, by well-known slurry-based film formation approaches. |
659bbb7be9ebbb4db9ada167 | 16 | CR2032 coin cells, containing the composite sulfur-graphene cathode, a separator, lithium metal, and a spring, were assembled after drying in a glovebox (<0.1 ppm O2). The separators were Whatman ® glass microfiber filters (Grade GF/B) punched into 18 mm discs in diameter, and the lithium metal discs (15 mm in diameter) were purchased from S4R, France. 80 μL of 1 M LiTFSI in 1,3-dioxolane (DOL)/dimethoxy ethane (DME) (1:1) as an electrolyte was used in all of the experiments (Sigma-Aldrich). The cells were kept at a constant temperature of 25 °C throughout all the experiments. |
659bbb7be9ebbb4db9ada167 | 17 | Electrochemical characterization of the sulfur-graphenic composite cathodes was conducted in order to assess the performance of the material as a potential cathode in LSB applications. Figure and show the cycle life of a typical cell comprised of the laser-induced cathode material, which were cycled at 0.07 and 0.7 A g -1 , respectively. The cell which ran at low rates exhibited excellent reversible capacity of 1000 mAh g -1 , with an average coulombic efficiency of 99 %, and capacity retention of 73.1% over 420 cycles. The cell portrayed in Figure , which was charged/discharged at significantly faster rates, showed initial reversible capacities of 575 mAh g -1 , and over 1500 cycles maintained remarkable capacity at a rate of 73.9%. |
659bbb7be9ebbb4db9ada167 | 18 | The remarkable electrochemical performance exhibited by the highly porous sulfur-graphene cathodes can be attributed to a combination of factors, including their unique porosity and the complete encapsulation of molecularly-dispersed sulfur within the graphene matrix. These characteristics synergistically contribute to the enhanced electrochemical properties of the resulting composite material. The high porosity of the graphene framework plays a crucial role in facilitating efficient electrolyte penetration and diffusion throughout the cathode structure. The interconnected network of pores offers a large surface area, enabling extensive contact between the electrolyte and the sulfur species embedded within the graphene matrix, as shown previously. This favorable electrolyte accessibility promotes swift ion transport, and facilitates the electrochemical reactions at the electrodeelectrolyte interface. As a result, the charge transfer kinetics are significantly improved, leading to enhanced overall electrochemical performance. More importantly, the complete encapsulation of molecularly-dispersed sulfur within the graphene matrix provides crucial benefits. By encapsulating the sulfur species, graphene acts as a physical barrier, preventing direct contact between sulfur and the electrolyte. This encapsulation strategy effectively addresses the issues associated with sulfur dissolution and the shuttle effect of polysulfides, which normally leads to rapid capacity fading and generally hindered cycling stability. The encapsulation shields the sulfur from undesired reactions, ensuring its stability and preserving its high theoretical capacity over repeated charge-discharge cycles. |
659bbb7be9ebbb4db9ada167 | 19 | Collectively, the highly porous nature of the graphene/sulfur cathodes, combined with the complete encapsulation of sulfur, engenders superior electrochemical performance. These characteristics make the graphene/sulfur composite cathodes highly promising for advanced energy storage applications, propelling the development of high-performance lithium-sulfur batteries. To further cement the promise of this approach, thicker cathodes, consisting of multiple layers have been assembled, by repeating the process used to manufacture the single-layered cathode after each ablation process, effectively mimicking additive manufacturing (3D printing), in order to achieve areal capacity market-end goals. |
659bbb7be9ebbb4db9ada167 | 20 | Two cells were assembled, consisting of 6 and 12 layers (SEM images of which are showcased in The rate performance of the cells was evaluated by cycling an identical sulfur-graphene cell, and was achieved by increasing the charging/discharging current densities incrementally every 8 cycles, between 0.71 and 28.57 A•g -1 . |
659bbb7be9ebbb4db9ada167 | 21 | To assess the ability of the prepared sulfur-graphene cathodes in suppressing polysulfides formation In order to further evaluate the efficacy of the cathodes in inhibiting the shuttle effect, the samples were then subjected to thorough analysis using Inductively Coupled Plasma Mass Spectrometry (ICP-MS), a highly sensitive technique capable of detecting and quantifying various elements present in a sample. |
659bbb7be9ebbb4db9ada167 | 22 | In this case, the primary focus was on verifying the presence of sulfur species within the electrolyte fractions. The raw results are shown in Figure , in blue. The pristine electrolyte contains sulfur species to begin with, and as such the data was normalized against it, and is shown in orange. Notably the sulfur content in the electrolyte did increase ever-so slightly, however due to the harsh cycling conditions and the overall miniscule increase in sulfur, these results demonstrate the promising potential of our sulfur-graphenic cathodes for Li-S applications. This result, combined with the straightforward synthesis method and the previously demonstrated favorable electrochemical performance, underscores the viability and appeal of this cathode design for advancing the field of lithium-sulfur batteries. |
659bbb7be9ebbb4db9ada167 | 23 | Postmortem analysis played a crucial role in unraveling the effects of sulfur-graphenic cathodes on their performance, and shedding light on various aspects of their behavior. Further analysis of the cell showcased in Figure (>1500 cycles), enabled a detailed examination of the sulfur-graphenic cathodes post-cycling, which allowed to assess any structural changes or degradation which may have occurred during the cell operation procedure, the results of which can be seen in Figure . weight percent of 2.6% and 6.7% respectively, as shown in the Table . |
66c4376c20ac769e5f22ac9e | 0 | Lithium hexafluorophosphate (LiPF 6 ) is the standard lithium salt for the state-of-the-art commercial lithium ion batteries (LIBs). LiPF 6 is the obvious winner over other lithium salts due to the virtue of its higher ionic conductivity (~10 -2 Scm -1 at 25ºC), excellent anticorrosion nature toward aluminum, and good anodic stability (> 4.2V vs. Li/Li + ) . |
66c4376c20ac769e5f22ac9e | 1 | Nevertheless, there remains longstanding concerns over the safety of LiPF 6 based lithium salts for LIB applications particularly at higher temperatures. LiPF 6 is unstable and moisture sensitive at higher temperatures (>60 ºC), which degrades to other decomposition products, and this eventually results in capacity fading and thermal runaway of the corresponding LIBs . For these reasons, therefore, there has been considerable research interest to replace the [PF 6 ] -anions with novel anions, that have better thermal stability and enhanced cycling performance. . Ionic liquids (ILs) based on bis(trifluoromethanesulfonyl)imide [TFSI] - and fluorosulfonyl(pentafluoroethanesulfonyl)imide [FSI] - are leading candidate anions for LIB electrolytes as they generally exhibit better thermal stability, relatively low viscosity and higher ionic conductivity . The [TFSI] -anion has been designated as a benchmark anion in ILs by IUPAC. As a result, the [TFSI] -anion has been the focus of several experimental and theoretical studies . There are several benefits of using the [TFSI] -anion containing ILs for LIB electrolytes. For example, lithium bis (trifluoromethanesulfonyl)imide salt (LiTFSI) has good thermal , chemical and electrochemical stability, which makes it a good candidate for LIBs. In addition, LiTFSI readily dissolves in various kinds of electrolytic solvents, which gives higher ionic conductivity at room temperature (e. g., ca. 10 -2 Scm -1 in EC/DMC). Nevertheless, LiTFSI is corrosive towards Al current collector, limiting its application as main conducting salt for LIBs . Another structural analogue of the [TFSI] -anion is the [FSI] -anion. The [FSI] -anion has been studied as a component of ILs . The [FSI] -anion containing ILs exhibit lower melting points, lower viscosities, and higher conductivities compared to [TFSI] -based ILs . For a given type of solvent, the LiFSI based electrolytes generally have higher ionic conductivities (~25 ºC) than LiPF 6 and LiTFSI based conducting salts . Therefore, the [FSI] -anion based conducting salts are worth of scientific researches as essential component of high performing LIBs . As such, LiFSI has been implemented as a co-salt to improve the performance of contemporary LIBs, particularly in electric vehicles (EV) field . |
66c4376c20ac769e5f22ac9e | 2 | ILs generally have some desirable features as electrolyte components for LIB. However, ILs also have undesirable features such as low wettability of the electrode surface and higher viscosity. The physicochemical properties of ILs can be enhanced by modifying ILs into hybrid electrolytes, either with lithium salts and/or organic solvents, which can effectively reduce the viscosity and ionic transferring resistance, and this thereby improves the ion transport and interfacial properties . One approach is to add a lithium salt with a different anion than in the original IL . For example, replacing [TFSI] -by [FSI] -, or combining the two, has been motivated by enhanced ionic conductivities and improved passivation of the aluminium current collector . Several IL based electrolytes with an imidazolium cation ([EMI] + ) have been investigated trying to find beneficial effect of mixing [FSI] -with [TFSI] -anions. A synergetic effect of having mixed anions was possible with the LiTFSI/[EMI][FSI] electrolyte giving the best overall performance . Another popular strategy is to mix ILs with organic solvents to form hybrid electrolytes with improved performance by decreasing the viscosity. Generally, the addition of ILs into organic solvent electrolytes was seen as a viable option to improve the electrochemical stability, which in most cases, is higher than organic solvent based electrolytes. For example, mixing of 70 wt% of 1-butyl-1-methylpyrrolidinium hexafluorophosphate ([Py 14 ][PF 6 ]) IL with conventional standard electrolytes based on 1M LiPF 6 /EC/DMC electrolyte solutions was found to improve the flash point of the electrolyte from 22˚C to 51˚C, while at the same time maintaining the ionic conductivity (10 -3 Scm -1 at 25ºC) . Over the past several decades, various ILs, in combination with organic solvents, have been proposed and investigated as electrolyte components for LIBs. Such approaches could represent a good compromise between fast ion transport properties (low viscosity) and improved safety (low flammability) of the resulting electrolyte . The incorporation of proper organic compounds into ionic liquid electrolytes was found to enhance the cycling performance of LIBs . Overall, the mixed ILs/organic solvent liquid electrolytes have the potential to be applied to LIBs with high output voltage at elevated temperatures, while at the same time satisfying the requirement of LIB safety. However, a compromise would exist between thermal stability and electrochemical performance at elevated temperatures. Thus, a great effort needs to be put on the adjustment of the proportion of ILs and organic solvent. |
