id
stringlengths 24
24
| idx
int64 0
402
| paragraph
stringlengths 106
17.2k
|
|---|---|---|
6352afdeecdad55d48e71510 | 33 | The determination of the right gene sequence in nucleic acid-based biosensor development is vital for distinguishing the pathogenic strains from the non-pathogenic ones. For the detection of general E. coli, 16s rRNA or uidA gene are the most common biorecognition elements among the selected articles. These genes, however, do not allow distinguishing between E. coli and Shigella spp., which share very similar genetic material to E. coli. Other genes such as yaiO, MalB or ybbW that show higher specificity towards E. coli can be considered in the future development of nucleicacid-based biosensors targeting these bacteria. The most specific detection of STEC should be based on either stx1 or stx2 genes coding for toxin production which has been studies by several researchers for the development of STEC biosensors. Usually, the eaeA gene, coding for intimin production, has been targeted, which is not sufficient to recognise all strains responsible of serious illnesses. Currently, there are several studies actively targeting the identification of more virulence factors of STEC and the development of multiplex biosensors capable of detecting several genes simultaneously. |
6352afdeecdad55d48e71510 | 34 | Various combinations of the materials and methods were used to develop nucleic acid-based sensors for the detection of general E. coli and STEC. This includes the electrode material, the surface chemistry application of the electrode, probe immobilisation and the chosen electrochemical method for detection of the hybridisation event. The biosensors collected in this review were divided into four categories depending on the electrode's materials (gold, ITO, or carbon) used and magnetic particles. Clearly, the electrode material, surface modification and detection technique have been key factors for the development of sensitive and selective biosensors. Carbon and gold were the most often applied electrode materials among selected literature and typically their modification with a mix of nanomaterials and polymers demonstrated even greater enhancement in sensitivity that resulted in achieving limits of detection in fM and aM range. Many researchers have shown the proof-of-concept with synthetic DNA strands while the real-life applications have been validated with a bacterial culture, naturally contaminated or spiked food samples. In the future, more focus should be placed on sensor's calibration using bacterial culture allowing for establishing an LOD in CFU/mL. This would give a better overview of sensor's performance. In addition, using high volumes of complex samples for the validation would show how such system may perform in the agri-food industry. |
6352afdeecdad55d48e71510 | 35 | The number of research articles published on electrochemical nucleic acid-based biosensors has been increasing every year, however there is still no commercial DNA biosensor available. The major roadblock of the success of a commercial DNA biosensor can be the limitations arising from sample preparation, possible contaminations, or a need for an amplification step that may require laboratory settings. Also, the studies concentrated on the fully integrated DNA biosensing devices allowing for DNA extraction, PCR, or isothermal amplification (if needed) and electrochemical detection require more research and development. Notably, the devices with multiplexing capabilities for the detection of several genes simultaneously may lead a successful commercialisation path for future diagnostics. Finally, CRISPR-Cas, a technology which is gaining much attention in molecular biology in recent years, is a highly promising tool for the development of highly efficient sensing platforms. More development is, however, needed to meet all the ASSURED criteria (Affordable, Sensitive, Specific, User-friendly, Rapid, and robust, Equipment-free and Deliverable to end-users) required to successfully develop a device for POC application. |
6531003287198ede079c56e8 | 0 | Conventionally, starting materials traverse through a photochemical and an electrochemical reactor connected in series, likely attributed to a blend of technological and chemical challenges in balancing the rates of the two concurrent transformations . However, from a chemical perspective, the transient nature of species formed in electrophotocatalysis necessitates the simultaneous execution of both the photochemical and electrochemical steps for effective outcomes. Technologically, the paragon reactor must adeptly navigate practical challenges inherent to both photo-and electrochemistry, each presenting its own scalability hurdles. Thus, in a unified reactor, electrodes must be installed while concurrently ensuring adequate illumination of the reaction medium. This implies that technological barriers, such as the Ohmic drop in electrochemistry and the attenuation effect in photochemistry, must be surmounted simultaneously to guarantee process scalability and reproducibility. |
6531003287198ede079c56e8 | 1 | In light of the aforementioned challenges, achieving the construction of an effective f-EPC reactor remains a pivotal yet unfulfilled objective. Recently, the Atobe group made strides in this domain, introducing a prototype of a flow cell for synthetic electrophotocatalysis. While it introduced valuable insights and advancements, there were aspects regarding robustness and productivity that requires further refinement and exploration (13). The desired reactor would ideally be i) modular and robust, ii) comprised of interchangeable, solvent-resistant materials, iii) safe and user-friendly, and iv) versatile, facilitating either independent or simultaneous execution of photochemistry and electrochemistry. Based on our experience, we posit that the challenges intrinsic to photochemistry and electrochemistry can be concurrently addressed through reaction miniaturization. Given the short-lived nature of the generated reactive species, it is plausible to suggest that all pertinent processes will occur in close proximity to the electrodes. Consequently, efficient photon delivery near the electrode surface becomes crucial. |
6531003287198ede079c56e8 | 2 | Thus, an optimal f-EPC reactor should feature at least one transparent electrode to ensure effective light penetration. Lastly, the reactor's volume should be adaptable to modulate productivity, such as by altering the inter-electrode gap using gaskets. Capitalizing on our experience (8c, 14), we present herein the development of such reactor to perform an efficient electrophotocatalytic heteroarylation of C(sp 3 )-H bonds. The achievement of this transformation was realized through the combination of hydrogen atom transfer (HAT) via LMCT photocatalysis and an electrochemical oxidation. |
6531003287198ede079c56e8 | 3 | In adherence to the previously outlined criteria, we constructed the f-EPC reactor, integrating stainless steel and PTFE (polytetrafluoroethylene) frames (160 mm × 95 mm × 10 mm) to house the electrodes within an insulating and solvent-resistant environment. This design ensures that the electrodes can be firmly compressed, and the inter-electrode distance can be finely tuned by utilizing serpentine-shaped PTFE gaskets of varying thickness (0.1 mm -0.5 mm), thereby allowing a cell volume of 0.28-1.4 mL. Liquids can be introduced directly into the reactor using Super Flangeless Nuts (PEEK, 1/4-28 Flat bottom, for 1/16"OD). Evidently, at least one of the electrodes must be transparent to ensure effective light penetration into the reactor. Thus, we implemented an FTO (fluorine-doped tin oxide) glass slab coated with a thin layer of Pt nanoparticles via a previously reported procedure to boost its catalytic activity to proton reduction (14b, 14c). |
6531003287198ede079c56e8 | 4 | With this apparatus in hand, we embarked on its validation, conducting a series of transformations, encompassing electrochemical, photochemical, and electrophotocatalytic processes (Scheme 2A). In verifying the reactor for electrochemistry, we opted for the synthesis of vinyl sulfones from cinnamic acids and sodium sulfinates, as detailed by Zha, Wang, and coworkers . Notably, pushing a mixture of cinnamic acid and sodium phenylsulfinate through the reactor yielded the desired product at 90%, with a residence time of 4.4 minutes and a total charge of 3 F. This outcome mirrors the results presented in the original paper, underscoring the feasibility of conducting electrochemical-only reactions with our new reactor concept. In a parallel vein, the f-EPC reactor was explored under photochemical-only conditions by executing a photocatalytic Giese-type radical hydroalkylation (18). When a mixture of dimethyl maleate and cyclohexane was flowed through the reactor, a good performance was observed, achieving a 60% 1 H-NMR yield of the anticipated product within a 10-minute residence time (further information are reported in the Supporting Information in Section S2) . While the preliminary results positively showcased the ability of the reactor to handle electrochemical and photochemical processes, achieving a successful outcome in electrophotocatalytic transformations presented a unique challenge, necessitating meticulous tuning of the kinetics within electron-and photon-driven elementary steps to increase efficiency. Inspired by our previous work (14d), initial investigations commenced by studying Table . Optimization of reaction conditions and control experiments for the electrophotocatalytic C(sp 3 )-H heteroarylation in flow. [a] Entry Variation from conditions Yield [b] 1 Yields determined by 1 H NMR spectroscopy using CH2Br2 as external standard. Yield of the isolated product is given in parenthesis. |
6531003287198ede079c56e8 | 5 | Venturing further, we subjected a diverse set of H-donor substrates to our flow-based electrophotocatalytic reaction conditions. Efficient α-to-O C(sp 3 )-H functionalization across various cyclic and linear ethers delivered the sought-after compounds in commendable yields . Notably, selective α-to-S and α-to-N C-H bond functionalization was observed as well , maintaining excellent selectivity over competing α-to-O C-H activation, as seen, for instance, with compound 18. Our method also demonstrated capability in functionalizing benzylic positions; xanthene was functionalized yielding notably high yields (21, 95%). Remarkably, functionalization of the nonsteroidal anti-inflammatory drug pranoprofen was accomplished, procuring product 22 (67% yield). Finally, our benchmark reaction involving ethyl pyrazole-4-carboxylate (1a) and tetrahydrofuran (2a) was readily scaled up in flow using the standard procedure, with an enhanced concentration (0.3 M) and a larger H-donor excess (10 equiv) for extended operation times (5 mmol, 70% isolated yield, productivity: 1 mmolꞏh -1 ). The reactor maintained adequate performance even after 210 minutes (corresponding to 14 residence times). |
