id
stringlengths 24
24
| idx
int64 0
402
| paragraph
stringlengths 106
17.2k
|
---|---|---|
62348e88d6d3edc4ae960358
| 2 |
PLVPs were synthesized according to the approach shown in Scheme 1 by using free radical polymerization (FRP). The detailed description of synthesis can be found in the Experimental Section. To ensure that the prepared polymer is free of any chromophore, azobis(isobutyronitrile) (AIBN) was employed as the initiator. The obtained polymers were further purified by precipitation of its solution three times in hexane. After drying in a vacuum oven at 75 °C overnight, a white powder was obtained.
|
62348e88d6d3edc4ae960358
| 3 |
The molecular weight of the polymers was examined by using gel permeation chromatography (GPC) and the monomer ratio was characterized by NMR spectroscopy (Figures and). In the present work, PLVPs with MLA/2VP ratios of 5:1, 3:1, 1:1, 1:3, and 1:5 were prepared, termed as PLVP-1, PLVP-2, PLVP-3, PLVP-4, and PLVP-5, respectively (Table ). Also, PLVPs with similar monomer ratios (1:1) but different molecular weights were also prepared, i.e., PLVP-3L and PLVP-3M. The absence of vinyl protons in the 1 H NMR spectrum indicates that residual monomers were completely removed by precipitation (Figure ). Also, the intactness of cyclic lactide in the copolymers was confirmed by the presence of CH proton signals at 5.7-5.2 ppm. In addition, the chemical structures of PLVPs were confirmed by FTIR (Figure ). It is found that PLVPs can be dissolved in high polar solvents, for example, tetrahydrofuran (THF), dimethyl formamide (DMF), acetonitrile (ACN), and it is difficult to be dissolved in low polar solvents, such as dichloromethane (DCM). Scheme 1. Synthetic Route to PLVPs.
|
62348e88d6d3edc4ae960358
| 4 |
The dilute DMF solutions of PLVPs are colorless, and blue emission can be observed when irradiated under 365 nm UV light from a handheld UV lamp (Figures and). Also, the solid powders of PLVPs exhibit bright blue emission under UV illumination at ambient temperature (Figures and). To confirm the observed PL is not caused by any impurity, PLVPs and the corresponding homopolymers, i.e., PMLA and poly(2-vinylpyridine) (P2VP) prepared by FRP with AIBN as initiator were examined. As shown in Figure , both PMLA and P2VP show very weak PL in solution, which is in agreement with the previous reports. Next, to verify that the observed PL of PLVP is not due to any byproducts generated during polymerization, PMLA homopolymer was dissolved in 1,4-dioxane (DOX) and heated with 2VP monomer at the temperature used for polymerization (75 °C) overnight (termed as PMLA@2VP).
|
62348e88d6d3edc4ae960358
| 5 |
It is found that only a very weak PL can be observed in PMLA@2VP (Figure ). In addition, almost no significant PL signal is detected in the blend of PMLA and P2VP homopolymers in solution (w/w = 1/1). All the above results confirm that the emission of PLVP is not due to any luminescent impurity or byproduct. Also, it implies that the intrinsic PL of PLVP is attributed to the copolymerization of two monomers, and the covalent linkage of MLA and 2VP along the main chains is critical to the intrinsic emission of PLVP (see below).