66c4376c20ac769e5f22ac9e | 3 | Nevertheless, ongoing from pure ILs to IL mixtures, there is an obvious increase in complexity arising from the presence of more ions. Therefore, molecular investigations on IL mixtures are often more difficult than pure ILs. The theoretical and experimental investigations on IL mixtures are often difficult in comparison with pure ILs. Hence, the progress of designing and utilizing the IL mixtures with more extraordinary properties will depend on advancing an in-depth understanding of their physical and chemical properties as well as the interactions between different cations and anions. At present, there have been a number of previous experimental studies on thermodynamic and transport properties of imidazoliumbased IL mixtures . It is particularly important to understand how the structural properties of IL mixtures change upon mixing, as well as the dependence of the properties of the resulting mixtures on their composition. The properties of IL mixtures arise from different cation-anion, cation-cation, and anion-anion interactions, including Coulomb, van der Waals as well as hydrogen bond interactions. . For example, a major finding of recent experimental studies reveals that the structure of IL mixtures is mainly determined by the random distribution of ions driven by Coulomb interaction . Due to the limited number of experimental studies on such interaction mechanisms, little is known on the origins of the thermodynamic and transport properties of imidazolium based IL mixtures. To date, molecular dynamics (MD) simulations in combination with experimental approaches have provided good insight into the structure and dynamics of ILs . Because of higher accuracy, ab initio molecular dynamics (AIMD) simulations may add valuable insight into the structures of solvation shells of ionic salts in a variety of organic solvents, but they require relatively higher computational demands. Due to the limited time-and size-scales of DFT methods, and high cost of AIMD simulations, classical molecular dynamics (CMD) has largely been the method of choice for investigations of IL structure . Early imidazolium based ionic liquid force fields by the groups of Maginn , Berne and Stassen helped shape initial thoughts as to the origin of their unique solvent properties. |
66c4376c20ac769e5f22ac9e | 4 | With this in mind, therefore, the purpose of the current study is to present detailed dynamical insights of such hybrid electrolytic systems, and their mixtures with ethylene carbonate (EC) and dimethyl carbonate (DMC), which are potential candidates for next generation LIB electrolytes. We report on the molecular level study of the self-diffusion coefficients, molar conductivity and transference numbers, and discuss on the influence of adding carbonate cosolvents, and explore the nature of these molecular transport properties. The article is structured as follows. First, we give a brief overview over the computational methods employed in this work. This is followed by the discussion of the mean square displacement of the ions (MSD). The details of self-diffusion coefficients (D s ), molar conductivity (ᴧ), and transference numbers are discussed by analyzing trajectories, and how this behavior is influenced by the EC/DMC content. We end this article with conclusions. |
66c4376c20ac769e5f22ac9e | 5 | In this work, MD simulations were performed using LAMMPS (Large Scale Atomic/Molecular Massively Parallel Simulator) . All potential models used to describe intermolecular interactions were based on the OPLS-AA force field (Optimized Potentials for Liquid Simulations/All Atom) . Although MD simulations using nonpolarizable force fields provide important insight into molecular level correlations and structure in ionic systems . However, nonpolarizable force fields do have poor dynamics , which has led to critical evaluations. As an alternative to the nonpolarizable force fields, methods such as AIMD , and polarizable force fields have varied advantages over their nonpolarizable counterparts, but this comes at a cost of higher computational demands. |
66c4376c20ac769e5f22ac9e | 6 | Alternatively, the cost-effective approach to account for polarization effects is to scale the atomic charges to mimic the average charge screening caused by the polarization as well as the charge transfer effects . In this work, charges are downscaled to an absolute value of 0.8 to account for the charge transfer and polarizability within ILs. Charge scaling is an efficient way to approximately include polarizability and charge-transfer effects in classical simulations at no further computational cost . |
66c4376c20ac769e5f22ac9e | 7 | The selected solvents were 1-ethyl-3-methylimidazolium bis(trifluoromethanesulfonyl) imide [TFSI] ionic liquid, ethylene carbonate (EC), diethyl carbonate (DEC) and 1M LiFSI as lithium salt. The bonded and non-bonded parameters for EC and DMC were obtained from the OPLS-AA force fields . The cations and anions of ILs ([EMI] + , [TFSI] -) and [FSI] - were modeled using parameters of the systematic all-atom force field of opes and adua he i ion parameters were adopted from ion-water potential calculations Compositions of the simulated ionic liquids are shown in (1.52 g/cm 3 ). Periodic boundary conditions were applied on all three sides of the cubic simulation box to represent the bulk solution. The system was simulated at temperature, T = 298 K and pressure, P = 1 atm. We used the velocity erlet integrator with a timestep of 1 fs to integrate the e uations of motion he ose-oover thermostat and barostat were used to keep the temperature and pressure constant. We used PACKMOL to create initial configurations. The Ewald method was used for treating electrostatic interactions. |
66c4376c20ac769e5f22ac9e | 8 | To equilibrate the IL solutions, the following procedure was employed. The system was first energy minimized using conjugate gradient algorithm. Then, the system was equilibrated for 4 short runs of 150 ps each in the NVT ensemble, 14 short runs of 150 ps each in an NPT ensemble at 298K, in which the density was recorded over the last 150 ps and averaged.Then, each system was further simulated for 20ns runs in the NVE ensemble. The equilibration was assured by analyzing the system potential energy and relevant physicochemical properties such as density as a function of simulation time. |
66c4376c20ac769e5f22ac9e | 9 | The adjustment of the numerical values of the density of IL/carbonate mixtures could have an effect on the dynamics of the system, and the results could suggest an intriguing point for understanding and designing electrolytes of lithium ion batteries for better performance and safety (see also Chapter 6) . The density of electrolytes in LIB technology can induce a dramatic effect on the dynamics of electrolytes. Fundamental studies of bulk electrolytes indicate that organic liquid electrolytes consisting of EC and DMC can induce larger effect on the diffusive dynamics and density of IL based electrolytes than other liquids . A common strategy to enhance the mobility at a given temperature of IL based electrolytes has been an increasing fraction of linear carbonates . However, increasing the content of linear solvents such as DMC is often limited by the ion-pairing of salts, which results in a decrease in the ionic conductivity. Therefore, it is recommended to find an optimal ratio of EC/DMC co-solvent mixtures to give a maximal ionic conductivity. Although density significantly affects the activation energy for diffusion, it does not, however, affect the 3 , respectively was in good agreement with other reported experimental and theoretical studies10,33 . Our study shows that the calculated value of 1M |
66c4376c20ac769e5f22ac9e | 10 | given time means a faster diffusion dynamics. The Ds of the center-of-mass of all cationic and anionic species for different mixing ratios of IL/carbonate co-solvent mixtures in the presence of 1M LiFSI conducting salt at 298K were calculated using the Einstein relation (Equation .1) through the MSD of the center-of-mass position for each ion . |
66c4376c20ac769e5f22ac9e | 11 | Although these two methods are both rigorous, they are, however, subjected to different numerical errors in finite time MD simulations of finite-size systems. Generally speaking, the numerical values generated from Green-Kubo expression are larger than that of the Einstein relation due to numerical imprecision . In this work, therefore, the values of Ds for different mixing ratios of IL/carbonate blends were estimated using the Einstein relation. The wt%. The relative values of the Ds were found to be in the order D cation > D anion > D Li , which was consistent with reported experimental and theoretical data . The small Li + ions interact strongly with IL anions, which consequently resulted in slower diffusivity. As opposite scenario to these reports of Li diffusion has also been reported that the Ds for the [PF 6 ] -anion was found to be greater than the Li + cation in mixed EC/DMC systems, which was attributed to strong electrostatic interaction between Li + ions and carbonyl oxygen of organic solvents (EC and DMC). The strong interaction between Li + ions and organic solvents causes higher tendency of the organic solvent to solvate Li + ions rather than the [PF 6 ] -anions, which consequently results in solvated-separated ion pairs (SSIPs), and this was found to reduce diffusivity of Li + ions. NMR experiments have shown that cations of the immidazolium family have larger Ds than that of the anions .The planar structure of the imidazolium cations results in greater friction for the translational motions of the anions . Although numerous experimental and computational studies have been reported so far, further exploration is however, needed to understand the properties of IL systems, their microscopic structures and dynamics, and the behavior of Li ions in ILs. |