6531003287198ede079c56e8 | 6 | We subsequently carried out a series of experiments to unravel the intricacies of the reaction mechanism, presenting our findings in Scheme 3. Our initial step involved conducting the reaction under optimized conditions while introducing TEMPO (4 equiv), a well-known radical scavenger. Notably, the reactivity was entirely quenched, and adduct 23 was identified via HRMS, confirming the presence of radical species. Parallelly, a reaction utilizing an electrondeficient olefin, namely dimethyl maleate, afforded product 24 (25% NMR yield) alongside product 3, further underlining the likely intermediacy of an α-oxyalkyl radical through a chlorine-mediated HAT step (Scheme 3A). Next, Kinetic Isotope Effect (KIE) evaluations, through both competitive and parallel approaches , consistently presented a value approximating 1. This suggests that the Hydrogen Atom Transfer (HAT) is not the ratedetermining step (Scheme 3B). A proposed reaction mechanism, illustrated in Scheme 3C, has been derived from these mechanistic observations. Initially, a chloride ligand associates with the iron center, and upon absorbing purple light, excites the complex to a higher energy state. This is followed by the homolysis of the Fe-Cl bond, generating a chlorine radical and a reduced Fe II species (21). The reduced catalyst is then oxidized at the anode, resulting in the formation of the Fe III center, which closes the catalytic cycle. Iron plays a pivotal role in this transformation. In the absence of iron, we still observe product formation, however, it is accompanied by a substantial production of molecular chlorine, resulting in the unwanted chlorination of pyrazole substrates. |
6531003287198ede079c56e8 | 7 | Importantly, the presence of iron effectively curtailed this unintended pathway. We hypothesize that this inhibition occurs due to the lower redox potential of Fe(II) compared to chloride (EFeIII/FeII = +0.53 V vs SCE, ECl•/Cl-= +1.12 V vs SCE), essentially making the iron catalyst function as a safeguard against both overoxidation and chlorination of the pyrazole substrates . Meanwhile, the chlorine radical executes C-H bond cleavage adjacent to a heteroatom (i.e., O, N, or S), thus generating a nucleophilic carbon-centered α-oxyalkyl radical (2a˙) . At this point, two scenarios can be envisaged. On the one hand, 2a˙ can be electrochemically oxidized to form a stabilized, electrophilic oxocarbenium ion (2a + ). The latter intermediate is subsequently trapped by the nucleophile, forming the desired C-N bond. |
6531003287198ede079c56e8 | 8 | On the other hand, the α-oxyalkyl radical (2a˙) can be trapped by minor amounts of solubilized Cl2-formed via the anodic oxidation of chloride-resulting in a chloride intermediate (2a-Cl) . Although the intermediate 2a-Cl was not observed in our experiments, it would theoretically lead to the formation of product 3 via nucleophilic substitution, aided by the anomeric effect . Despite extensive efforts, the C(sp 3 )-H heteroarylation method could not be extended to unactivated aliphatic C(sp 3 )-H bonds (see Supporting Information). The fact that these substrates are unreactive suggests that the oxidation from intermediate 2a˙ to 2a + is only realized when the latter is adequately stabilized, as observed with an oxocarbenium ion. |
60c74faf0f50db01a1397445 | 0 | Reducing the ecological and economic footprint of electrochemical energy storage requires battery and storage concepts beyond standard intercalation-type Li-ion batteries. Current technologies suffer from the need of expensive, toxic, and flammable materials that are often obtained under harsh environmental and socioeconomic conditions . Amongst the possible alternatives, iodine-based aqueous systems, such as iodide hybrid supercapacitors , zinc iodine batteries , or zinc iodide flow batteries are highly promising considering their performance, sustainability, and environmental aspects. However, to see their more widespread use in mobile or stationary applications, energy density, rate capability, and long-term stability need to become competitive with current Li-ion battery technology. |
60c74faf0f50db01a1397445 | 1 | At a planar platinum surface, triiodide (I3 -) forms by electrochemically oxidizing iodide (I -) to iodine (I2), followed by comproportionation of I -and I2 . In aqueous hybrid supercapacitors and iodine batteries, the reaction occurs in the nanopores of positively polarized carbon electrodes . Confinement in such electrodes may change rates of individual reaction steps and hence the stability of I2 and I3 -. While the polyiodides I3 - and I5 -are generally accepted to form during I -oxidation, described mechanisms are ambiguous . The rather slow self-discharge of iodine-based electrochemical energy storage devices is currently attributed to immobile I3 -and I5 -confined in the narrow carbon pores of below 1 nm diameter . However, based on the known reaction mechanism on planar platinum electrodes and the use of battery electrodes with physically impregnated I2 , alternative electrodeposition of solid I2 in carbon nanopores ought to be considered. |
60c74faf0f50db01a1397445 | 2 | Resolving the mechanism requires in situ techniques with chemical and structural sensitivity for all involved species in the nanoporous carbon. In situ Raman spectroscopy probes species in electrochemical cells and is sensitive to polyiodides (I3 -, I5 -) . In situ small and wide-angle X-ray scattering (SAXS/WAXS) is sensitive to concentration changes and structural arrangements of molecules and ions in the nanoporous system . Given the high scattering power of all iodine species (I -, I3 -, I5 -, I2), in situ SAXS/WAXS data should provide rich kinetic and structural information. However, the complexity induced by the multiphase character of these systems makes the SAXS data analysis highly challenging . |
60c74faf0f50db01a1397445 | 3 | Here, we show that I -oxidation in microporous carbon (pore size < 2 nm) produces solid I2, which can reach pore fillings of at least 30 % and which in the iodide electrolyte partly dissolves into I3 -and I5 -. The latter are responsible for self-discharge via shuttling. In situ Raman and ex situ ultraviolet-visible (UV-vis) spectroscopy confirm I2, I3 -and I5 -during positive polarization and their reduction into I -during negative polarization. In situ SAXS/WAXS data show that a majority of these species is confined to the carbon nanopores. Combined with stochastic modeling, in situ SAXS quantifies the amount of solid iodine deposit and visualizes its structural evolution in the pores. Based on the derived reaction mechanism, we show that high capacity with low self-discharge requires a small concentration of mobile polyiodides and a large fraction of immobile iodine deposits. |
60c74faf0f50db01a1397445 | 4 | In situ Raman and SAXS/WAXS were done in two-electrode hybrid supercapacitor cells comprising activated carbon (AC) electrodes and aqueous 1 M NaI electrolyte. The iodide/iodine redox reaction takes place at the positive electrode, while the negative electrode stores charge exclusively via electrical double-layer capacitance. Faradaic capacity (current) at the positive electrode would be limited by the capacitance (current) of an equally sized negative electrode . We increased the amount of iodine within the positive electrode by oversizing the negative AC electrode about 10-times in mass to exploit the Faradaic capacity and enhance the in situ Raman and SAXS/WAXS signals. The microporous AC has a mean pore size of 0.81 nm, a specific surface area of 1947 m 2 g -1 , and a specific pore volume of 0.84 cm 3 g -1 (Supplementary Table and Supplementary Fig. ). The experimental set-ups and cell assemblies are discussed in the Methods and sketched in Supplementary Fig. . |
60c74faf0f50db01a1397445 | 5 | In situ Raman data were measured during cell voltage sweep between -0.25 V and +0.55 V at a scan rate of 0.32 mVs -1 (Fig. ). At negative voltages, the cell shows pure double-layer capacitance, with a current limited by the capacitance of the positive electrode. Upon positive polarization, I -is oxidized at the positive electrode with a significantly enhanced current, which is limited by the capacitive current of the oversized negative electrode. The negative current of virtually the same magnitude during negative polarization above 0 V indicates a high coulombic efficiency of the reversible Faradaic iodide/iodine reaction. |
60c74faf0f50db01a1397445 | 6 | The band forming around 224 cm -1 is an overtone of the I3 -band . At larger Raman shifts, changes in the carbon-related G-and D-band intensity (band width and position) can be observed (Fig. , Supplementary Fig. ). With increasing cell voltage, the widths of the G-and D-modes increase, while they are slightly redand blue-shifted, respectively and the ID-to-IG ratio is lowered. This can be explained by a superposition of charge transfer and an increase of defects , which is qualitatively consistent with ex situ results reported earlier . |
60c74faf0f50db01a1397445 | 7 | As positive polarization proceeds, initially dominant I3 -band intensity gives way to the dominance of I5 -band intensity (Fig. , curves from red to blue). Note that the I3 -and I5 -band intensities do not directly correlate with their concentration, since the scattering cross-sections of the polyiodides may vary significantly (resonant vs. non-resonant scattering) . To quantify the Raman band intensity changes, we deconvolute the Raman spectra as shown in Supplementary Fig. 26 . The band intensities as a function of the cell voltage (Fig. ) |
60c74faf0f50db01a1397445 | 8 | show an increasing I5 -to I3 -band intensity ratio with increasing cell voltage. Hence, relative to I3 -, I5 -forms at accelerated rates at higher cell voltages. To check whether I2 had formed (next to I3 -and I5 -), we soaked charged (positively polarized) and washed AC electrodes for several minutes in either pure water or 1 M NaI and recorded ex situ UV-vis spectra (Fig. , details in the Methods). While pure water would leave I2 confined to the AC nanopores, it should be dissolved in the 1 M NaI solution via I + I ⇌ I . Indeed, the 1 M NaI solution turned brownish, while pure water remained colorless (Supplementary Fig. ). The UV-vis spectra in Fig. confirm a large quantity of I3 -in the 1 M NaI solution and hence the presence of confined I2 in the charged positive AC electrode . As an independent proof, we detected iodine after immersing the charged positive AC electrode for several minutes in ethanol, which directly dissolves I2 (Supplementary Fig. ) . |
60c74faf0f50db01a1397445 | 9 | In situ Raman and ex situ UV-vis provide evidence for the formation of polyiodides (I3 -, I5 -) and I2 during I - oxidation. However, these methods fail to quantify the absolute amount of the species, to locate them in the nanopores, and to clarify whether I2 is a solute or solid. In situ SAXS/WAXS can afford all this (Fig. ). |