|
62348e88d6d3edc4ae960358
| 6 |
Furthermore, it is found that the emission of PLVP-3 is almost independent of the solvent circumstances (Figure and Table ), implying that the emission is not attributed to the intermolecular charge transfer (ICT), nor the formation of polymer/solvent complex. Moreover, the lifetimes of emission bands at 350 nm were around 6 ns for PLVPs with various MLA/2VP ratios, suggesting that the emission of PLVP is fluorescence, rather than phosphorescence (Figure and Table ). The ΦF of PLVP-1 to PLVP-5 in solution are measured as 20.6%, 21.8%, 20.5%, 19.5% and 20.1%, respectively. As a non-conjugated polymer, PLVPs give a fairly high ΦF value in solution, which is comparable to some traditional luminophores. assigned to the 0-0 and 0-1 vibrionic transition, respectively. Figure shows the normalized absorption and PL spectra of PMLA, P2VP, and the blend of PMLA and P2VP (w/w = 1/1) in ACN solution. The absorption peaks in P2VP and the blend at around 260 nm are ascribed to the intrinsic absorption of pyridine rings. For PMLA, no characteristic absorption peak can be observed. As shown in Figure , normalized PL peaks of PLVPs exhibit a slight red-shift from 400 to 412 nm as the MLA/2VP ratios varied from 5:1 to 1:5. More importantly, it is found that the wavelength of PLVP emission is very close to that of P2VP, despite the emission of P2VP is very weak (Figure ). Therefore, it implies that the emission of PLVPs is attributed to the induced/enhanced emission of pyridines by copolymerizing with lactone rings. nm. In addition, the emission intensity reduced with the increased high concentration.
|
62348e88d6d3edc4ae960358
| 7 |
Moreover, PL of PLVP-3 in the mixed solvents of DMF/H2O with different H2O fractions at a concentration of 0.1 mg/mL is examined. It is found that, with the increasing of H2O fractions, the PL intensity gradually decreased (Figure ). These results indicate that PLVP possesses an ACQ property. This is different from most NCLPs reported previously, if not all, that usually possess AIE feature. Also, it's noteworthy that the emission wavelength of PVLP-3 in the dilute solutions (0.001-1.0 mg/ml) is independent with the concentration (Figure ), which implies the emission of PLVP is determined by the intra-chain interactions, rather than the inter-chain interactions (see more discussion below).
|
62348e88d6d3edc4ae960358
| 8 |
As shown in Figure , PLVP-3 solution with low concentration (0.1 mg/mL) can only a Abbreviation: λab = absorption wavelength, λex = excitation wavelength, λem = emission wavelength be excited in a narrow range, i.e., 330-360 nm. However, the concentrated solution of PLVP-3 exhibited obvious excitation-dependence on emission. For example, the emission of 5 mg/mL PLVP-3 solution red-shifted from 410 nm to 480 nm when the excitation wavelength was changed from 330 nm to 400 nm. As the concentration reached 50 mg/mL, the emission at 410 nm disappeared and the strongest emission at 480 nm can be obtained with the excitation wavelength of 390 nm (Figure ). Also, as the concentration of PLVP-3 increases, the UV absorption band edge red-shifted (Figure ). Similar results are observed in the other PLVPs (Figure ). These results suggest that new luminophores with smaller bandgaps could generate in PLVP solutions with increased concentration.
|
62348e88d6d3edc4ae960358
| 9 |
Next, the structures and photophysical properties of PLVPs in solid states are explored. Figure than those in the precipitated powders, revealing the presence of higher-order of chain packing. The emission wavelength of cast films of PLVPs is around 400 nm (Figure , S16, and S17), whereas the absorption peaks appear at 300-450 nm (Figure ).
|
62348e88d6d3edc4ae960358
| 10 |
However, the precipitated powders of different PLVPs exhibited distinct excitation wavelengths. The emission wavelength of PLVP-1, PLVP-2, and PLVP-3 is around 465 nm, whereas the emission wavelength of PLVP-4 and PLVP-5 is about 400 nm (Figure and S19). The above results reveal that, as the powders of PLVPs were prepared by fast precipitation from non-solvent, the photophysical properties are expected to be influenced by their meta-stable structures.