66c4376c20ac769e5f22ac9e | 12 | The numerial values for the Ds of the cations and anions in ILs are affected by numerous factors. For example, the relative cationic and anionic sizes, geometric shape, the strength of vander Waals and Coulombic interactions between the cation and anion of the corresponding IL are just some of the factors that can affect the Ds of ILs. On top of this, differences in the type of the force fields, equilibration times, length of production runs, and a method for calculating the Ds may have a huge impact on the computed Ds. Therefore, the direct comparison of our computed Ds values with other MD simulations, as well as reported experimental results, is rather difficult. Recent scientific reports have emphasized the effect of the force field on differences in the calculated Ds of ILs. Our own experience from the current work shows that simulation conditions such as the potential truncation at the cutoff radius, finite-size effects due to the use of small simulation cells, numerical imprecision in the integration of the equation of motion, and most importantly, insufficient equilibration of the system before the production phase of a simulation are the most important factors than need to be considered during classical MD simulations . These factors may potentially introduce systematic errors in the MD simulation results, which consequently may give rise to substantial differences in the numerical values of the calculated Ds. Therefore, it requires the careful choice of such simulation parameters in order to eliminate these systematic errors . Most importantly, the choice of the MD simulation method for calculating the Ds is a trade-off between computational costs, the complexity of the IL, the specific property to be computed, the temperature of simulation, and the accuracy required . Because ILs do have large polarizable ions, and thus they require polarizable force fields to represent their behavior. The work of Bhargava and Balasubramanian shows that equilibrium simulations may give Ds values close to experimental values for ILs if a custom designed force field without polarizability is employed. |
66c4376c20ac769e5f22ac9e | 13 | Ionic conductivity is one of the most important properties that characterize the electrochemical performance of electrolytes. Ionic conductivity quantifies how mobile the ions are for the electrochemical reactions . It is determined by the number of ions, the magnitude of the ionic charge and the mobility of ions . For any given ions, therefore, the value the ionic conductivity essentially depends on both diffusivity and the number of ions participating in carrying charges . The larger the diffusivity of ions, the higher will be the value of ionic conductivity. However, formation of ion pairs of cations and anions decrease ionic conductivity because the formation of pairs of cations and anions is closely related with a decrease in diffusivity due to an increasing size of ionic clusters in addition to a decrease in the number of ions contributing the ionic conductivity. Solvents may solvate the ions preventing the cations and anions from forming pairs and even clusters. Solvents can enhance the mobility of ions by forming a proper solvation shell of cations. Generally speaking, solvents of larger dielectric constant enhance the ion pairing process, but this comes at the cost of decreasing the mobility of ions due to its large viscosity. In sharp contrast, solvents of lower dielectric constant can enhance the mobility of ions due to their lower viscosity, but such solvents with lower dielectric constant do not enhance the solvation process of the ions. Thus, no single-solvent strategy satisfies all the requirements of LIB electrolytes. For this reason, therefore, state-of-the-art LIB electrolytes adopt multiple-solvent strategy to compromise both properties: the mobility and the ion-pairing . For example, EC has a large dielectric constant (ε~9 at 4 °C) owever, its high viscosity (η~1 9 c at 4 °C) dis ualifies it from being chosen as a sole solvent On the other hand, DMC has low viscosity (η~ 59 c at 2 °C) but smaller dielectric constant (ε~3 1 at 25 °C). Therefore, a mixture of both EC and DMC has been suggested as a candidate of efficient electrolytes to be satisfied with two important properties . In this work, we explore the effects adding a mixture ratio of EC: DMC=50%:50% co-solvents on the achieviable ionic conductivity of [EMI][TFSI] consisting of 1M LiFSI salt at 298K. We consider the ionic conductivities of these mixtures and investigate how they are affected by solvent properties. In accordance with prior reported studies, we comput the ionic conductivity by using the ideal ionic conductivity as a proxy. The limiting value of the selfdiffusion coefficient of an ion (D Si ) at infinite dilution (denoted by ∞) is given by ernst's relation : |
66c4376c20ac769e5f22ac9e | 14 | where λ i ∞ is the limiting ionic molar conductivity, F and R are the Faraday and gas constants, and T is the absolute temperature. Since the salt molar conductivity is the sum of the ionic contributions, it can be written in terms of the sum of the ion self-diffusion coefficients as: |
66c4376c20ac769e5f22ac9e | 15 | for a solution of a 1:1 salt his is widely known as the ernst-Einstein relation he ernst-Einstein relation relation does not accurately describe the ionic conductivity at a uantitative level he ernst-Einstein relation does not take into account ion-ion interactions, but this is a sufficient approximation to use for the purpose of qualitatively evaluating trends and is commonly employed in the literature . |
66c4376c20ac769e5f22ac9e | 16 | The numerical values of molar ionic conductivity are summarized in [TFSI] Ils have higher total molar ionic conductivity than 1M LiFSI doped pure EC/DMC carbonate solvents. Furthermore, the relative ionic contribution to the total molar conductivity of each individual cationic and anionic species increase with higher contents of IL in the IL/carbonate mixture, showing that higher carbonate co-solvents have the effects of reducing the molar ionic conductivity. The organic solvents restrict the free motion of the ionic species, resulting in a low conductivity for the organic-based electrolytes, while the IL solvents facilitate the ionic association of the ions, causing a high conductivity for the IL-based electrolytes. Additionally, we realized that the molar ionic conductivity changes in the order ᴧ Li > ᴧ EMI > ᴧ FSI > ᴧ TFSI , indicating that Li + ions have the highest molar ionic conductivity. |
66c4376c20ac769e5f22ac9e | 17 | The relationship between ionic diffusivity, viscosity, and molar conductivity for different cationic and anionic structures was highlighted by the works of Watanabe et al and coworkers . Through self-diffusion and conductivity experiments, it has been shown that the diffusion of the ionic species in IL was an anion driven process. In those studies, it has been shown that the high rate of ions contributing to the ionic conductivity within the diffusion component was related to the poor interaction of the anion with the cation. Thus, it is evident from these studies that size, shape and geometry of the anion determine the strength of the electrostatic interaction between the cation and the anion. The effect of the salt concentration on the ionic conductivity in different ILs has also been reported for some IL systems. For example, some interesting properties were observed for [EMI][TFSI]/EC/DEC mixtures . In those studies, it has been shown that the ionic conductivity increased when the IL in the mixtures increases from 0 to around 60%, as would be expected if the "salt" (I ) concentration increases by adding it to the conventional solvent EC/EDC. However, further increase above 60% IL in the mixture, the study reported that the conductivity decreased, due the absence of sufficient organic solvent available to "solvate" all the I present, and the conductivity approached that of pure IL. Furthermore, the addition of EC in to IL based electrolyte exhibited a significant increase of ionic conductivity due to the dilution effect of electrolytes, which helped decrease the ion-ion interaction between Li + and TFSI -anion. |
66c4376c20ac769e5f22ac9e | 18 | The numerical value of the ionic conductivity of ILs does not reflect their ability to conduct Li + ions. Lithium transference number (T Li ) reflects the relative contributions of the charged species to the transfer of the total charge . The higher the value of the T Li , the greater will be the relative contribution to the charge transport. The T Li , i.e., the contribution to the ionic conductivity due to the Li + ion transport, can be approximated from D through an equation [151]: |
66c4376c20ac769e5f22ac9e | 19 | Where Ni is the number of ions of species i, and D i is the corresponding self-diffusion coefficient.This equation is valid where no ion correlations are present, but it is justifiable to use it here to obtain a rough estimate of conductivity of Li + ions for ILs. In this work, Table . > E 1 where as for [TFSI] -it follows in the order G 1 > F 1 > E 1 . The T Li increased with the increase of EC/DMC concentrations. In addition, the T Li transference is higher than T EMI , T TFSI and T FSI for higher contents of EC/DMC. Therefore, the lithium ion dynamics are relatively faster in in the EC/DMC content of IL/ content. |