60c74faf0f50db01a1397445 | 10 | As before, we use an asymmetric two-electrode cell with oversized negative electrode and ran cell voltage sweeps with the scan rate controlled by the total cell voltage. The cell exhibits pure double-layer capacitance limited by the capacitance of the positive AC electrode at cell voltages below +0.2 V (Fig. ). Above this voltage, I -is oxidized at the positive AC electrode, leading to a significantly enhanced current (in line with electrochemical data of the in situ Raman measurements). The X-ray transmission of the electrolyte-filled positive AC electrode was simultaneously recorded by a photodiode placed behind the in situ cell, and quantifies the material flux in and out of the positive AC electrode (Fig. ). The significant increase of the Xray attenuation (details in Ref. ) indicates the accumulation of large amounts of iodide, iodine, and polyiodides within the AC pores. The relative scattering intensity as a function of time follows the applied voltage and resulting current (Fig. ). During positive polarization, the scattering intensity in the SAXS regime at q < 5 nm -1 and in the WAXS regime has a clear maximum at the peak voltage. |
60c74faf0f50db01a1397445 | 11 | Before analyzing the scattering intensity changes during positive cell polarization, we first must consider the data from the ex situ SAXS of the empty and electrolyte filled nanoporous carbon (Fig. ). The scattering intensity of the nanoporous carbon can be split into three additive terms: one accounting for spatial correlations on the molecular level, called structure factor, a second one accounting for the spatial correlations of the nanopores, which we refer to as nanopore scattering, and a third one accounting for the scattering contribution of the AC particles (with a size around 1 µm) . Given the reciprocal relationship between the size of realspace objects and their appearance on the scattering curve, the structure factor contribution is most prominent at large scattering vectors q (large scattering angles). The nanopore scattering appears as a distinct intensity hump at intermediate scattering vectors q, with the shape and the position containing information about the pore morphology, size distribution, and mean pore size. The increased scattering intensity at q < 0.8 nm -1 is caused by the scattering of the AC particles. The nanopore scattering intensity depends on the squared electron density difference between pores and carbon skeleton; therefore, filling of the nanopores with electrolyte leads to a significant decrease of the scattering intensity (Fig. , inset). At the same time, the electrolyte structure factor increases the scattering intensity in the WAXS regime. The distinct intensity peak around 20 nm -1 can be attributed to the first water structure factor peak (Supplementary Fig. ). |
60c74faf0f50db01a1397445 | 12 | Qualitatively, the large SAXS/WAXS intensity increase at low and large scattering vectors q (Fig. , red to blue for increasing polarization) can be assigned to different contributions. The distinct intensity hump around 3 nm -1 has a shape comparable to the nanopore scattering contribution of the empty AC shown in Fig. . This can only be explained by large amounts of a high-density compounds accumulating in its characteristic nanopores. The electron densities ρ of carbon skeleton and electrolyte-filled nanopores (insert in Fig. ) point at the formation of solid iodine, since dissolved polyiodides alone could not account for the necessary electron density increase. |
60c74faf0f50db01a1397445 | 13 | In the WAXS regime (q > 6 nm -1 ), distinct intensity peaks form around 7.5 nm -1 , 16 nm -1 , and 21.5 nm -1 (Fig. ). To better visualize the peak formation, we subtracted the structure factor contribution of the 1 M NaI electrolyte (that is, the WAXS intensity at 0 V cell voltage). We attribute the peak around 7.5 nm -1 to I3 - -I3 -correlations of by a highly concentrated triiodide solution in the cavities formed by nanoporous carbon and solid I2, justified by the mean distance between I3 -species being in the order of 2π/7.5 nm -1 = 0.84 nm. |
60c74faf0f50db01a1397445 | 14 | The peaks at 16 nm -1 and 21 nm -1 do not exactly match diffraction peaks of crystalline I2, indicating that solid I2 forms in a crystal structure slightly different from bulk I2, partially in an amorphous state and/or the confined nanocrystals are strongly distorted. The diffraction peak widths (FWHM), as obtained by Gaussian peak fitting (Supplementary Fig. ) are located in the region of 10.5°-19° 2θ. According to Scherrer's equation, this translates into a crystallite size between 0.55-0.9 nm, which fits well to the mean carbon pore size of 0.81 nm and the intensity changes found in SAXS. Qualitatively, the in situ SAXS/WAXS and Raman data analysis provide correlating complementary information. Based on these data, we see that during oxidation, solid iodine particles are deposited in the carbon nanopores. At the same time, significant amounts of I3 -and I5 -are generated. The WAXS correlation peak at 7.5 nm -1 suggests a highly concentrated solution of polyiodides that is, to a large extent, confined to the nanopores. Solid iodine domains are likely to be present in crystalline form, since diffraction peak widths point at crystallite sizes up to 0.9 nm, in accord with the mean carbon pore size of 0.81 nm. |
60c74faf0f50db01a1397445 | 15 | To quantify the amount of the formed I2 in the nanopores, we first separate the electrolyte/carbon structure factor (high q) and the particle scattering (low q) from the pure nanopore scattering (intermediate q). The separation procedure is described in the Methods and Supplementary Fig. . In a second step, we use the concept of plurigaussian random fields to fit the reduced experimental in situ SAXS curves and generate 3D lattice models of the carbon nanopore structure filled with the solid iodine phase. This involves modeling the empty carbon nanopore structure using a Gaussian random field (GRF) ( ) and the solid I2 via a second Gaussian random field ( ). After the carbon nanopore structure has been obtained from a fit to the SAXS intensity of the empty carbon electrode (Supplementary Fig. and Ref. ), the in situ SAXS data are fitted using the I2 pore occupancy and two structural parameters of the I2 phase as fit parameters. These are the correlation parameter lz of the GRF ( ) and the parameter accounting for carbon-iodine correlations (Supplementary Fig. ). A detailed description of the plurigaussian random field modeling and fitting is given in the Methods. |
60c74faf0f50db01a1397445 | 16 | The modeled in situ SAXS intensities fit well to the reduced experimental in situ SAXS intensities measured during positive polarization (Fig. -oxidation proceeds from red to blue). The increasing scattering intensity is caused by the increasing amount of solid iodine as visualized on cross-sections and 3D cut-outs in Fig. (see Supplementary Video of the 3D model at maximum pore filling). The overall shape of the scattering curve remains similar to the shape of the empty nanopore structure (Fig. and bottom red curve in Fig. at the beginning of I -oxidation). This finding is attributed to the similar characteristic feature size of the solid iodine structure and the empty carbon nanopores on the one hand (compare fit parameters lZ and lY, Supplementary Fig. ), and the correlation between the iodine and carbon structure on the other hand (fit parameter ). Given the size of the confining carbon pore structure and the crystallite sizes obtained from the Scherrer analysis (Supplementary Fig. ), the iodine particles should not be much larger than the mean pore size of 0.81 nm. The I2 pore occupancy reaches a maximum of 30% at the maximum positive cell voltage of 0.8 V (Fig. ). This I2 pore occupancy fits well with the value estimated from Faradaic capacity, assuming that all capacity forms I2 (blue line; in reality, a certain fraction also reacts to polyiodides). Given the poor electronic conductivity of I2, such high degree of pore filling is remarkable. We conclude that solid I2 rather than dissolved polyiodides represent the primary, capacity-relevant fraction of oxidized species. While in situ Raman data confirms the existence of significant amounts of I3 -and I5 -, relative to solid I2, their absolute amount is small at least up to charging capacities of 200 mAh gC -1 . The quantification via plurigaussian random fields, presented here, confirms the qualitative data interpretation from above. Contrary to the current state-of-the-art, oxidation of iodide in nanoporous carbon electrodes results in solid iodine nanoparticles, as well as dissolved I3 -and I5 -polyiodides. Processes in the positive electrode of a hybrid NaI supercapacitor are comparable to those in common conversion-type battery electrodes, such as Li-S or Li-O2 battery cathodes, where solid Li2S or Li2O2 precipitate from solution species . During charging and discharging, the solid active material is deposited within the nanopore network of a carbon cathode. The pore occupancy of active material hence determines achievable capacities and reversibility. |
60c74faf0f50db01a1397445 | 17 | The in situ and ex situ data allow deriving a detailed reaction mechanism of the (poly)iodide/iodine redox chemistry in nanopore confinement (Fig. ). Specifically, (i) in situ Raman, ex situ UV-vis and in situ SAXS (Fig. ) confirm the formation of I2, I3 -and I5 -during I -oxidation. (ii), I5 -generation is delayed with respect to I3 -upon oxidation process as seen in their Raman intensities (Fig. ). This indicates that larger amounts of generated I3 -and I2 accelerate I5 -formation. (iii), In situ SAXS/WAXS data confirm the formation of solid I2 nanoparticles or clusters in the nanopores. |
60c74faf0f50db01a1397445 | 18 | At first, I -is oxidized to I2 at the carbon-electrolyte interface (Equation ). This reaction leads to solid I2 nanoparticles with a size limited by the electron tunneling/conduction thickness of the insulating I2 and the confining carbon cavities. Concurrently, I2 comproportionates to some extent with I -to I3 -(Equation ), with the amount of the latter growing with the amount of I2 (steadily increasing amounts of I3 -and I2, Fig. ). As I3 -and I2 amounts grow in the nanopores, the I5 -formation via the comproportionation (Equation ) accelerates. |
60c74faf0f50db01a1397445 | 19 | The direct electrochemical generation of I3 -via 3 I → I + 2 e can be excluded since the further precipitation of I2 via Equation 2 would require to first reach an I3 -concentration which would drive disproportionation; hence, I3 -would level off while I2 grows. Instead, in situ Raman and SAXS show steadily and concurrently evolving amounts of I3 -and I2. Based on the equilibrium constant of Equation , the concentration of I3 -is orders of magnitudes higher than the I2 concentration at all times . The generation of large amounts of I2 via precipitation (Equation ) would require extremely high I3 -concentrations, which appears unlikely. |
60c74faf0f50db01a1397445 | 20 | The detected amount of solid I2 produced in the carbon nanopores during voltage cycling is quite remarkable, with a specific capacity reaching 200 mAh gC -1 . To gauge the maximum amount of iodine possible to be electrodeposited in the positive AC electrode, neither the negative AC electrode capacitance nor total I -in the system must limit iodine electrodeposition. This is realized with a large, about 100 times oversized AC counter electrode and a large electrolyte volume (details in Methods, cyclic voltammetry in Supplementary Fig. ). |
60c74faf0f50db01a1397445 | 21 | We oxidized I -at a constant potential of +0.6 V vs Ag/AgCl at the positive working electrode up to a certain charging capacity. After leaving the cell at open-circuit for different resting times (1-32 h), the remaining I2 was measured by a potentiostatic discharge to 0 V for 4 h (Fig. ). |