|
62348e88d6d3edc4ae960358
| 11 |
All above results indicate that PLVPs possess intrinsic emission in solution and solid state, but the PL behavior is different from NCLPs reported previously. For one thing, PLVPs possess ACQ property, rather than AIE property. For another thing, no apparent molecular-weight dependence can be observed for the emission wavelength and the ΦF of PLVPs (Table ). These results suggest that the intrinsic PL of PLVPs is
|
62348e88d6d3edc4ae960358
| 12 |
given by an underlying mechanism that is distinct from typical CTE. We speculate that the emission of PLVPs is attributed to the presence of MLA comonomers enhanced the luminescence of pyridine rings via intra-chain interactions and the improved chain rigidity of copolymers. In this case, a single PLVP chain acts as an individual luminophore, namely, single-chain luminogen. With the increased concentration of the solution, the aggregation of PLVP chains leads to the energy transfer via inter-chain interactions, resulting in the quenching and red-shifting of emission. Also, the aggregation of PLVP chains gives through-space conjugations (TSC)
|
62348e88d6d3edc4ae960358
| 13 |
among aromatic and non-aromatic moieties, resulting in distinct emissive species with varied bandgaps. To confirm our speculation, the intra-chain interactions and the glass transition of PVLPs are investigated by solid-state nuclear magnetic resonance (NMR) and differential scanning calorimetry (DSC). Figure shows the high-resolution doublequantum/single-quantum (DQ/SQ) chemical shift correlation of PLVP-3 and PMLA/P2VP blend (w/w = 1/1) at a magic-angle-spinning rate of 12 kHz. Note that both samples have a similar monomer ratio. It is observed that there are strong autocorrelations among protons of the pyridine chemical group (peak 1). Interestingly, the DQ correlation between pyridine (peak 1) and -OCH (peak 2) was not observed in both samples, indicating the distance between these two protons was larger than about 5 Å.
|
62348e88d6d3edc4ae960358
| 14 |
It also implies that the correlation between peak 1 and peak 3 mainly comes from the intra-monomer correlations. In addition, the DQ correlation peak at around 13.5 ppm (i.e., auto-correlation of pyridine) is much stronger than the DQ peak at around 3.4 ppm (i.e., auto-correlation of aliphatic protons) in PLVP-3, however, these two peaks have similar intensity in the blend. It is because the intra-chain interactions between two monomers in PLVP can lead to better dispersion of monomers and thus avoid the aggregation of aliphatic groups. The solid-state NMR results indicate that the copolymer, PLVP, possesses a stronger intra-chain interaction between lactones and pyridines compared with the blend of PMLA and P2VP homopolymers. As evidenced in Figure , no emission can be observed in PMLA/P2VP blend. It indicates that the effective intra-chain interactions of PLVPs are critical to the intrinsic PL in solution. In this case, a single polymer chain can play a role as an individual luminophore, rather than the clusters of certain groups. This is also the reason that strong emissions and high quantum yield can be obtained even in a dilute solution of PLVPs. Furthermore, theoretical calculations were carried out for PLVP based on a single polymer chain with four alternant units (Figure ). As a comparison, P2VP and PMLA homopolymers with four repeating units were also investigated (Figure ). It is found that the pyridine and lactone rings could have a small distance (3.227 Å) under certain conformation. Note that the C-C backbones of PMLA are atactic, and different local conformations are expected to coexist. The HOMO energy levels of PLVP-1~PLVP-5 are -5.86 eV, -5.94 eV, -5.94 eV, -5.95 eV and -5.87 eV, respectively, as measured by photoelectron spectrophotometer (Figure ). Based on the corresponding UV absorption band edge, the LUMO energy levels of PLVP-1~PLVP-5 are calculated as -2.38 eV, -2.48 eV, -2.48 eV, -2.37 eV and -2.29 eV, respectively. In addition, the HOMO energy levels of PLVP-3L and PLVP-3M are -5.85 eV and -5.77 eV, respectively, and the LUMO energy levels are -2.47 eV and -2.39 eV, respectively. To evaluate the dependence of PL intensity of PLVPs on chain rigidity, the Tgs of PLVPs with different MLA concentrations were examined by DSC. Figure shows the DSC heating curves of P2VP, PMLA, and PLVPs. The Tgs of P2VP PMLA are 107 and 235 ℃, respectively, which are consistent with the previous reports. All the PLVPs exhibit single Tg in their DSC profiles, indicating the random and homogeneous distribution of two monomers along the main chains. The Tgs of