66c4376c20ac769e5f22ac9e | 20 | No single-solvent electrolyte can satisfy all the requirements of LIB electrolytes. For this reason, therefore, LIB electrolytes are often complex mixtures of multiple co-solvents with optimized properties. Here in, we present the effects of different mixing ratios of the IL solvents, the computed density of the pure IL/1M LiFSI system (no EC/DMC content) is higher than that of the EC/DMC/1M LiFSI system (no [EMI][TFSI] content). The relative increase in density with higher contents of IL/1M LiFSI system is suggestive of arising from the stronger interactions of [TFSI] -and [FSI] -anions with the Li + ions as compared to the interaction of EC/DMC with Li + ions. The behavior of the MSDs for the center-of-mass of the ions as a function of IL/carbonate co-solvent mixtures indicated that the ions exhibited slow dynamics (diffusity) with higher carbonate content. The observed MSD curves show that ions exhibite faster dynamics (diffusity) with higher IL content. Our study on the diffusion coefficient analysis of Li + , [FSI] -and [TFSI] -ions have revealed that the organic solvents restrict the free motion of the ions, reducing the dynamics (diffusity) of the electrolytes. In this work, the Ds values predicted for the ions in the IL/carbonate blends were found to be of the order of 10 -12 m 2 /s. The Li + ions have higher diffusion rates in the [EMI][TFSI]/EC/DMC mixture than all other iomic species present. This finding is in sharp contrast with previously reported studies which indicated that the D s values change in the order D EMI > D TFSI > D Li . In those studies, the Li + ions had the lowest Ds values, and the [EMI] + cations have the highest value.Our finding shows that the diffusion rate of Li + ions changes with the concentration of the EC/DMC co-solvents. Our results also show that the diffusion rate of the [FSI] -anions is higher than [TFSI] -anions. In all cases, the diffusion rates of the cation are higher than those of the anions, which is consistent with experimental data. Our MD simulation results revealed that the total molar ionic conductivity for the different mixing ratios of IL/carbonate blends decrease with higher contents of EC/DMC cosolvents. Our results show that 1M LiFSI doped [EMI][TFSI] Ils have higher total molar ionic conductivity than 1M LiFSI doped pure EC/DMC carbonate solvents. Further more, the relative ionic contribution to the total molar conductivity of each individual cationic and anionic species increase with higher contents of IL in the IL/carbonate mixture, showing that higher carbonate co-solvents have the effects of reducing the molar ionic conductivity. The organic solvents restrict the free motion of the ionic species, resulting in a low conductivity for the organic-based electrolytes, while the IL solvents facilitate the ionic association of the ions, causing a high conductivity for the IL-based electrolytes. Additionally, we realized that the molar ionic conductivity changes in the order ᴧ Li > ᴧ EMI > ᴧ FSI > ᴧ TFSI , indicating that Li + ions have the highest molar ionic conductivity. The numerical value of the ionic conductivity of ILs does not reflect their ability to conduct Li + ions. Lithium transference number (T Li ) |
6775e7f3fa469535b97ec414 | 0 | First principles thinking (or reasoning from first principles) is a way of thinking and problem-solving that breaks down a complex problem into its most basic assumptions, facts, concepts, or ideas to gain new knowledge and find new solutions through reassembling them from the bottom-up. There were attempts to introduce this way of thinking into engineering education in recent years . Problem-based learning, a student-centered approach in which students learn a subject by working in groups to solve an open-ended problem, offers great platforms for utilizing first principles thinking in deepening their learning and producing solutions. Here we reported the elaborate design of two PBL projects to facilitate the students´ learning by practicing first priciples thinking in two PBL small student groups. The aim of the design was to enable students to develop the first principles thinking by deconstructing problembased learning projects to their fundamental concepts and rebuilding their solutions from the ground. |
6775e7f3fa469535b97ec414 | 1 | According to Aristotle, new knowledge is known from very basic concepts or assumptions that are first principles, thus 'the first principle is the basis from which a thing is known' and the building block of true knowledge . For our daily life and work, the most popular and usual way of thinking is to think/reason by analogy which is proven to be very useful and easy. However, 'first principles thinking' (or 'reasoning from first principles) is a way of thinking and problem-solving requiring breaking down a complex problem into its most basic assumptions, facts, or ideas and then reassembling them from bottom-up. In other words, it is a way to build up the solution and knowledge from the foundational truths through rational reasoning other than analogous reasoning. Rational reasoning is built based on logic, consistency, and evidence, which include deductive or inductive processes. However, analogous reasoning draws conclusions based on similarities between two things or situations, inferring that what is true in one case may also be true in the other, which might fall to 50 logical fallacies and errors. |
6775e7f3fa469535b97ec414 | 2 | In first princinples thinking, these foundational truths are basic building blocks like Lego pieces, with which we can create and build up marvelous structures and architectures with great freedom and creativity. First principles thinking can be applied to solve problems in our daily life, and also complex problems in science, technology, and engineering. It is regarded crucial to innovation and creativity . |
6775e7f3fa469535b97ec414 | 3 | The problem-based learning (PBL) pedagogy provides an ideal platform to use first principles thinking for students to solve problems and work on projects. Here we explored how first principles thinking was applied in PBL in chemical education carried out at the Department of Chemistry and Bioscience through a one-year pedagogical program provided to researchers at the Aaborg University (AAU). |
6775e7f3fa469535b97ec414 | 4 | Students were divided into small groups to solve problems of patient cases or biomedical phenomena and practice, and applying knowledge and skills . In Denmark, two young universities, Roskilde 65 University and Aalborg University (AAU) were founded in the 1970s, initiating the innovative PBL pedagogy . At AAU, the Aalborg Mode has been established, characterized by a syllabus of 50 % percent problem-centered project work and 50 % traditional lecture in an interdisciplinary research environment . The PBL approach is organized as group project work and is implemented from the first semester until the completion of the master's degree. Throughout the university program, the group 70 sizes generally decrease. Initially, groups typically consist of 6-7 students in the first year, gradually reducing to a maximum of 2-3 students in the final semester . Learning is thus organized around problems and naturally frames a student-centered pedagagy to drive the students to be autonomous learners. The other benefits for students include the potential to reach a high academic level in interdisciplinary areas, the development of the team-work capacity, and better preparation for the labor market. One special advantage of PBL is its great relevance to the practice of the project work, thus allowing students to move from theoretical knowledge to practical application seamlessly. Logically, PBL offers an excellent pedagogical platform for employing first principles thinking to enhance students' learning and their capacity to solve real-world problems . |
6775e7f3fa469535b97ec414 | 5 | It was reported that syllabus coverage of fundamental theory and knowledge could sometimes be insufficient in Problem-Based Learning (PBL), resulting in the so called "holes" in the conceptual knowledge base . At the Department of Chemistry and Bioscience in AAU, most of the PBL projects in the chemistry section are research projects, especially experimental projects related with the on-going research that are mainly in the area of applied supramolecular chemistry, disordered materials, and separation processes (). The training in experimental skills and the operation of instruments are crucial and indispensable for completing PBL projects. In most cases, students can effectively carry out their projects in an analogous way by applying the experimental skills and instrumental operation techniques. However, gaps often exist between the theoretical foundation and the practical requirements for performing research projects, highlighting the need to bridge this divide during the learning process. They may know well the "hows" in conducting the project, but not need to know "whys" from the the aspects of fundamental principles. Considering the syllabus of 50 % traditional lecture, the learning of fundamental theories and knowledge is insufficient for students, which also limits their learning depth in chemistry and the exertion of their creativity. Overall, the general pedagogical problem that we targeted was to strengthen in-depth active 95 learning of the fundamental principles and excite creativity in PBL. First principles thinking has recently been introduced into engineering education , which offered a solution to improve the depth of learning and creativity. Hereby we described our recent practice and achievements in strengthening first principles thinking in PBL of chemistry. |
6775e7f3fa469535b97ec414 | 6 | In the fall semester of 2022, two projects were designed. The first project (Project 1: Kinetic modeling of chemically fueled assemblies) was to build kinetic models from experimental data in the domain of theoretical chemistry for the 3 rd semester students for 4 months. For the details of chemically fueled assemblies, several research works are referenced . Bacially, to construct chemically fueled assembly, the building elements should include an activation/deactivation reaction cycle, a non-assembling precursor, and a self-assembling product (Figure ). Specifically, a molecular precursor that cannot self-assemble is converted to a product that can self-assemble into ordered structures after reacting with a high-energy molecule. This conversion reaction is the activation reaction and the high-energy molecule is the chemical fuel that is irreversibly converted to waste. The deactivation reaction which converts the product back to the disassembled precursors constitutes the second reaction in the activation/deactivation reaction cycle. Consequently, the assemblies are only present for a finite time until the chemical fuel is depleted and can be regenerated by resupplying the chemical fuel. Sustaining the non-equilibrium assembly thus requires continuous input of the chemical fuel. Such transient assemblies are accessible to accurate temporospatial programming by controlling the kinetics of the reaction cycles. The participating students at the same time were taking the courses of physical chemistry and organic chemistry. This PBL project required the students to build the kinetic model based on the understanding of the fundamental principles in interdisciplinary area of physical chemistry and organic chemistry. |