60c74faf0f50db01a1397445 | 22 | The self-discharge increases as the charging capacity of the positive AC electrode is increased (Fig. and Supplementary Fig. ). Independent of the positive electrode loading, already after a few hours resting time, the discharge capacities approached a similar level. This is also shown by charging to three different capacities (219, 330 and 441 mAh gC -1 ) and waiting for 4 h at OCV (Supplementary Fig. ). After 32 h of resting, the discharge capacities reached a value of 150 mAh gC -1 . Assuming that most capacity is stored in the form of solid I2 and by considering the known electrode porosities (Supplementary Table ) this corresponds to an I2 pore occupancy of 0.17. The proposed reaction mechanism explains the increasing self-discharge with increasing amounts of iodine loading/charging capacities. The more iodine is deposited, the more mobile I3 -and I5 -are formed chemically. |
60c74faf0f50db01a1397445 | 23 | Owing to their negative charge, a certain fraction of polyiodides will remain adsorbed in the nanopores of the positively polarized WE. The more polyiodides are generated, the higher is also their fraction that diffuses to the CE, where they are reduced to I -to promote self-discharge. Hence, future strategies to increase the iodine pore occupancy and reduce self-discharge should focus on reducing the concentration of shuttling polyiodides. |
60c74faf0f50db01a1397445 | 24 | Based on the reaction mechanism in Fig. , this involves reducing the I -concentration by, for example, using additional, electrochemically inert anions like NO3 -, which allow precipitating a larger fraction of the I -without anion starvation. Using aqueous 0.5 M NaI and 0.5 M NaNO3 as electrolyte (red curve, Fig. ), I2 pore filling and discharge capacities improve significantly. This not only demonstrates a practical way towards improved energy densities of carbon-iodine electrodes used in hybrid supercapacitors or batteries, but also independently confirms the reaction mechanism, derived from in situ Raman and SAXS/WAXS. |
60c74faf0f50db01a1397445 | 25 | In situ Raman and in situ SAXS/WAXS have shown that oxidation of I -in the confinement of carbon nanopores generates not only I3 -and I5 -polyiodides, but to a large extend solid iodine nanoparticles. Induced by the confinement of the carbon nanopores solid I2 dissolves only slowly during self-discharge. The derived reaction mechanism explains all recorded in situ spectroscopic, in situ scattering, and electrochemical data and suggests pathways to increase capacity densities and further reduce self-discharge. We demonstrate that an effective strategy to improve the iodine pore filling capacities and reduce self-discharge lies in reducing the concentration of iodide in the carbon nanopores in the charged state using an inert supporting electrolyte. |
60c74faf0f50db01a1397445 | 26 | Understanding that I -oxidation in carbon nanopores electrodeposits solid iodine calls for a paradigm shift in two aspects. First, increasing the capacity of such systems relies on strategies to increase the pore filling with solid iodine. Second, avoiding self-discharge means reducing the concentration of mobile polyiodide species generated by comproportionation. Until now, confined immobile polyiodides were thought to be responsible for the excellent self-discharge properties of iodine-based energy storage. |
60c74faf0f50db01a1397445 | 27 | Considering the shown strategies to improve capacity and self-discharge, the similarities with other conversion-type batteries, such as Lithium-sulfur, are remarkable. Carbon electrodes with a high degree of iodine pore filling are thus alternative, sustainable, and environmentally friendly conversion-type battery electrodes that may be used in combination with a supercapacitor counter-electrode (hybrid supercapacitor) or a battery anode material (e.g., zinc-iodine battery). |
60c74faf0f50db01a1397445 | 28 | From a broader perspective, the work shows the necessity of applying advanced in situ techniques for detailed insights into electrochemical mechanisms and to derive practical conclusions for real devices. We show that properties and function of complex emerging systems for electrochemical energy storage are equally controlled by chemistry (in situ spectroscopy) and the nanoscale structure (in situ scattering). In combination with stochastic modeling, in situ SAXS/WAXS allowed for quantifying the iodine pore filling and visualizing the (sub-)nm iodine phase evolution during electrochemical cycling. Hence, in situ SAXS proves as a powerful experimental method to study the phase evolution of active materials in nanoporous carbons and beyond. |
60c74faf0f50db01a1397445 | 29 | Free-standing carbon electrodes were prepared by mixing 90 mass% of the microporous KOH-activated carbon (MSP-20, Kansai Coke and Chemicals, Supplementary Fig. , Supplementary Table ) with 5 mass% carbon black (C65 from Imerys), and 5 mass% of polytetrafluoroethylene (60% dispersion in water from 3M Chemicals) in ethanol. After stirring at +70°C, the obtained dough was pressed and rolled on a glass plate into a thin sheet. Disc electrodes were punched from the sheet and dried at 120°C, resulting in a final thickness of about 150 μm. Sodium iodide (NaI, 99.5%) was purchased from Alfa-Aeser and dried at 110°C overnight before preparing the aqueous 1.0 M NaI electrolyte with de-ionized water. The pH of the electrolyte was 6.5 and the conductivity 82 mS cm -1 . |
60c74faf0f50db01a1397445 | 30 | In situ Raman spectroscopy was carried out using a LabRAM HR 800 spectrometer combined with an Olympus BX41 microscope. The spectra were measured from the top, with the focus plane underneath the standard microscope glass slip (covering the cell assembly) and the electrolyte film on top of the carbon electrode. A schematic drawing of the two-electrode electrochemical in situ cell ECC-Opto-Gas from EL-CELL (Hamburg, Germany) is given in Supplementary Fig. . As an objective lens a 40× Olympus LUCPlanFL N (NA=0.6; corrected for the cover thickness) was used. To avoid sample damage, the laser with 532 nm wavelength worked at reduced power (0.5 mW). With the given grating/slit setup (300 mm -1 ; 200 µm) the spectral (pixel) resolution is about 3.6 cm -1 . The acquisition time per spectrum was 30 s (4-fold accumulations) for a total measurement time and time resolution of 120 s/spectrum. In addition, the DuoScan System was applied to continuously scan the laser spot over a 20×20 µm area. With regard to the spectral analysis the band deconvolution of the I3 -/I5 -bands (Supplementary Fig. ) was done using three Lorentz peaks (initial position 110 cm -1 , 143 cm -1 , and 224 cm -1 ) and one Gaussian peak (initial position 165 cm -1 ), this routine is based on the band assignments from Ref. . The deconvolution of the G-and D-band was done according to the method introduced in Ref. . |
60c74faf0f50db01a1397445 | 31 | Positive carbon electrodes (mass = 3.0 mg) were charged in a two-electrode cell with oversized counterelectrode (mass = 34.0 mg) by voltage sweep (2.0 mV s -1 ) to a cell voltage of 1.0 V. After charging, the positive electrodes were taken out, washed for 1 min with deionized water and placed for 3 h in three different solutions: |
60c74faf0f50db01a1397445 | 32 | In situ SAXS/WAXS measurements were carried out on a custom-built two-electrode electrochemical in situ cell. The cell consisted of the positive AC electrode, a glass fiber separator (Whatman), a negative AC electrode and platinum foil current collector (CC). The assembly was sealed in a stainless-steel housing with an adhesive film (Tixo) X-ray windows. To enhance the recorded signal by increasing the amount of oxidized I -the negative electrode was oversized 8-times (mass = 32 mg, thickness = 600 µm, diameter = 12 mm) compared to the positive electrode (mass = 4 mg, thickness = 200 µm, diameter = 8 mm). Oversizing the negative electrode requires increasing its thickness rather than increasing the diameter to avoid detrimental, large ion diffusion pathways. To investigate only processes in the positive electrode, a hole with 2 mm diameter was cut into negative electrode, separator and CC. SAXS/WAXS data were recorded on an in-house SAXS facility using Cu-Kα radiation and a 2D areal detector (Dectris Eiger R 1M) with a nominal sample-to-detector distance of 100 mm. The primary 2D scattering data were azimuthally averaged and normalized by transmission values and recording time. Further, intensity data were multiplied by polarization 2 (1 + (cos 2 ) ) ⁄ and absorption 1 -!" #$ % & ln -% )* + ln & , correction factors to adequately normalize scattering intensities at large scattering angles 2 . The background scattering intensity was recorded after the in situ SAXS measurements and removing the WE; and subtracted from in situ SAXS/WAXS data after primary data treatment as described above. |
60c74faf0f50db01a1397445 | 33 | To separate the nanopore scattering contribution from the electrolyte (+ carbon) structure factor and the lowq particle scattering (Supplementary Fig. ), we first determine the electrolyte structure factor using a modified Porod fit of the form -.*/*0 = 2 3 4 + 5 • 78(3) ⁄ at 6.5 nm -1 < q < 9 nm -1 . 78(3) is the measured scattering intensity of the bulk electrolyte, 5 a multiplicative constant, and P the porod constant. The latter two parameters are determined by the fit. Since the nanopore scattering intensity is fitted with a simple two-phase model (see below), density fluctuations within the carbon matrix are neglected. After subtracting the electrolyte structure factor contribution 5 • 78(3), the particle scattering contribution is determined from a power law fit of the form -. /9:);< = = 3 .? + @ ⁄ at 0.4 nm -1 < q < 1 nm -1 , with B and C being the unknown fit parameters. The exponent of 3.73 is determined from a power law fit on the empty AC electrode. The obtained particle scattering = 3 .? ⁄ is then subtracted from the experimental SAXS intensities. |
60c74faf0f50db01a1397445 | 34 | The in situ SAXS data fits apply the concept of plurigaussian random fields , as explained below. We assume the carbon nanopore structure and the corresponding GRF ( ) to be constant during the in situ SAXS experiment. The GRF parameters are taken from a fit to the experimental ex situ SAXS intensity of the empty nanoporous carbon electrode (Supplementary Fig. and Supplementary Table ). Therefore the experimental SAXS curve was normalized by its integrated intensity |
60c74faf0f50db01a1397445 | 35 | The in situ SAXS intensities (Fig. ) were fitted to three fit parameters: the I2 pore occupancy F P , the correlation parameter M Q of the iodine phase GRF ( ( )), and the parameter , accounting for ( ) (iodine) and ( ) (carbon) correlations. The I2 pore occupancy values were constrained to values smaller than 0.35, estimated from electrochemical capacity values (Fig. ). The value O Q was kept constant at a value of 30 nm (O Q ≈ ∞). To fit the in situ SAXS intensities, we minimized the sum of the squared residuum of all experimental SAXS intensities during a positive potential sweep. In other words, the parameters M Q and were held constant while the iodine pore occupancy was used as fit parameter for each SAXS curve at its specific state of charge. The procedure was repeated for several values of M Q and (Supplementary Fig. ) and the parameter set with the minimum sum of squared residuum values taken as the solution. |