PLVP-2, PLVP-3, PLVP-4, and PLVP-5 are 97, 107, 140, and 178 ℃, respectively. The Tg of PLVP-1, the one with the highest concentration of MLA, was unable to be detected, since its Tg is above the degradation temperature (Figure ). Figure shows the PL spectra of PLVPs in DMF solution at the same concentration (1 mg/mL), and the emission intensity of PLVPs is varied with the monomer ratio. It is found that the maximum emission is obtained as the MLA/2VP ratio is 3:1 (Figure ). This is because the emission intensity is a "trade-off" between the chain rigidity and the concentration of luminophores. Namely, the low content of MLA will reduce the chain rigidity, and the low content of 2VP will give to a low concentration of luminophores. Similar results were also observed in some previous reports. In addition, it is found that PLVPs with different monomer ratios have almost the same quantum yield and fluorescence lifetime, but distinct fluorescence intensities in solution. This is because the absorbance of PLVPs is varied with the monomer ratios. According to the absorption spectra of PLVPs shown in Figure , the molar extinction coefficients of PLVP-1 to PLVP-5 are calculated as 4.2×10 3 , 5.8×10 3 , 4.5×10 3 , 1.4×10 3 , 1.0×10 3 cm 2 /mol, respectively. As a result, the fluorescence intensity of PLVP-2 is higher than the others due to its larger value of molar extinction coefficient.
|
62348e88d6d3edc4ae960358
| 15 |
The above results give the proof-of-concept for the single-chain luminogen of PLVPs and further understanding of its PL mechanism. As pyridine, as well as carbonyl groups, are well-known organic ligands, we speculate that the intra-chain interactions of PLVP will be interfered by introducing metal ions. Namely, the intra-chain interactions of PLVP may be broken by the presence of metal ions, which will result in the change of photophysical properties. To this end, the response of PLVP to a variety of metal ions is investigated. Figure shows the PL spectra of PLVP-3 in DMF (0.05 mM) with 20 mM metal ions. It is found that the presence of Co 2+ , Na + , and Zn 2+ could slightly enhance the PL of PLVPs, whereas the other ions, i.e., Ni 2+ , Eu 3+ , Ga 3+ , Cu 2+ , and Fe 3+ , would reduce the PL of PLVPs (Figure and). In particular, Fe 3+ significantly quenched the PL of PLVP-3. Generally, the quenching effect is determined by the metal ion charge/radius ratio. Namely, the stability of the metal complexes increases with increasing of the metal ion charge/radius ratio. As Fe 3+ has the largest charge/radius ratio than the other ions used in this work (Table ), it possesses stronger interaction with pyridines. Figure show the PL spectra of PLVP-3 in DMF with varied concentration of Fe 3+ , With the increase of Fe 3+ concentration, the fluorescence intensity of PLVP-3 gradually decreased. The UV-Vis absorption spectra of PLVP-3 with the addition of different metal ions (Figure ) shows a new absorption band at 300 -450 nm upon adding Fe 3+ , suggesting that Fe 3+ could coordinate with the carbonyl groups and pyridine, breaking the intra-chain interactions of PLVP, leading to the quenching of PL. Again, it evidences that the intra-chain interaction of PLVPs is critical to its intrinsic PL. Moreover, it is noteworthy that PL of PLVP is not very sensitive to the presence of Fe 3+ . This may be because the strong intra-chain interactions and the steric hindrance weaken the coordination abilities between ligands and Fe 3+ ions.
|
62348e88d6d3edc4ae960358
| 16 |
In summary, a new type of non-conjugated luminescent copolymer, PLVP, was synthesized by free radical copolymerization of two non-luminescent monomers, MLA and 2VP. In both solution and solid state, PLVPs exhibit intense blue emissions at 408 nm, and the quantum yield (ΦF) in the solution can reach 20%. As a non-conjugated luminescent polymer, however, the intrinsic PL of PLVP is attributed to the enhanced emission of pyridines by intra-chain interactions between pyridines and lactones. Such 'single-chain luminogens' of PLVPs exhibits ACQ properties. With increased solution concentration, the aggregation of polymer chains gives rise to inter-chain interactions, and new luminophores with smaller band gaps generate via TSC. Also, the emission of PLVPs can be tuned by molecular weight and ratio of monomer. The optimal molar ratio of MLA and 2VP is determined as 1:3, at which the maximum PL intensity can be obtained. This study not only expands the scope of luminescent polymers, but also provides a general approach to induced emission from non-luminescent monomers by copolymerization with proper comonomers.