6775e7f3fa469535b97ec414 | 7 | The second project was designed for 3 students from Ecole Technique Supérieure de Chimie de l'Ouest in France for two months. They exchanged to AAU with the support of Erusmus+ project for 2-month traineeships in the winter of 2022. They were in the level of 2 nd year of the 'brevet de technician supérieur' (an equal of the second year in college). They applied for the internship in our lab to gain some experimental techniques in organic synthesis. As they had taken the organic chemistry course in France, thus they were grouped into a small PBL group. A PBL project (Project 2: Synthesis of a peptide aldehyde) in organic chemistry was designed for them. The research goal set for them is to synthesize a peptide aldehyde with the amine and aldehyde group protected (Thr-Phe-Thr-Phe-acetal) in the solution phase (Figure ). The synthesis included suitable protection of the N, and C-terminal, amidation reaction, converstion of weinrebamide to aldehyde and further protection of the aldehyde by converting aldehyde to dimethyl acetal. The main traning goal was to promote the understanding of the basic principles in organic chemistry in addition to the training of organic synthesis skills. |
6775e7f3fa469535b97ec414 | 8 | For project 1, a group of 7 students selected this PBL project. In inquiring why they selected this project, they explained that this project could allow them to use what they have learned from lectures of physical chemistry and organic chemistry. This has partly proven that the project design is appropriate to practice first principles thinking. To strengthen the usage of first principles thinking, cascades of coupled inquiries (inquiry-based technique) were designed to drive the students to reversely breakdown the complex problems to the fundamental principles in the supervision meetings. The cascade of main questions is shown as follows. |
6775e7f3fa469535b97ec414 | 9 | 1) Experimental aspects: what are the building elements for creating chemically fueled assemblies? This question aimed to drive students to look for the basic compositional elements for construting dissipative assemblies. What is the precondition to make the experiment work? This question aimed to drive the students to think of the requirements of the reaction rates in the reaction cycles (i.e. the activation reaction must be faster than deactivation reactions) which logically requires the students to understand reaction kinetics in physical chemistry. Through reverse-engineering the project, the group could break down the big project to fundamental principles in organic chemistry and physical chemistry, these basic pieces was recombined to 160 successfully produce the kinetic models for this project (Figure ). The students firstly determined the rate equations by analyzing and identifying the mechanisms of activation and deactivation reaction in the reaction cycle. Then kinetic models were built for data fitting and MATLAB codes were produced for the data correlation. Based on the first turn of sucessful data fitting, they further found that a simplified model with a steady state approximation could also be used to fit this set of data nicely. Moreover, they For project 2, we aimed to build the experimental project on the understanding of the basic principles in organic chemistry. As this group of students from France did not have any experience in research projects or PBL and had no clue to create a synthetic route for synthesizing the targeting compound, the project was thus broken down and reframed to 3 sub-problems by us to help them work out the synthetic route. Jigsaw technique was employed to enable the generation of the synthetic route by this group. The jigsaw technique is a collaborative learning strategy where students are divided into small groups to work on specific assignments . These groups are then reshuffled into mixed groups. Each person in the groups teaches the rest of the group what he/she knows to complete a comprehensive task, integrating all the pieces to form a complete understanding, like a jigsaw puzzle. Jigswa techniques have been proved effective in chemistry education . Specifically, in this small PBL group, each student was assigned to solve one individual sub-problem through self-study and the basic organic chemistry principles. One student was assigned to work on how to make amide bond and its reaction mechanism, the second strudent worked on the basic protection/deprotection concept and the protection/deprotection methods in peptide synthesis, and the third student focused on the chemistry and mechanism to transform the carboxylate group into aldehyde group (Figure The major aim in the supervision of the two student groups was to drive the student to go deeper the fundamental principles of organic reaction mechanism beneath the empirical project and find new solutions other than to develop experimental techniques in PBL. |
6775e7f3fa469535b97ec414 | 10 | Through the elaborate design of PBL projects, first principles thinking was applied through the implementation of the projects. For the project 1, according to the pedagogy request at AAU, the group of students wrote a report together. In the final 118-page report, the students not only presented their concepts, principles, and knowledge, formulated with their own words which included the basics in reaction mechanisms of nucleophilic substitution, reaction kinetics, the driving forces behind molecular assembly, and the distinctions between equilibrium and non-equilibrium molecular assembly from a thermodynamic perspective. This indicated their active engagement in learning the fundamental 200 prinicples in chemistry. The successful building of the kinetic model also proved a solid understanding of the first principles in chemistry. An oral exam was subsequently conducted for this group, with 3 examiners including the supervisors over a span of 5 hours. The questions ranged from fundamental principles in general chemistry and physical chemistry to reaction mechanisms and the details of research methods and results in their report. All the questions were answered with a certain degree of 205 clarity and relevance although the answers are given by different students. One of the students could even identify what types of non-covalent interactions are involved in the formation of supramolecular structures. The students could also give reasonable explanations and judgments through reasoning from the fundamental principles. Their performances were graded respectively. Denmark's grading system is based on a scale that ranges from -3 to 12 to grade the performance from unacceptable to 210 excellent . Score 2 corresponds to a performance meeting only the minimum requirements for acceptance. In this study group, 3 students are scored 12, 2 students scored 10, and 2 students scored 7. Overall, their collective performance was above the average compared to all student groups in the fall semester. The report and the evalution indicated first principles thinking was effectively enabled in PBL to increase students' learning depth and their ability to solve problems. For the project 2, by reframing the the problem of synthesizing a target compound, the synthetic problem has been broken down to 3 sub-problems. Each student has finally worked out the sub-problems and taught with each other how the synthesis could be performed and what the reaction mechanisms and underpinned chemistry were. |
6775e7f3fa469535b97ec414 | 11 | By combining their learning and solutions, this group has designed multiple plausible solutions (several potential synthetic routes). Finally, they identified and selected the simplest synthetic route, and the target product was successfully synthesized. This specific design of the PBL which combines proble m breakdown and the jigsaw technique enables the production of multiple solutions, primarily indicative of the possibility to the generation of creative solutions. Moreover, this group expressed high satisfaction with their training and performance via an obligatorily survey carried out by their project coordinator at the École Technique Supérieure de Chimie de l'Ouest. As a result, our lab has formalized an agreement with the project coordinator to host intern students regularly each year. |
6775e7f3fa469535b97ec414 | 12 | In summary, two PBL projects were elaborately designed to enable students practicing first principles thinking to promote both learning of fundamental principles in chemistry and using the fundamental principles to generate solutions for research problems. PBL project 1 was focused on constructing kinetic models based on chemically fueled assemblies. To strengthen the practice of first principles thinking for students through working on the project, cascades of coupled inquiries (inquiry-based technique) were designed and used to drive the students to reversely breakdown the problems to the foundations in 235 physical and organic chemistry through regular supervision meetings. The learning of the foundations was strengthened and kinetic models were sucessfully built. PBL project 2 was focused on synthesizing a target compound. In this project, Jigsaw method was used to allow students to learn the fundamental principles in organic chemistry and apply the foundations to product synthetic routes. Through applying first principles thinking, the learning depth has been increased. We have also seen the sign of 'creativity' |
623179fd8ab37333ac6662a7 | 0 | Exciplexes are short-lived, heterodimeric "excited complexes"' that are stable in the electronic excited state while dissociative in the electronic ground state. Their formation can be considered as the association between an excited monomer * M and a second monomer N in the ground state, as depicted by the following rea School of Chemistry, The University of Melbourne, Parkville, Australia. ; Tel: +61-(0)3-8344 6784; E-mail: [email protected]. |