60c74faf0f50db01a1397445 | 36 | We model the in situ SAXS data using the concept of plurigaussian random fields . This allows retrieving 3D real space models of the solid iodine phase evolution within the carbon nanopore structure and quantifying the degree of iodine pore filling (Fig. ). A detailed description of how plurigausian random fields are applied to in situ SAXS data of three-phase systems is given by Gommes et al. . |
60c74faf0f50db01a1397445 | 37 | Before modelling the iodine phase evolution, we generate a 3D model of the empty nanoporous carbon structure using a fit to the ex situ SAXS curve of the empty carbon electrode and the concept of clipped Gaussian random fields. Detailed descriptions of the procedure are given in Refs. . In brief, a Gaussian random field ( ) is the sum of cosine waves with wave vectors distributed according to their power spectral density `N(a) and phase factors b ^ randomly distributed between 0 and 2c. ( ) = d 2 e f cos (g h • -b |
60c74faf0f50db01a1397445 | 38 | Three-electrode Swagelok-type cells with a Ag/AgCl reference electrode were assembled for cyclic voltammetry (Supplementary Fig. ) and to test the maximum amount of iodine that can be electrodeposited in the positive AC electrode (potentiostatic charge/discharge, Fig. ). We used an AC working electrode (diameter = 3 mm) with a mass of 0.3 mg, an AC counter-electrode (diameter = 10 mm) with a mass of 30 mg and about 250 µl electrolyte (three Whatman separators with 10 mm in diameter). This ensures that neither the total amount of I -in the cell nor the double-layer capacitance of the counter-electrode is limiting the I2 uptake in the AC WE pores. |
6465cb15a32ceeff2dd41455 | 0 | Designing the spatial environment around a metal center is a critical issue in terms of the electronic and/or steric properties and hence the reactivity of organometallic complexes. Chemists have thus focused on the design of the structural, electronic, and dynamic properties of supported ligands, as demonstrated in the field of ,e.g., homogenous catalysis, supramolecular chemistry, and materials science. One well-known strategy for reversibly modulating the electronic/spatial environment around metal centers is based on a ligand-substitution process on metal centers bearing multifunctional ligands including a hemi-labile coordination moiety (L') in the presence of an external Lewis base (LB) (Figure , left). However, these processes often yield an equilibrium mixture that is difficult to separate. In this context, we have reported the reversible, recyclable, and pressure-responsive room-temperature-chemisorption of carbon monoxide (CO) on a zero-valent nickel complexes that bear multifunctional multipurpose carbene ligands, namely N-phosphine oxide-substituted imidazolinylidenes (SPoxIms) and the corresponding imidazolylidenes (PoxIms) (Figure , left). (S)PoxIm ligands in tetrahedral (S)PoxImNi(CO)n complexes selectively adopt either a -C-(n = 3) or -C,O-(n = 2) coordination mode, which can be interconverted through the addition/exclusion of CO. |
6465cb15a32ceeff2dd41455 | 1 | A strategy based on the use of Lewis acids (LAs) is another potential option for reversibly modulating the electronic/spatial environment around metal centers that bear multifunctional ligands (Figure , right). The LA-mediated process can cause a substantial change in both the electronic state and spatial environment of the metal center due to the change in the number of ligands. This feature clearly distinguishes LA-mediated processes from LB-mediated ones, as the latter commonly proceeds without changing the number of ligands. However, examples of such LA-mediated reactions have been severely limited, because LAs often trigger the irreversible decomposition of the organometallic complex via the abstraction of ligands to form an LA-LB adduct. A judicious combination of multifunctional ligands and LAs is thus essential for constructing an LA-mediated system that can reversibly regulate the electronic/spatial environment around the metal centers. It should be noted that Fan et al. have reported an example that relies predominantly on an electrostatic interactions involving the reversible complexation of Na + and a crown-ether moiety included in a Rh complex that bears Aza-CrownPhos; the reactivity of the Rh complex can be regulated through the Na + -mediated interconversion. The authors evaluated the change in the spatial environment around the Rh center using multinuclear NMR spectroscopy and ESI mass spectrometry, even though a quantitative evaluation of how much the field expanded/diminished via the reaction was not discussed. |
6465cb15a32ceeff2dd41455 | 2 | Herein, we present a novel strategy to reversibly transform the spatial environment around a nickel center by combining (S)PoxIms 1a-d (Figure ) as multifunctional ligands and tris(pentafluorophenyl)aluminum as an LA, where the geometry of the Ni center is interconverted between tetrahedral and trigonal planar (Figure , right). We quantitatively evaluated the variation of the spatial environment, i.e., the volume and shape of the space, and the electronic state around the nickel center based on single-crystal X-ray diffraction (SC-XRD) and X-ray absorption spectroscopy (XAS) analyses. |
6465cb15a32ceeff2dd41455 | 3 | Previously, we have reported that (S)PoxIms include synand anti-conformers with respect to the N-P bonds, which can interconvert easily at room temperature via rotation of the N-phosphinoyl groups (∆E ‡ ~12 kcal mol -1 for previously reported PoxIms), even though the anticonformers are thermodynamically favorable (Figure ). Importantly, the volume and shape of the space can be drastically varied via rotation of the N-phosphinoyl groups, and a relatively limited space is present around the carbene carbon atoms in the anti-(S)PoxIms. Due to this limited spatial environment, only group-11 metals such as Cu and Au have been confirmed to form complexes that bear anti-(S)PoxIms in a -C fashion, given that these metals tend to adopt two-coordinated linear complexation geometries (Figure ). We have also reported the synthesis (IAd)Ni(CO)2 (I t Bu = 1,3-di-tert-butylimidazol-2-ylidene; IAd = 1,3-di-adamantylimidazol-2-ylidene) has already been reported which inspired us to explore the preparation of a Ni(0) complexes bearing bulky anti-(S)PoxIm ligands (Figure ). |
6465cb15a32ceeff2dd41455 | 4 | We started our study with the reaction between (syn--C,O-1a)Ni(CO)2 (2a) and an equimolar amount of Al(C6F5)3(tol)0.5, which resulted in the selective formation of heterobimetallic Ni/Al complex 3a in 90% yield (Figure ). We also confirmed that the Al(C6F5)3-mediated rotation of N-phosphinoyl moieties in 2b-d afforded the corresponding heterobimetallic Ni/Al complexes (3b-d) in good-to-high yield. The molecular structures of 3ad were fully characterized using NMR and ATR-infrared (IR) absorption spectroscopy as well as SC-XRD analysis (vide infra). Subsequently, we treated 3a with slightly excess amounts of 4dimethylaminopyridine (DMAP) and confirmed the quantitative regeneration of 2a with the concomitant formation of the adduct DMAP-Al(C6F5)3 (Figure ). These results demonstrated that the interconversion between 2a and 3a can be successfully mediated by the addition/removal of Al(C6F5)3. |
6465cb15a32ceeff2dd41455 | 5 | Next, we compared the electronic and geometric parameters of 2a and 3a. In the 31 P NMR spectrum of 3a, the resonance corresponding to the N-phosphinoyl moiety is observed at P 82.0 in toluene-d8, which represents a significant downfield shift compared to that of 2a (P 68.0 in THF-d8), while it is almost identical to that of (-O-1a)Al(C6F5)3 (P 79.8 in CD2Cl2). These results indicate that the 1a moiety in 3a should adopt predominantly the anti-conformation. The geometric parameters of 3a obtained from the SC-XRD analyses are shown in Figure . As previously reported, the Ni center in 2a, which bear the -C,O-syn-1a ligand (C-N-P-O torsion angle: 8.5(1)°), adopts a tetrahedral geometry. In stark contrast, 3a features a trigonal-planar Ni center (sum of bond angles around Ni: 358.9°) bearing anti-1a (C1-N2-P-O1 torsion angle: 159.7(2)°), as shown in Figure . The Ni(CO)2 unit is located slightly out of ideal alignment with the carbene lone pair (Z1-C1-Ni angle: 169.9°; Z1 = centroid of the imidazolinylidene ring), most likely due to the high steric demand of the t Bu groups. |
6465cb15a32ceeff2dd41455 | 6 | Interestingly, a non-covalent interaction between the Ni and H1 atoms was confirmed in the gas-phase-optimized structure of 3a (level: PBE0/def2-TZVPD//M06L/def2-SVPD for O, F, and Ni; def2-SVP for all other atoms) via a topological analysis of the electron density, which was calculated using the quantum-theory-of-atoms-in-molecule (AIM) method (Figure ). It should also be noted here that the geometric parameters of the solid state structure of 3a confirmed by SC-XRD analysis were closely reproduced in the theoretically calculated gasphase-optimized structure, e.g., the interatomic distances between the Ni and C5 atoms was 3.18 Å in the gas-phase and 3.240(4) Å in the crystalline state. In fact, the Ni•••H1 distance (2.13 Å) and the relatively high electron density ( = 0.026 e rBohr -3 ) as well as the positive value of the Laplacian of the electron density (∇ 2 = +0.061 e rBohr -5 ) at the bond-critical point (BCP) between these atoms, suggests the participation of an agostic Ni•••H interaction. Dissociation of the N-phosphinoyl moiety from the tetrahedral 2a to afford trigonal planar 3a also caused obvious changes in the wavenumbers of the stretching vibration of the CO ligands. |
6465cb15a32ceeff2dd41455 | 7 | In fact, the signal corresponding to the stretching vibration of the CO ligands in 3a was observed at 2019 cm -1 , i.e., at higher wavenumber than that of 2a (1967 cm -1 ). The AIM analysis clarifies the increase in the delocalization indices (Ni, CCO) and (Ni, OCO) in 3a compared to those in 2a, which represent a number of electron pairs delocalized between the Ni and carbonyl carbon atoms (CCO) and the Ni and carbonyl oxygen atoms (OCO), respectively. (Figures S7-S8. The mean values of (Ni, CCO) and (Ni, OCO) increase from 0.51 (2a) to 1.20 (3a) and from 0.063 (2a) to 0.23 (3a), respectively, indicating that electrons become effectively delocalized through the Ni, CCO, and OCO atoms in 3a. On the other hand, the values of (CCO, OCO) are nearly identical between 2a (1.70-1.71) and 3a (1.60-1.62). Thus, we attribute the aforementioned blue-shift to the increased polarization of the CO bonding orbitals due to an increased cationic nature on the Ni in 3a (AIM atomic net charge = 0.50 e) compared to 2a (AIM atomic net charge = -0.1 e). Based on the aforementioned geometric parameters of 2a and 3a in the crystalline state, the impact of the N-phosphinoyl rotation on the spatial environment surrounding the Ni center was quantitatively evaluated on the basis of percent buried volume (%Vbur) (Figure ). Topographic steric maps that visualize the %Vbur values in the north-western (NW), north-eastern (NE), south western (SW), and south-eastern (SE) quadrants, as well as their average (Av), are also depicted; these maps were produced using the program SambVca 2.1. The Av(%Vbur) values in 2a and 3a were calculated to be 43.9 and 51.2, respectively, clearly demonstrating that the space around the Ni center is significantly altered via rotation of the N-phosphinoyl moiety. |