|
62348e88d6d3edc4ae960358
| 17 |
Solution 1 H NMR spectra were measured on a Bruker ARX 400 NMR spectrometer using either DMSO-d6 and chloroform-d as solvent and tetramethylsilane (TMS, δ = 0) as an internal reference. Weight-average molecular weights (Mw) and polydispersities (Mw/Mn) of the polymers were estimated on a Waters gel permeation chromatography (GPC) system using THF as eluent. UV absorption spectra were taken on a Shimadzu UV3600PLUS array spectrophotometer. PL spectra were recorded on a Shimadzu RF6000 spectrofluorometer. The fluorescence quantum yields (ΦF) were measured by using a Hamamatsu C11347-12 Quantaurus-QY (C11347-12). The fluorescence lifetime results were taken from Hamamatsu Quantaurus-Tua (C11367-35). IR spectra were recorded on a PerkinElmer 16 PC FTIR spectrophotometer. WAXS measurements were executed by Rigaku Nanopix-SP with an ultrahigh-intensity microfocus rotating anode X-ray generator. The wavelength X-ray is 0.154 nm. The energy of HOMO was measured by using a photoelectron spectroscopy in air (PESA) with an AC-3 spectrometer (RIKEN KEIKI Co., Ltd.).
|
62348e88d6d3edc4ae960358
| 18 |
PLVP. In a 50 ml Schlenk bottle, add 0.5 g methylene lactide (MLA), 2 mg AIBN, 0.46 g 2-vinylpyridine and 10 ml 1.4-dioxane. The mixture in the bottle was frozen and deoxygenated three times, and the reaction bottle was placed in an oil bath at 75 °C to react for 24 hours. After the reaction is over, cool to room temperature, and add 10 ml C frequency of 399.78 MHz and 100.53 MHz, respectively. The magic-anglespinning (MAS) rate was set as 12 kHz, and POST-C7 pulse sequence was used to excite and reconvert DQ coherences with a DQ recoupling time of 0.167ms. Windowed z-rotation Phase-modulated Lee-Goldburg (LG) decoupling pulses were used during the t1 and t2 evolution periods to suppress proton dipolar couplings, leading to significant enhancement of proton spectral resolution. The experimental pulse sequences can be found in the literature. The 1 H 90 o pulse length was 2.2 s. 128 t1
|
66431a4221291e5d1d33f214
| 0 |
Gold nanoparticles (GNPs) conjugated to biomolecules are crucial for various applications in biotechnology, medicine, and diagnostics . The distinctive optical properties of gold nanoparticles, including their strong surface plasmon resonance and intense scattering and absorption properties, make them useful as contrast agents for various imaging techniques, including electron microscopy, optical microscopy, and lateral flow assays . The conjugation with antibodies, antigens, proteins, or nucleic acids enables to bind the GNPs to desired biomolecular targets minimizing interference from other components present in the sample. When conjugated with specific biological molecules, gold nanoparticles (GNPs) provide high sensitivity, and multiplexing capabilities in biosensing facilitating the detection of various targets through colorimetric, fluorescent, or surface-enhanced Raman scattering (SERS) detection methods . GNP-based sensor platforms with microfluidic devices result in the development of portable, point-of-care diagnostic tools applicable across healthcare, food safety, and environmental monitoring sectors . The necessary stage in incorporating GNPs into any sensing platform requires their modification with biomolecular ligands. However, the process of conjugation presents numerous obstacles. Preserving the functionality of biomolecules during conjugation is paramount to maintain assay performance and accuracy. The functionalization of GNPs with large biomolecules or multiple ligands may introduce steric hindrance effects, which could impact the accessibility of binding sites or the efficiency of target recognition. Therefore, optimizing the spacing and orientation of biomolecules on GNPs is crucial for minimizing steric hindrance effects and maximizing assay sensitivity .