623179fd8ab37333ac6662a7 | 1 | The excited-state properties of the dimeric species are unique from those of either monomer. The stabilising effect in exciplexes and excimers can be rationalised with molecular orbital (MO) theory and a collision model of the reaction in Eq. 1. When the two monomers collide, the predominant interactions occur between the highest occupied (HOMO) and lowest unoccupied (LUMO) molecular orbitals of each species to form a new set of orbitals as shown in the simplified diagram in Fig. . In the case of two ground-state molecules, the resulting constructive and destructive-interference contributions cancel, resulting in minimal to no net stabilisation. In an excimer or exciplex, the electronically excited species causes two interacting orbitals to be singly occupied resulting in stabilisation through formation of the complex. This electronic stabilisation effect is short-lived, as relaxation to the ground state occurs quickly, causing repulsive intermolecular forces to oppose the relatively weak attractive forces. Energy decomposition analysis of the excimer/exciplex stabilisation energy revealed the following contributing components: electrostatics, Pauli repulsion, CT, exciton coupling and London dispersion. Aromatic excimers were discovered experimentally in the fluorescence spectrum of pyrene in cyclohexane solution by a broad, structureless emission band that occurred at a lower energy than the associated monomer emission. Since then, excimers have also been revealed to be crucial species in contemporary applications of technological and biological relevance. Contrary to their notoriety as an undesired energy trap for singlet-fission, some studies have shown potential advantages in organic electronics, such as excimer states mediating intramolecular electron transfer and charge separation or broadening the emission for white organic light emitting diodes; see Ref. 15 for a review. Additionally, excimers have applications as chemosensors, molecular rulers, and industrial-scale lasers. Excimers also occur in biological systems, playing a role in the photo-damage of DNA as they can occur between nucleobases such as adenine and guanine . |
623179fd8ab37333ac6662a7 | 2 | Computational chemistry techniques, particularly Kohn-Sham Density Functional Theory (DFT) approaches, also called density functional approximations (DFAs), are frequently used in applications of technological relevance, including the description of exciplexes and excimers. In order to have predictive character, DFAs must be robust; i.e., they must yield results with equal accuracy or error margins across a variety of different chemical problems. Comprehensive benchmark studies have identified such generally applicable DFT approaches for ground-state problems including for thermochemistry, kinetics, and geometries, to name a few examples. In all studies, it was evident that the accurate treatment of noncovalent interactions (NCIs), in both inter-and intramolecular cases, is crucial to achieve the desired accuracy and robustness of a method. In the context of DFT, it is particularly important to properly address the correct treatment of London dispersion effects. It has been known since the mid-1990s that conventional DFAs do not capture those effects correctly. This sparked the development of various dispersioncorrected DFT techniques, some of which later turned out to be less effective than initially claimed, and others having now become the recommended default for ground-state DFT applications. We refer interested readers to recently published reviews directed at users that are unfamiliar with the field of dispersion-corrected DFT applications for ground-state problems as well as reviews that recommend the currently best practice in the field, which includes dispersion-corrected doublehybrid DFAs (DHDFAs) when feasible. |
623179fd8ab37333ac6662a7 | 3 | The binding of excimers and exciplexes inherently relies on the system's excited-state properties and the dominating types of NCIs, which introduces additional complexity compared to the ground-state case. Linear-response time-dependent DFT within the adiabatic approximation (TD-DFT) has become the method of choice to treat excited-state problems. Recent advances and detailed benchmark studies on singlemolecule cases have shed light on the quality of TD-DFAs for the calculation of excitation energies. As recently summarised for readers unfamiliar with the field, lower rungs on the Jacob's Ladder of DFT should be avoided due to the emergence of artificial ghost states and large red shifts in excitation energies. Global-hybrid DFAs can be suitable for local valence excitations, subject to having the right amount of nonlocal Fock exchange (FE)-about 40% -but fail for CT and other long-range transitions. While the range-separation (RS) technique applied to the exchange part of DFAs solves the long-range problem, local valence excitations tend to be blue-shifted. Time-dependent DHDFAs depend additionally on a nonlocal, perturbative electron-correlation component that compensates for many of the systematic errors of hybrids. In fact, our group recently developed RS-DHDFAs that belong to some of the currently most robust TD-DFT methods for a variety of different local and long-range valence excitations in organic molecules. A slightly dif- ferent group of RS-DHDFAs have also proven to be very promising when used within the Tamm-Dancoff Approximation 80 (TDA-DFT). Compared to single-molecule cases, there are fewer systematic studies that explore NCIs in excited states of molecular complexes. However, it is expected that TD-DFT methods inherit the problems of the underlying ground-state DFA, such as the inability to properly describe London dispersion interactions. Additionally, the TD-DFT formalism itself carries problems that may further complicate the treatment of excited-state NCIs. In 2008, Huenerbein and Grimme presented a study of excimers and one exciplex across a very limited number of TD-DFAs. The study was unique in the sense that a London dispersion correction was rigorously applied to an excited-state study, namely the DFT-D2 correction developed in 2006. Although the dispersion coefficients used in DFT-D2 were based on ground-state polarisabilities, the combination of DFT-D2 with high-FE global-hybrid DFAs, showed a qualitatively better agreement with experimental data compared to dispersion-uncorrected TD-DFT, both for the stability of the dimers and the inter-monomer equilibrium distance. The authors' argument was that a ground-state based correction should be seen as a lower bound for the dispersion energy in an excited-state system, as C 6 coefficients are different, and most likely larger, for more polarisable excited states. Some of the subsequent TD-DFT based studies on excimer and exciplex systems either do not explore the dispersion problem of DFT or erroneously justified the use of Minnesota functionals 87 for consideration of dispersion effects, despite ground-state based studies that disprove claims they are able to treat such interactions. Some other studies recognised the dispersion problem, but for a lack of a better choice, followed Huenerbein and Grimme's recommendation to use a global-hybrid combined with DFT-D2, , while fewer applied the newer, groundstate based DFT-D3 method in its zero-damping [DFT-D3(0)] or in its Becke-Johnson-damping [DFT-D3(BJ) ] form. To our knowledge, only three studies have considered adjusting dispersion corrections for excited states. In the first, Ikabata and Nakai 107 calculated excited-state dispersion coefficients within the local-response dispersion 108,109 (LRD) model to explore the interaction energies of exciton-localised molecular complexes from the S66 110 benchmark set as well as three molecular excimers. Then, Briggs and Besley 111 empirically varied the dispersion coefficients and van der Waals radii in DFT-D2 for small model systems of ethene-argon and formaldehydemethane complexes. And finally, Johnson and co-workers calculated dispersion coefficients within the exchange-hole dipole mo-ment (XDM) model for excited states of conjugated hydrocarbons, pull-pull chromophores, and CT complexes. While these studies considered suitable model systems, only a limited selection of DFAs were applied. These three studies, among others, would allow the conclusion that state-specific dispersion corrections for excited states should be developed, but further characterisation of current methods is necessary to provide guidance towards such future developments. The present paper intends to provide such characterisation. |
623179fd8ab37333ac6662a7 | 4 | Most of the aforementioned exciplex studies used older globalhybrid DFAs, while some used rangeseparated hybrids. Of the two studies that investigated DHD-FAs both utilised TDA-DFT, instead of the full TD-DFT scheme, but only Krueger and Blanquart 102 considered a dispersion correction. Interestingly, most of the methods previously used to study exciplex systems are not those recommended by some of the latest single-molecule excitation studies, which by itself is reason enough to investigate the latter in the exciplex/excimer context. The majority of excimer and exciplex studies investigated these methods via applications, or carried out benchmark analyses against reference values that were not well defined and mostly based on experiment. Both benchmarks and application-based excimer studies focussed on comparison of methods by the calculation of the dissociation energy, as this explores the binding of the complex, while fewer studies have explored other spectrometric parameters such as the fluorescence, absorption and repulsion energies. Potential energy curves of the lowest excited state reported relative to the asymptote of the dissociation curve of the ground state 5 are standard practice to derive these spectroscopic parameters. Groundstate studies commonly explore NCIs by focusing entirely on interaction energies that can be calculated, e.g., as the difference between the dimer and individual monomer energies or alternatively as the difference between the total energies of the dimer and a system in which the two monomers have been dissociated to a large distance from one another. By this definition, a negative interaction energy indicates a stable complex. The dissociation energy of an excimer is inherently an interaction energy and provides a convenient metric to discuss excited-state binding and NCIs in one go. However, the computational studies of excimers and exciplexes do not report excited-state interaction energy curves, other than a handful of exceptions we are aware of; we advocate to adopt it more frequently, as it allows direct comparison with what has become the standard in groundstate treatments of NCIs. Note that according to that definition of an interaction energy, the dissociation energy D e is simply its absolute value. |