6465cb15a32ceeff2dd41455 | 8 | Importantly, the drastic expansion/contraction of the space can be precisely and reversibly triggered by the addition/removal of Al(C6F5)3. In addition, this rotational transformation alters the shape of the space around the Ni center, as clearly confirmed by the comparison of both the %Vbur(NE) and %Vbur(SE) values of 2a and 3a. Å; bond radii scaled by 1.17; mesh spacing 0.05; H atoms are omitted) based on the structural parameters obtained from the SC-XRD analysis. |
6465cb15a32ceeff2dd41455 | 9 | We then carried out solution-phase XAS measurements to determine whether the results obtained from the SC-XRD analysis could be extended to the solvated states of 2a and 3a. A magnified view of the pre-edge and edge region of the Ni K-edge X-ray absorption near edge structure (XANES) is shown in Figure . In the case of 2a, a pre-edge peak corresponding to the electric dipole transition from the Ni 1s orbital to the Ni 4p(-3d) orbitals was observed at 8326 eV and no characteristic peak was detected at the absorption edge, which is typical for nickel complexes with tetrahedral geometry. Meanwhile, a remarkable peak appeared in the edge region (8331 eV) for 3a, along with the greater pre-edge peak intensity compared to 2a. We attribute the characteristic edge peak to the Ni 1s→4pz transition in 3a, given the presence of the non-bonding 4pz orbital (a2'' symmetry) (Figure ). Considering that such non-bonding 4pz orbitals are commonly found in trigonal-planar 16-electron complexes, the results of the solutionphase XAS experiments clearly rationalize the adoption of a trigonal-planar geometry by 3a, even in toluene, as observed in the crystal structure. |
6465cb15a32ceeff2dd41455 | 10 | To gain further insight, the Ni K-edge XAS spectra of 2a and 3a were simulated using the time-dependent DFT (TDDFT) calculation at the B3LYP level with the zeroth-order regular approximation (ZORA) to take the relativistic effect into consideration. The CP(PPP) basis set was used for Ni, while the ZORA-def2-TZVP(-f) basis set was used for all other elements. |
6465cb15a32ceeff2dd41455 | 11 | In the case of 2a, the transition from the Ni 1s orbital to the LUMO+2 significantly contributes to the pre-edge peak, where the LUMO+2 mainly consists of the Ni 3d (6.7%), Ni 4p (7.8%), and CO 2p (30.7%) orbitals (Figure ). On the other hand, in the case of 3a, the increased intensity of the corresponding pre-edge peak can be rationalized predominantly by the transition from the Ni 1s orbital to the LUMO. The LUMO of 3a features increased contribution from the Ni 4p orbitals (17.1%) and CO ligands (52.3%) compared to their contributions to the LUMO+2 in 2a (Figure ), while the Ni 3d orbitals exhibit a nearly identical degree of contribution to the respective MOs in 2a and 3a. We therefore attributed the increase in the pre-edge peak intensity of 3a induced by the Al(C6F5)3-mediated geometric change to the increased efficiency of the Ni 3d-4p mixing and contribution of CO ligands (e' symmetry) in the trigonal planar geometry. |
6465cb15a32ceeff2dd41455 | 12 | Finally, we further explored the electronic structures of the Ni 3d orbitals in the solid state for 2a and 3a based on Ni L2,3-edge XAS experiments using the partial fluorescence yield (PFY) method. In the cases of 2a and Ni(cod)2, the absorption maxima on the Ni L3-edge appear at 856.0 and 856.2 eV, respectively, with a shoulder peak in their lower-energy regions (Figure ). These Ni L2,3-edge XAS spectra represent spectroscopic features characteristic of tetrahedral nickel complexes with a d 10 electron configuration (formally Ni(0)). In contrast, for 3a, the peak maximum shifts toward the lower-energy region to 852.9 eV, clearly indicating that the 3d orbitals are stabilized through the formation of the 16-electron complex with a vacant 4pz orbital. In fact, the energy levels of the unoccupied frontier MOs of 3a are substantially lower than those of 2a, showing good agreement with the experimental Ni L2,3-edge XAS data (Table ). Furthermore, a broad signal is observed in the higher energy region at 854-858 eV. Based on the results of the In summary, we have reported a strategy that reversibly modulate the electronic state and spatial environment around metal centers that bear multifunctional ligands based on the use of Lewis acids. To this end, a series of N-phosphine oxide-substituted N-heterocyclic carbenes, referred to as (S)PoxIms, was employed, as (S)PoxIms can afford heterobimetallic species in a - O-Al(C6F5)3} through the complexation-induced rotation of the N-phosphine oxide moiety, while the addition of DMAP quantitatively triggered the formation of the former complex. We experimentally and theoretically confirmed that the shape and size of the space around the Ni(0) center drastically expanded/contracted through this Lewis-acid-mediated procedure. Furthermore, a detailed discussion based on multinuclear NMR, IR absorption, and X-ray absorption spectroscopy shed light on the changes in the electronic states of the Ni centers. Thus, this work manifests a conceptually novel and effective approach to design and modulate the electronic and spatial environment surrounding metal centers in organometallic compounds using a combination of multifunctional ligands and Lewis acids. |
60c73e14337d6ca46ae26282 | 0 | In the 60+ years since the introduction of Cahn-Ingold-Prelog Sequence Rules in 1956, the "CIP Rules" have become an integral part of chemical nomenclature, providing a way to identify the spatial arrangement of atoms of a molecule using simple mostly atom-or bond-based stereodescriptors. Over the course of this time, various authors have pointed out deficiencies in the rules and proposed solutions in the form of modifications and subrules, to the point where today we have eight distinct Sequence Rules: 1a, 1b, 2, 3, 4a, 4b, 4c, and 5. |
60c73e14337d6ca46ae26282 | 1 | In order to provide a single resource summarizing the state of the evolving rules, the International Union of Pure and Applied Chemists (IUPAC) published the first comprehensive description of the CIP Rules in Nomenclature of Organic Chemistry: IUPAC Recommendations and Preferred Names 2013 (referred to below as "BB 2013"). This description, an impressive 197-page chapter with more than two hundred examples, presents a complete set of CIP Rules, along with detailed procedures for their application. |
60c73e14337d6ca46ae26282 | 2 | The "CIP descriptors" generated by these rules are primarily intended for use in chemical nomenclature. It would be possible to use them in other applications, such as the removal of redundant stereo specifications, the determination of structure equivalence, or canonical labeling. However, CIP-based algorithms may be more resource consuming and complex than simpler, more efficient cheminformatics algorithms developed specifically for those purposes. |
60c73e14337d6ca46ae26282 | 3 | We note that discoveries of deficiencies in the original rules have been made before in the context of developing computer-based implementations. This was certainly the case in 1993, when subrules 4a, 4b, and 4c were proposed by a group developing a stereochemical module for the LHASA computer-aided synthesis analysis program. Not surprisingly, predating the open-source collaborative environment and not having a concise reference in hand, machine implementations of the CIP Rules to date have been only marginally successful. Certainly, many software developers have implemented the CIP Rules to one extent or another, but recent analysis of available software packages clearly demonstrates that there is much disagreement among these implementations, even for relatively simple compounds, among several highly respected software packages. In the spring of 2017, concurrent discussions started in two forums, the IUPAC Blue Book Project, focused on preparation of errata for BB 2013, and the Blue Obelisk Group, focused on implementation issues of CIP rules in algorithmic form. The result of these lively discussions has been a reason for coordinated and thorough analysis of the CIP Rules, with the aim of concurrent development and improvement of four software packages: Jmol , Centres , ChemSketch , and Balloon . Our joint efforts ensured that multiple, validated, independent machine implementations of the CIP Rules are made available to the cheminformatics community, as well as to anyone interested in stereochemistry and chemical nomenclature. In fact, the results of our work have already been incorporated into an interactive web site utilizing JSmol. As we sought consensus when there were issues and questions as to interpretation of the rules or correctness of BB 2013 examples, we turned to the use of finite acyclic digraphs, the assumption being that analysis of digraphs should always be the final arbiter in any dispute relating to stereochemistry. The use of digraphs in our discussions allowed for one of three possible conclusions: (a) that the disagreement was due to different interpretations of CIP rules among software developers, (b) that there was a problem with an algorithm or its implementation in code, or (c) that the CIP rules themselves were flawed. In fact, all three of these possibilities were encountered in the process of coming to consensus, including the discovery of a small number of errors in the Blue Book, two minor flaws in the CIP rules, and a proposal for a new rule. |
60c73e14337d6ca46ae26282 | 4 | Before discussing the Sequence Rules, we note that BB 2013 is inconsistent in its use of terminology in relation to duplication of nodes. In the discussion that follows, and in our proposed revision of these rules, we use the following terminology exclusively: duplicate node A digraph node that has been added as a copy of a "real" atom. duplicated atom A digraph node that represents the real atom that has been duplicated. |
60c73e14337d6ca46ae26282 | 5 | Digraphs and Auxiliary Descriptors. The importance of using finite acyclic directed graphs ("digraphs") and auxiliary descriptors (temporary assignments made in the process of determining a specific center's descriptor) was introduced by Prelog and Helmchen in 1982 in relation to cyclic structures and emphasized later by Mata et al. in relation to the process of developing Rule 4: |