|
66431a4221291e5d1d33f214
| 1 |
Copolymers derived from N,N-dimethylacrylamide (copoly-DMA) have emerged as highly versatile coatings suitable for a wide range of applications . Notably, they have found extensive use in the development of microarrays and multispot biosensors on various substrates, as supported by relevant references . This class of copolymers offers rapid and durable surface adhesion , facilitating the covalent attachment of biomolecules while exhibiting excellent antifouling properties . The key precursor of these copolymers, N-acryloyloxysuccinimide (NAS), features a functional moiety that readily reacts with functional groups conducive to click-chemistry reactions, such as azide/alkyne reactions . Furthermore, click-chemistry reactions facilitate the biorthogonal orientation of immobilized probes .
|
66431a4221291e5d1d33f214
| 2 |
Since the polymer is known to form a nanometric film on the surface of silica, we have utilized gold nanoparticles (GNPs) encapsulated within a silica shell through a sol-gel process. The silica shell provides stability, biocompatibility, and a versatile platform for further functionalization with copolymer of dimethylacrylamide. In this work, we demonstrate that dimethylacrylamide polymers can also be used to coat AuNP lacking a silica shell. Uncoated gold nanoparticles present a unique opportunity for modification with thiol-bearing reagents. However, the inherent challenge lies in the limited colloidal stability of AuNPs during the conjugation process, which may lead to aggregation. Aggregation propensity was observed also during the polymer coating stage. Indeed, the polymers employed in this investigation offer a significant advantage, thanks to their exceptional versatility. This adaptability allows for modifications to their composition, either during synthesis or through post-polymerization processes. To coat AuNPs without inducing aggregation, we introduced an ionizable monomer into the polymer backbone. This strategic modification not only ensures the stable coating of AuNPs but also creates opportunities for diverse functionalization, thereby enhancing their applicability across various fields. The polymer film bearing PEG-azide functionalities allows covalent binding of proteins and DNA modified with dibenzocyclooctyne (DBCO), a group characterized by its high reactivity towards azides via a copper-free strain-promoted alkyne-azide cycloaddition (SPAAC) reaction . This reaction proceeds rapidly and efficiently without the need for a copper catalyst, making it particularly attractive for bioconjugation and labeling applications in biological systems where copper could be cytotoxic.
|
66431a4221291e5d1d33f214
| 3 |
(Sunnyvale, CA, USA). Nanoparticle Tracking Analysis was performed with NanoSight NS300 using 3.2 Dev Build 3.2.16 software (Malvern Instruments Ltd, Malvern, United Kingdom). Antibody Aptamer Conjugates were purified using proFIRE instrument (Dynamic Biosensors GmbH, Munchen, Germany). Spectrophotometric characterizations were performed with Multiskan SkyHigh instrument from Thermo Scientific. Silica-Coated Gold Nanoparticles with an outer silica shell of 3 nm, were purchased from CD Bioparticles (Shirley, NY, USA). Gold nanoparticles of 40 nm of diameters were purchased from DCN (Carlsbad, CA, USA). Sample containing gold nanoparticles were sonicated using Omni Ruptor 250-Watt Ultrasonic Cell Disruptor (OMNI International, GA, USA).
|
66431a4221291e5d1d33f214
| 4 |
Images were created using Biorender (www.biorender.com). The Single Particle IRIS (SP-IRIS) images were acquired in every 2 minutes by scanning 20 µm with 1 µm step size using a conventional SP-IRIS setup . After every defocus stack was normalized by its median along the defocus dimension, the SP-IRIS signal was constructed by calculating the difference between maximum and minimum value of every stack.