623179fd8ab37333ac6662a7 | 5 | There have been many rigorously conducted NCI studies for ground-state systems that are usually based on a high-quality reference and compared different computational approaches on an equal footing; the benefits of using high-level computational data as opposed to experimental references has been well-established and explained elsewhere. To the best of our knowledge, similar studies are missing to systematically study NCIs in excited states. In this work, we intend to initiate the first step towards more systematic studies of excited-state NCIs by restudying excimer model systems under incorporation of the latest state-of-the art TD-DFT methods. The four model systems discussed herein are the benzene, naphthalene, anthracene and pyrene dimers (see. Fig. ) which have been studied in combination before by two prior studies. We discuss the change of the interaction energy for each dimer system upon dissociation, i.e. we discuss equilibrium and non-equilibrium geometries. In each case, we discuss the first excited state, which is the excimer state. In the following section, Section 2, we briefly outline general computational details before analysing various wavefunction theory (WFT) levels based on truncated basis sets and at the complete-basis-set (CBS) limit in Section 3. The purpose of this analysis is to identify a suitable level of theory that can serve as a reliable, yet fast benchmark for the subsequent discussion of TD-DFT methods in various subsections of Section 4. The latter is split into different aspects: the impact of non-local FE, the behaviour of dispersion-uncorrected TD-DFT methods, the impact of using ground-state based corrections of the DFT-D3(BJ) and DFT-D4 122,123 type, and an analysis of method performance averages over the four systems. In a nutshell, our study is sufficiently comprehensive to fill an important gap in our knowledge of current TD-DFT methods and additive dispersion corrections for the description of excimers. |
623179fd8ab37333ac6662a7 | 6 | 2 General computational details TURBOMOLE 7.3 was used for geometry optimisations of the lowest-lying singlet excited states of each dimer at the spin-component-scaled 127,128 approximate coupled cluster singles doubles (SCS-CC2 129 ) level with the def2-TZVP 130 tripleζ (TZ) atomic-orbital (AO) basis set and a geometry convergence criterion of 10 -7 E h . Use of the RICC2 module in TURBO-MOLE limits the symmetry consideration to Abelian point groups such that all excimers are calculated with D 2h symmetry. The excited-state optimised structures were used to generate nonrelaxed dissociation curves as follows: the internal coordinates of each monomer were frozen, then total energies of the lowestlying singlet excited states were calculated across a range of intermonomer separations up to 16 Å, and then each energy was taken relative to 16 Å, which represents the asymptote for the given state. The geometry optimised excimer structures were not comprised of perfectly planar monomers, which has been previously noted during both TD-DFT and CC2 131 optimisations of these excimers and exciplexes comprised of the same monomers. Therefore, we defined the inter-monomer separation for our dis-sociation curves as the distance between two central carbon atoms, opposite each other on each monomer; this parameter is visually defined for each excimer structure in Fig. ). D e , is defined as the difference between the total energy of the first excited state at the 16 Å point and the total energy of the same state at the minimum point. The effect of geometric relaxation falls outside the scope of this study. However, frozen-monomer structures at the SOS-CIS(D 0 ) 132 level have been shown to give good estimates of the equilibrium inter-monomer distance, with excimer dissociation energies that are generally overestimated by ca. 2 kcal/mol in the case of the benzene excimer. As this study uses the same structures throughout, internal comparison of the dissociation energies is reasonably justified. |
623179fd8ab37333ac6662a7 | 7 | Herein, the figures display interaction energy curves such that a negative interaction energy indicates a stable dimer. In the context of this paper, D e corresponds to the dissociation energy of an excimer and therefore the terms interaction and dissociation energy are used interchangeably. Equilibrium inter-monomer distances r e , corresponding to the distance at which the minimum energy arises, are also discussed in parts of our discussion. |
623179fd8ab37333ac6662a7 | 8 | TURBOMOLE was also used for all SCS-CC2 and CC2 singlepoint calculations along the dissociation curves. The frozen-core and resolution of the identity 134 (RI) approximations were employed with appropriate auxiliary basis sets 135 . Coupled cluster singles doubles with perturbative triples excitation correction [CCSDR(3)] calculations for select points along the dissociation curves were carried out with Dalton2016, 137 also utilising the frozen-core approximation. The WFT calculations were used to identify suitable reference data for the subsequent assessment of TD-DFT methods. |
623179fd8ab37333ac6662a7 | 9 | Single-point, linear-response TD-DFT calculations within the adiabatic approximation were conducted for all dissociation curves for a series of DFAs as listed in Table . For the study of the influence of FE we apply a series of PBE 138 -and BLYP 139-141based methods with varying amounts of non-local exchange. However, as lower-rung DFAs have shown to be unreliable for excited-state problems, we limit our final benchmarking study to global hybrids, range-separated hybrids, global double hybrids and range-separated double hybrids; the exact functional types are detailed in Table . Those DFAs have been chosen either based on popularity or known accuracy for excited states of organic molecules; see Ref. 66 for recommendations and insights. Note that time-dependent double-hybrid calculations require a conventional TD-DFT step for its hybrid portion followed by a configuration interaction singles with perturbative doubles [CIS(D)] correction as initially suggested by Grimme and Neese; 77 see Ref. 66 for a comprehensive, free-access review on this methodology and some of its latest advances. |
623179fd8ab37333ac6662a7 | 10 | Self-consistent-field (SCF) steps in this work were carried out with energy convergence criteria of 10 -7 E h (TURBOMOLE) and 10 -8 E h (ORCA). The frozen-core approximation was applied across all TURBOMOLE calculations, and in ORCA 147,148 calculations it was used along with the RI approximation for the perturbative parts of the double hybrids. Large quadrature-grid options "7" (TURBOMOLE) and "grid6 finalgrid7" (ORCA) were used to ensure smooth dissociation curves. Dispersion corrections of the type DFT-D3(BJ) DFT-D4 122,123 were carried out with the DFTD3 and DFTD4 standalone programs. Damping parameters for the various functionals were taken from the respective reference that published them first. A series of different basis sets-ranging from double-ζ (DZ) to quadruple-ζ (QZ)-were employed throughout this work, with some of them used for extrapolations to the CBS limit. More information is provided in the following sections. |
623179fd8ab37333ac6662a7 | 11 | Prior to analysing the performance of TD-DFT methods, a sufficiently reliable reference method that is still feasible for larger systems must be identified. An appropriate reference method should minimise the computational cost while maintaining reasonable comparability to a higher level of theory. Coupled-cluster WFT methods are considered the gold standard of chemical accuracy. For ground states, coupled-cluster singles-anddoubles with perturbative triples, CCSD(T), 156 at the CBS limit is the ideal that many aim to achieve in contemporary benchmarking. It constitutes a very accurate approximation to the true interaction energy of noncovalently bound complexes in their electronic ground state. One excited-state equivalent to ground-state treatments with CCSD(T) is linear-response CCSD enhanced with a different type of perturbative triples correction, such as the aforementioned CCSDR(3). It delivers excitation energies that are similar to the more costly linear-response approximate coupled cluster singles doubles triples (CC3); for examples of CCSDR(3) and CC3 being established as benchmarks and comparisons between both, see Refs 51,52, and 162-165. While the ground-state gold standard has, in recent years, become increasingly feasible for large systems of up to several hundred atoms, 166,167 the excited-state equivalent remains prohibitively expensive. For most of our systems, CCSDR(3) is therefore not achievable for computational reasons, and in the following we identify if instead CC2 or its spin-component-scaled 128 vari-ant, SCS-CC2, can be used as a low-cost alternative. |
623179fd8ab37333ac6662a7 | 12 | Establishing a reference method also requires the careful choice of an appropriate basis set. Large basis sets quickly become computationally prohibitive, especially for large systems, while small basis sets are rife with errors due to being incomplete. Studies on obtaining vertical excitation energies, including studies on DHD-FAs, have shown that large TZ basis sets are often sufficient to obtain nearly converged results, with little change when using QZ basis sets. That being said, as interaction energies in the excited state have not been studied frequently, little is known about the effects of truncated basis sets on interaction energies in excited states. |