60c73e14337d6ca46ae26282 | 6 | Our consensus interpretation of this statement is that, at least for cyclic compounds and compounds with more than one stereocenter, a unique digraph must be generated for each center in question. This includes double-bond as well as axial stereochemistry and is true whether the center in question is in the ring or not. Temporary auxiliary descriptors are assigned solely on the basis of this digraph and may or may not be the "final" descriptors ultimately used to describe those centers using their own digraphs. Figure shows an example (VS279) where only a minority of the auxiliary descriptors are the same as the final descriptors for the corresponding atoms. Generation of a complete digraph, including all auxiliary descriptors, including seqcis and seqtrans, is required prior to Rule 3. The entire sequence of all rules must be carried out for each auxiliary center on the same digraph. Generation of auxiliary descriptors must start from the highest sphere, proceeding toward the root. In this way, all auxiliary descriptors in higher spheres than the one being determined are already assigned. This is sufficient, as the descriptor for an auxiliary center does not depend upon any descriptor between it and the root. The auxiliary center ligand leading to the digraph root can always be ranked by Rule 1a exclusively, as auxiliary centers are offset from the root of a digraph, and so the path back to the root is always unique in connectivity and atomic numbers. |
60c73e14337d6ca46ae26282 | 7 | However, from an implementation perspective, simplified digraphs must be used with caution. In complex examples, for instance, a critical aspect of the algorithm must be finding the first difference in two ligands, and this may not be obvious from a simplified digraph. For example, there is no rule that "real atoms have higher priority than duplicate nodes." This is intentional. While generally true, this statement hides the fact that duplicate nodes lose to their real counterparts only in the next higher sphere, where their associated phantom atom, with atomic number zero, always loses to any real atom. That being the case, some other consideration may be missed. An example of a compound for which proper consideration in this regard is required is the bicyclic alkene shown in Figure (VS172). Naïve application of the pseudo-rule "real atoms have higher priority than duplicate nodes" leads to a 1R descriptor rather than correct 1S assigned in accord to Rule 1b. Sequence Rules and Subrules. While BB 2013 specifies that each of Rules 1-5 must be applied "exhaustively in the order given," it may not be clear what exactly "exhaustively" means or how this relates to subrules such as 1a and 1b. In fact, each subrule must be applied sequentially. |
60c73e14337d6ca46ae26282 | 8 | Effectively, there are eight (nine, if one adds our proposed Rule 6) fully independent explicit Sequence Rules: 1a, 1b, 2, 3, 4a, 4b, 4c, and 5. The fact that these are not "Sequence Rules 1-8" is simply a result of the historical evolution of the Sequence Rules. "Exhaustively" simply means "until a decision is reached, or it is determined that no such decision is possible." |
60c73e14337d6ca46ae26282 | 9 | Ranking and Comparing Ligands. As described clearly in BB 2013, application of all eight rules involves the same two processes: Ligands are ranked sphere by sphere, branch by branch in a breadth-first fashion. Then, in pairs, two ligands are compared atom by atom, in order of that ranking. In practice, it is not always necessary to completely rank a ligand, including hydrogen and phantom atoms. Rankingat least through Rule 2can be carried out on a "need-to-know" basis, skipping whole sub-branches of the digraph where a decision has already been made. |
60c73e14337d6ca46ae26282 | 10 | On the face of it, Rule 1a sounds simple enough to implement. And it is, except for the special cases discussed in BB 2013 P-92.1.4.4 in relation to compounds and ions with multiple chemically equivalent Kekulé structures, such as benzene, pyridine or cyclopentadienyl anion. That section briefly introduces the idea of "atomic number averaging" for mancude-ring systems, which are "rings having (formally) the maximum number of noncumulative double bonds, e.g. benzene, indene, indole, 4H-1,3-dioxine". The idea is to average the atomic number of the duplicate node when it is involved in multiple resonance structures. Our reading of Section P-92.1.4.4 is that it is not a wellcrafted guideline, with many unanswered questions. What exactly defines the pertinent cases for which this rule applies? What about acyclic cases such as allyl anion or acetoacetate? How exactly is the averaging to be done -Do we need to know the exact weighting of all the possible resonance structures, or is it sufficient to average over just the adjacent atoms involved in the electron delocalization? |
60c73e14337d6ca46ae26282 | 11 | Without going into detail here, suffice it to say that atomic number averaging is a difficult procedure to describe or implement, other than the case of all-carbon neutral species, for which it is unnecessary, and some simple heterocyclic systems, such as pyridine derivatives. Thus, consider substituted pyridine VS032/ VS033, Figure . Notice that without this consideration, Rule 1a gives two different results in this case, depending upon the choice of Kekulé structure. The correct result, S, derives from assigning the duplicate node for the aromatic nitrogen an averaged atomic number of 6.5, which loses to the unaveraged atomic-number 7 duplicate node of the imine nitrogen. |
60c73e14337d6ca46ae26282 | 12 | Shortly into our study it became clear that the current IUPAC recommendation for Rule 1b is not sufficient. The problem is that although Rule 1b was designed to solve a problem with ring-closure duplicate nodes, the rule as stated also applies to multiple-bond duplicate nodes. As such, we again have in Rule 1b the same issue involving multiple Kekulé structures as for Rule 1a. The problem involves cases such as shown in Figure , where the application of Rule 1b as currently recommended can lead to the inappropriate introduction of stereodescriptors. The issue is just one specific case of more general problem of absent procedures to assign root distances for duplicates resulted from averaging of atomic numbers. Assigning distance to root using the current Rule 1b gives the same problem as found for Rule 1a if an averaging criterion is not applied. |
60c73e14337d6ca46ae26282 | 13 | Our discussion revolved around how to address this issue algorithmically: (a) Should we implement an averaging scheme as in Rule 1a for all six duplicate nodes? (b) Do we omit multiplebond duplicate nodes from consideration--just assigning them "n/a" and skipping them entirely? (c) Should we assign the distance to the root of their sphere? (d) Do we assign the distance to the root of their attached atom? Our group decided that the first of these options was too complex, the second would lead to ambiguities in the algorithm, the third would not work, and the simplest and surest solution would be the last of these --to assign to a multiple-bond duplicate node the distance to the root of its corresponding attached atom, not its corresponding duplicated atom. |
60c73e14337d6ca46ae26282 | 14 | Thus, the digraph for bis-(2-hydroxyphenyl)methanol is shown in Figure , where numbers are distances to the root assigned for each node needed for Rule 1b (not the usual atomic numbers seen on digraphs in BB 2013). The specific representation of the bonding is no longer significant, and the center is found to have no descriptor, as expected. There is an additional aspect of Rule 1b we suggest revising. The original statement of Rule 1b includes an additional criterion ranking any duplicate node higher than any node that is not a duplicate node for these purposes. The consensus of our group was to recommend changing the language of Rule 2 to be specific, using exact isotopic mass when an isotope is indicated, and atomic weight when it is not. |
60c73e14337d6ca46ae26282 | 15 | As, 89 Y, Nb, 103 Rh, 127 I, Cs, 141 Pr, Tb, Ho, Tm, Au, Bi, Pa, and 232 Th) to be equivalent whether their isotope number is given explicitly or not, as is the case chemically. In addition, the elements Tc, Pm, Po, At, Rn, Fr, Ra, Ac, and all elements with atomic number > 92, have no natural abundance; their "atomic weight" found on the periodic table is just one of their integer isotope mass numbers, as though that isotope were 100% naturally abundant. |
60c73e14337d6ca46ae26282 | 16 | It may seem that the switch to exact isotope mass from integer isotope number might be difficult to implement, requiring access to a complete table of isotopes, but that is not the case. It turns out that we can use integer isotope mass numbers provided we take account of just four anomalies: O, Cr, Mo, and 175 Lu. These four isotopes are the only ones that have exact masses slightly below their element's atomic weight even though their mass number is above it. For example, the exact mass of O is 15.994, which is below the element's atomic weight of 15.999, even though its mass number ( ) is higher. In practice, this is no problem. We simply use integer isotope numbers, but reduce these four by 0.1 for the purpose of setting priorities. For example, since the atomic weight of oxygen is 15.999, when O is compared to O, there is no problem --17 > 15.999. But when O is specified, we use 15.9 instead of its actual value of 15.994. This allows 16 O to have the required lower priority than "O" itself, with atomic weight 15.999. No other isotopes have this problem, and we can just use their unadjusted integer mass number as a surrogate for isotopic mass. In this way, there is never a need to check a table of exact isotope masses. |
60c73e14337d6ca46ae26282 | 17 | The exclusion of duplicate nodes by assigning their mass to be zero guarantees this. It also removes an issue similar to the one discussed for Rules 1a and 1b, that different arrangements of double bonds in conjugates systems must not affect the chirality. For example, isotopically labeled alcohol VS007, shown in Figure , is achiral. To ensure this result, we simply assign all duplicate node masses to be zero, allowing the mass difference to be carried only by their corresponding duplicated atom. As mentioned above, generation of a complete digraph, including auxiliary descriptors, is required prior to Rule 3 (not just Rule 4a, as mentioned in BB 2013). This is because seqcis and seqtrans also describe double bonds that involve two diastereomorphic ligands on the same end (with stereodescriptors RR and SR, for example). Inverting about a plane changes the comparison (to SS and RS, in this case), but this change does not reverse priorities. |
60c73e14337d6ca46ae26282 | 18 | Placement of Rule 3 before Rule 4a ensures that only enantiomorphic (seqCis and seqTrans) comparisons involving double-bonds and cumulenes with an odd number of double bonds are left to consider in Rules 4 and 5. From an implementation point of view, application of Rule 3 is simply the comparison of two ligand pairs, one pair on each end of an alkene or cumulene. |
60c73e14337d6ca46ae26282 | 19 | Rule 4b is by far the most difficult rule to comprehend and implement. One simplification is that although Rule 4b, as stated in BB 2013, refers to all possible mixes of R/S, M/P, and seqCis/seqTrans descriptors, for implementation purposes, all auxiliary descriptors can be normalized by labeling them either R or S. For example, any of R, M, or secCis can be assigned R for the purpose of processing Rules 4b. In this way, all discussion can be expressed in terms of "equal" or "not equal" to a reference R or S, rather than "like" vs. "unlike". It is critical that an implementation assign auxiliary descriptors involving double bonds --seqCis/seqTrans and M/P -to the sp 2 node closer to the root (or, alternatively, to both nodes equally). Otherwise the second phase of Rule 4b may fail. |