623179fd8ab37333ac6662a7 | 13 | The basis set superposition error (BSSE), for example, plagues the treatment of ground-state NCIs with small AO basis sets in both WFT and DFT methods. This well-known error is caused by the limited number of AOs available on a molecular fragment 'borrowing' the basis functions from other fragments, which artificially stabilises the multi-fragment system-such as a dimerrelative to its separate fragments. Small basis sets also suffer from basis set incompleteness error (BSIE) which can artificially destabilise complexes through a failure to correctly describe intermolecular electron density. In the case of interaction energies, BSIE is not consistent with varied inter-monomer separation and will therefore not be cancelled by similar error of infinitely separated monomers in the calculation of interaction energies. If use of a larger basis set is not suitable, additive corrections have been developed for ground states to minimise the BSSE. 170-172 BSIE corrections 173,174 have also been developed although they are generally scare and more computationally demanding. To our knowledge BSSE corrections have not yet been developed for excited states, and basis set convergence studies for such systems are sparse. 121,169 Some previous exciplex studies opted to utilise the counterpoise correction to account for BSSE in exciplex binding. Our own basis set study below utilises BSSE-uncorrected interaction energies in interest of time and avoiding corrections that have not been thoroughly assessed specifically for excited states. In that way we are also able to provide a picture of current methodology that can be easily been applied by method users. Moreover, we strive to use large basis sets for CBS extrapolations, which further reduces the impact of BSSE. That being said, it is worthwhile to explore the influence of counterpoise or similar corrections in later studies. A BSIE correction for excited states was recently developed Loos and co-workers that was shown to recover chemically accurate vertical excitation energies of small organic molecules with augmented double-ζ basis sets, with the exception of diffuse excited states. This suggests that work to establish basis set error corrections for excited states is underway, however, current methods do not seem to be robust enough yet to provide reliable corrections for our present study. |
623179fd8ab37333ac6662a7 | 14 | For this section, non-relaxed dissociation curves for the first excited state of the benzene dimer were generated with CC2 and SCS-CC2 and various AO basis sets. While CC2 and SCS-CC2 both yield reliable excitation energies, albeit still above the related chemical-accuracy threshold of 0.1 eV, benchmark studies suggest that the latter does not show consistent improvement to vertical excitation energies in general. However, enhanced excited-state geometries and vibrational frequencies could lead to more accurate 0-0 transition energies for π → π * and n → π * excitations. Therefore, any potential improvements for SCS-CC2 seem to be problem specific, hence we need to carry out a comparison with CC2. |
623179fd8ab37333ac6662a7 | 15 | To decide between CC2 and SCS-CC2, we conduct a study DZ-, TZ-, and QZ-quality basis sets both with and without additional diffuse functions. This comparison involves both Ahlrichs [def2-nZVP(D)] 130,182 and Dunning [(aug-)cc-pVnZ] 183,184 basis sets; where n corresponds to 'S/D', 'T', 'Q', respectively. All dissociation curves are shown in Fig. including select points for CCSDR(3)/def2-TZVP near the minimum to allow for an initial evaluation. Note that due to its computational cost, generating complete dissociation curves with CCSDR(3) was technically not feasible. Numerical values for interaction energies and intermolecular distances at the respective curve minima are shown in Table . It is obvious that all CC2 minima-regardless of the basis setare deeper than the SCS-CC2 ones, meaning that CC2 consistently gives larger dissociation energies (Fig. ). CC2 results indicate systematic overstablisation, exhibiting dissociation energies that are almost twice as large as CCSDR(3). For instance, with the def2-TZVP basis set, the CCSDR(3) interaction energy at an intermolecular distance r of 3.00 Å (a point in proximity of the expected minimum for this level of theory) is -11.96 kcal/mol (Table † ) and the well depths of SCS-CC2 and CC2 are -13.18 kcal/mol (r e =2.99 Å) and -20.19 kcal/mol (r e =2.90 Å), respectively. While the differences between CC2 and SCS-CC2 are quite striking, the general trends show parallels to ground-state studies, where it has been established that conventional MP2 overestimates interaction energies in dispersion-driven complexes, while SCS-MP2 provides a more balanced description. ,128 CC2 has also been shown to overestimate the dispersion contribution in the same excimer complexes as studied here and it appears, by its closer proximity to CCSDR(3), that SCS-CC2 reduces this overestimation. Despite SCS-CC2 being in closer proximity to CCSDR(3), both CC2 variants across all basis sets explored overstabilise the excimer relative to CCSDR(3)/def2-TZVP. |
623179fd8ab37333ac6662a7 | 16 | Basis sets including diffuse functions-aug-cc-pVTZ and def2-TZVPD-produce minima that are considerably lower than basis sets without diffuse functions, meaning that the absolute interaction energies are larger for both CC2 variants in those cases. As CC2 more greatly overestimates the interaction energy minima, additional diffuse functions give well depths furthest from the CCSDR(3) reference. Interestingly, def2-TZVPD produces a deeper minimum than aug-cc-pVTZ. For SCS-CC2, def2-TZVPD and aug-cc-pVTZ differ by 0.58 kcal/mol, while for CC2 they differ by 0.38 kcal/mol. Without diffuse functions, Dunning and Ahlrichs basis sets are closer in energy; for instance, def2-TZVP and cc-pVTZ minima differ by 0.19 kcal/mol and 0.24 kcal/mol for SCS-CC2 and CC2 respectively (Table † ). The choice to include diffuse functions is highly system and method dependent as in some cases their addition can increase the BSSE of the system. In the case of exciplex binding, diffuse functions giving a worse result may suggest that the BSSE and BSIE do not decrease at the same rate. The impact of diffuse functions is of interest for future studies. Herein, we choose to discard them for prag-matic reasons, as the CCSDR(3) data could not be obtained with diffuse functions. However, this decision will not influence our subsequent TD-DFT benchmark study, as long as the same type of basis set is used therein. Our final findings and conclusions are therefore unlikely to be affected by this decision. |
623179fd8ab37333ac6662a7 | 17 | Based on the herein discussed findings, we rule out using the CC2 method in the remainder of the study due to its greater tendency to overestimate the well depths relative to CCSDR(3). Interaction-energy curves for Ahlrichs and Dunning basis sets without diffuse functions, show reasonable agreement for SCS-CC2 with the def2-TZVP well being only by 0.19 kcal/mol deeper than for cc-pVTZ. Given that Ahlrichs basis sets are computationally more efficient due to relying on fewer primitive Gaussian-type orbitals, they are our preference in this study, particularly when considering the larger dimers. In order to further clarify the best choice of basis set for SCS-CC2 as a wave function reference, CBS values are generated for CCSDR(3) and SCS-CC2 in the following section. |
623179fd8ab37333ac6662a7 | 18 | The total energy calculated by a given method is expected to converge to a finite value with an increase in AO basis set size. This also leads to converging interaction energies, as indicated by the series of interaction energies ranging from DZ to QZ basis sets discussed in the previous section, with changes between TZ and QZ being smaller than between DZ and TZ. CBS extrapolations take advantage of this convergence behaviour to estimate the fully converged energy for a given family of basis sets. While the practice of CBS extrapolation is well established for ground-state studies, CBS extrapolations are scarcely conducted for excitedstate problems due to a lack of established extrapolation methods. Prior to the development of established excited-state extrapolation methods, application of ground-state extrapolations should at least improve the ground state at the base of excited-state energies. This improvement of the ground-state energy is expected to extrapolate the convergence behaviour enough to give us further insight into the best comparison of SCS-CC2 to CCSDR(3). We therefore chose to employ the same extrapolation formulae and the same extrapolation exponents as in ground-state studies. Further studies to comprehensively investigate basis set convergence behaviour of NCIs in excited states, such as that conducted by Krueger and Blanquart for exciplex systems, 169 are strongly encouraged but such characterisation falls outside the scope of this study. |
623179fd8ab37333ac6662a7 | 19 | In ground-state extrapolations the HF and correlation energies are extrapolated separately due to their different convergence behaviours. The equivalent to ground-state HF energy in excited-state coupled-cluster treatments is the CCS total energy as it does not include electron correlation. We conducted linearresponse calculations, so CCS total energies (E CCS ) were obtained by adding the CCS excitation energy to the HF ground-state total energy. Such excited-state total energies for two truncated basis sets were then used to obtain the resulting CBS-limit energy with the familiar formula: 186 |
623179fd8ab37333ac6662a7 | 20 | where X and Y are the successive cardinal numbers of the two basis sets used, and α is a constant specific to the family of basis set used. We adopted the value of α for ground-state Ahlrichsbasis set extrapolations, namely α = 10.39 for DZ-TZ and α = 7.88 for TZ-QZ extrapolations. The electron-correlation contribution E C was obtained from the differences between CCS and SCS-CC2 or CCSDR(3) excitation energies. Those contributions were extrapolated to the CBS limit with the familiar formula for the correlation energy (E C ): 188 |
623179fd8ab37333ac6662a7 | 21 | where the basis-set specific constant β has values of 2.40 (DZ-TZ) or 2.97 (TZ-QZ), respectively. The resulting energies E CCS (CBS) and E C (CBS) were added together to obtain the excited-state total energies. These were then used to calculate the interaction energies at the CBS level. Ideally, the basis sets used in such extrapolations should be as large as technically possible. For CCSDR(3), only DZ and TZ calculations were feasible for the benzene dimer, while TZ and QZ calculations were possible for SCS-CC2. The resulting CCSDR(3)/CBS(2,3) interaction energies for the same four select intermolecular distances discussed for CCSDR(3)/def2-TZVP in the previous section are shown in Fig. (numerical values in Table † ) alongside the SCS-CC2 curves with truncated Ahlrichs basis sets and CBS (3,4). SCS-CC2/CBS interaction energies are slightly more negative than CCSDR(3)/CBS for three of the four points with the differences ranging from 0.34 kcal/mol (r=2.90 Å) to 0.1 kcal/mol (r=3.00 Å). Both levels of theories agree for r=3.05 Å with ∆E= -13.42 kcal/mol. The agreement between both levels is therefore better near the minimum region of CCSDR(3)/CBS, which lies close to 3 Å (see Fig. ). |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.