60c73e14337d6ca46ae26282 | 20 | The process for ranking ligands in Rule 4b is a more complex process than for previous rules, involving a two-stage process. First, the nodes are re-ranked in a way that may cross digraph branches. Second, the nodes are scanned in rank order for auxiliary descriptor similarity to both R and S reference descriptors. The higher priority ligand is the one with the highest score after both of these comparisons are made. We have found the easiest way to conceptualize this algorithmically is to "read" each of the four rankings (two for each ligand) as a series of 0s and 1s, which can be implemented as an integer, an array, or a bit set. In our examples, we will use integers, though our different implementations actually use different representations. The basic idea is to create four lists for each pair of ligands that can be compared together. If there is no winner, we go on to the next rule. |
60c73e14337d6ca46ae26282 | 21 | This criterion must incorporate the full path from the root to this node, including all Rule 4a-priorities as well as the similarity or dissimilarity of the node's descriptor to the reference. Pseudoasymmetric descriptors r, s, m, p, seqcis, and seqtrans are ignored in this process. As always, only previously identically-ranked nodes are resolved. |
60c73e14337d6ca46ae26282 | 22 | The example in Figure illustrates the process used in Rule 4b for VS262. Note that branches change order, depending upon the reference. In this case, some of the branch orderings have already been set, due to a Rule 4a comparison, r vs. ns. The reading of the S-ranked ligand A, SRSRRR (read from centers 1-6, in order), is encoded as 101000 in base-2, giving a value of 40. |
60c73e14337d6ca46ae26282 | 23 | Similar encoding gives 24 for S-ranked ligand B, 27 for R-ranked ligand A, and 43 for R-ranked ligand B. So ligand B, with a high score of 43, is given higher priority than ligand A, and the designator is R. (each encoded as the number 3 by one of the reference options) will be selected in preference to either RS or SR (each of which will be encoded as 2 by one of the options and 1 by the other). And yet, identical readings or opposite readings overallsuch as RRSRRS vs. SSRSSR, which need to pass on to Rule 5are not distinguished, as they will be encoded as the same two numbers with one or the other reference options. |
60c73e14337d6ca46ae26282 | 24 | An important facet of the ranking in Rule 4b is that identically-ranked nodes can come from different branches of a ligand. So, for example, in Figure we have a case (VS242), which has four "highest-ranked nodes" that must be sorted by R and S as a group. The result that the two ligands cannot be distinguished in Rule 4b. Note that all of the "rules for sorting" discussed in other works, such as, "The first step of this procedure is the critical choice of the first descriptor for each ligand," are unnecessary in terms of code implementation. These "critical choices," such as determining the descriptor associated with highest-ranking node or the descriptor that occurs the most in the set of highest-ranking nodes or sequentially evaluating both and S as references, simply fall out of the mathematics of the numerical rankings described above. Thus, no such critical choices need to be implemented, though if they are, they might speed the processing. |
60c73e14337d6ca46ae26282 | 25 | If a center passes Rule 4b undecided, it means that there are only three possibilities: (1) There is no ligand chirality; (2) two or more ligands have identical chirality descriptors; or (3) the two ligands each have sub-branches with opposite chirality. Rule 4c takes care of case ( ), where we assign r over s, and m over p. The implementation of Rule 4c is a straightforward extension of the implementation of Rule 4a. |
60c73e14337d6ca46ae26282 | 26 | Rule 5 does a final check for enantiomorphic ligands. Note that implementation of Rule 5 is not just a check of the lists generated using the procedure of Rule 4b, as priorities may have changed after application of Rule 4c. Consider the digraph in Figure , which shows the digraph of Figure , ranked by R-and S-reference for both Rule 4b and Rule 5. Here we see that after application of Rule 4c, the sorting is changed, and Ligand A has preference over Ligand B by Rule 5. |
60c73e14337d6ca46ae26282 | 27 | Figure . The same digraph as in Figure , for VS242, here also indicating the four pseudoasymmetric centers. In this case, sorting and listing of ligands gives different results after application of Rule 4c, which sorts both main ligand branches by r > s. The Rreference, Ligand A, with RSRSR… has higher priority than B, with RSRSS…. Due to the fact that both asymmeric ligands are their own enantiomorph, the final designation is R rather than r. |
60c73e14337d6ca46ae26282 | 28 | If all ligands are finally distinguished after application of Rule 5, an additional test should be done to count the number of pairs of enantiomorphic ligands. The final descriptor will be r/s, m/p, or, in the case of akenes, seqCis/seqTrans, if and only if this number is one (Figure ), otherwise it will be R/S, M/P, or seqcis/seqtrans (Z/E). In terms of implementation, the criterion for pseudoasymmetry at a tetrahedral center is that, when comparing otherwise identical ligands, there is an odd number of pairs that reverses priority when comparing like/unlike sequences using an S reference vs. using an R reference. So, for example, in Figure , the right-hand R ligand has higher priority with the R-reference, but the lefthand S ligand has higher priority with the S-reference. The priority switches, and we have the normal outcome for Rule 5pseudoasymmetric. But in Figure , priority switches twice, so the result is asymmetric. |
60c73e14337d6ca46ae26282 | 29 | In Figure , we have a different story. Sorting by R gives A > B. But sorting by S also gives A > Bno switch! A naïve application of Rule 5, only checking the two R-reference like/unlike lists, would have assigned r to that center. However, this center is asymmetric, not pseudoasymmetric, because each ligand is its own enantiomorph. The ultimate descriptor will be R, not r. The analysis is the opposite for alkenes and even-atom cumulenes (Figure and). Thus, if the pair on only one end of the alkene reverses priority when using the S reference vs. using the R reference (Figure ), then the result is the asymmetric seqCis or seqTrans; if neither or both pairs reverse priority (Figure ), then the result is the pseudoasymmetric seqcis or seqtrans. |
60c73e14337d6ca46ae26282 | 30 | A Proposal for Rule 6: Spiro and other axially-symmetric compounds. Early on in the development of the CIP system, it was recognized that certain cases involving C2, D2, and C3 point groups require additional consideration for assignment of stereodescriptors; an S4 case was described later 3 (Figure ). Specifically, an algorithm must distinguish between both enantiomers of compounds VS285, VS281, and VS283, and yet still deliver no stereodescriptors for cubane (VS009) and the S4 compound VS012. Only simple spiro structure VS285 is mentioned in BB 2013, in section P-93.5.3. |
60c73e14337d6ca46ae26282 | 31 | Cases to be considered by Rule 6 are identified by having, after application of Rule 5, two pairs of identical ligands or three or four identical ligands. Thus, in the first case in Figure , we have two identical amino ligands and two identical keto ligands; in the second and fifth cases, we have four identical ligands. In the third and fourth cases, we have three identical ligands. |
60c73e14337d6ca46ae26282 | 32 | Basically, by arbitrarily breaking the symmetry in this way, the problem is immediately resolved upon inspection of the digraph. We note for double spiran VS281, the result of application of the proposed Rule 6 generates the opposite descriptor to the one assigned previously. We believe this is a due to either a misassignment or a typographical error. (b) After application of Rule 6, all ligands are distinguished. The center receives a descriptor. Such will be the case only for compounds that have rings that involve the root atom and three or more ligands. A full application of Rule 6 tests all possible promotions, though this is necessary only for certain symmetries. Any matching R and S pairs are ignored; if a descriptor remains, it is valid. |
60c73e14337d6ca46ae26282 | 33 | Thus, a simple test that indicates fewer than three duplicate nodes with root distance 0 is sufficient for skipping Rule Note that Rule 6 must be applied in the determination of all auxiliary descriptors as well as in the final root-node determination, as for all other Sequence Rules, because it is possible for an auxiliary descriptor to result from C3 symmetry (e.g. VS300). Even if it is determined that Rule 4a -5 can be skipped due to the lack of auxiliary descriptors, Rule 6 should not be skipped without further evidence that Outcome (a) is the only possibility. |
60c73e14337d6ca46ae26282 | 34 | Rule 6 also allows assignment of stereodescriptors for centers with axial symmetry, such as those involving allenes and biphenyls. No further special consideration of high-symmetry compounds or groups is necessary. Interestingly, an easy extension of Rule 6 solves the heretofore unresolved issue of "in/out" stereochemistry such as shown in Figure . These compounds can be treated successfully simply by taking into account the temporary re-assignment of auxiliary descriptors after each promotion of a node. However, we stop short of recommending this modification at this time, as it only applies to rather esoteric structures mostly of only theoretical interest. Its implementation would result in additional (unnecessary) descriptors for common all-r "all-out" structures, many of which are known compounds and have already been arbitrarily assumed to have no stereodescriptors. |
60c73e14337d6ca46ae26282 | 35 | Key to our development process was the production of a robust validation suite that could be used by any developer to ensure that a CIP implementation follows the rules to whatever degree that software claims to implements them. Though scattered attempts have been made to develop such model collections, primarily for in-house testing of various software packages, the supplemental material to this paper includes the first fully tested openly available collection of models that can test the full range of issues presented in the CIP Sequence Rules. The 300 models are in annotated SDF format, providing both 2D and 3D models, as well as SMILES. Standard SDF data annotations provide reported (or corrected) descriptors for all relevant models in Chapter 9 of BB 2013, along with a significant number of additional "challenges" for developers. Accompanying this collection is a table that correlates the specific test models with specific sections of BB 2013 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.