|
66431a4221291e5d1d33f214
| 5 |
DNA sequences were modified as reported (amine and azide linkers were linked at 5' end, while DBCO and biotin were linked to 3' end or 5' end. Stabilizer DNA sequence was used without modification. Table The immobiline buffer was dried under reduced pressure and all the monomers were suspended in 0,5 mL of anhydrous THF. To the resulting reaction mixture was then added AIBN (5 mg) under argon atmosphere and the polymerization was conducted at 65°C for 2 hours.
|
66431a4221291e5d1d33f214
| 6 |
The reaction mixture was cooled at room temperature and 10 mL of anhydrous THF were added (polymer concentration 10% w/v). The polymer was then precipitated in 200 mL of petroleum ether and left stirring for 1.5 hours. The polymer was then filtered on a buchner and then dried under vacuum at room temperature.
|
66431a4221291e5d1d33f214
| 7 |
The polymer was synthesized by post-polymerization modification of copoly NAS positive 4%, similarly to what described in . In a two neck round bottom flask, copoly NAS positive 4% (0.5 g, 0.0483 mmol) was solubilized with 5 mL of anhydrous THF and degassed for 5 minutes by insufflating argon. 11-azido-3,6,9-trioxaundecan-1-amine (0.105 g, 0.483 mmol) was then added to the reaction mixture and the reaction proceeded at room temperature for 5 hours.
|
66431a4221291e5d1d33f214
| 8 |
60 nm silicon oxide SP-IRIS supports were pretreated with oxygen plasma to clean and activate the surface. The oxygen pressure was set to 1.2 bar with a power of 29.6 W for 10 min. Then chips were dipped into a 1% w/v aqueous solution of MCP-4 in 0.45 M ammonium sulfate. The supports were immersed into the coating solution for 30 min at room temperature, rinsed with bidistilled water, dried under nitrogen stream, and finally cured at 80 °C for 15 min.
|
66431a4221291e5d1d33f214
| 9 |
To prepare spotting solutions, oligonucleotides were diluted in a solution of 150 mM sodium phosphate buffer containing 0.01% sucrose monolaurate at pH 8.5. After spotting, chips were stored overnight in a sealed chamber filled at the bottom with sodium chloride saturated solution (40 g/100 mL, 65% humidity). Finally, chips were treated with a blocking solution containing ethanolamine (50 mM in 0.1 M Tris/HCl buffer pH 9 and 2 mM MgCl2) at room temperature for 1 h, rinsed with MgCl2 2 mM and dried.
|
66431a4221291e5d1d33f214
| 10 |
GNPs and Silica coated GNPs were analyzed using Nanosight NS300 (Malvern Panalytical, Malvern, UK). Videos were analyzed by the in-built NanoSight Software NTA 3.4 Dev Build 3.2.16. The Camera type, Camera level, and Detect Threshold were sCMOS, 9 and 5, respectively for 80 nm nanoparticles and 11 and 5 respectively for 40 nm nanoparticles. The number of completed tracks in NTA measurements was 5 (a 60 second movie was registered for each measurement). Samples were diluted in MQ water to a final volume of 1 mL. The ideal particle concentration was assessed by pre-testing the optimal particle per frame value (20-100 particles per frame).
|
66431a4221291e5d1d33f214
| 11 |
The mixture is incubated for 30 minutes at 25°C. After the incubation, 30 µL of Tris-HCl 1 M pH 8 were added, and the reaction was allowed to proceed for 5 minutes. The DBCOmodified streptavidin is then purified using Amicon Ultra 10MWCO centrifugal filters ( x 5 min at 13400 rpm) and finally PBS was added to bring the volume to 300 µL.
|
66431a4221291e5d1d33f214
| 12 |
Silica-coated gold nanoparticles were vortexed for 30 seconds and then sonicated using a bath sonicator for 10 minutes. Subsequently, a solution of Copoly Azide 4% at 1% w/v was prepared. SiGNPs were diluted to a final concentration of 1.4 OD by suspending 20 uL of SiGNPs in 480 uL the polymer solution; the sample was incubated for 1 hour at 25°C under stirring. At the end of the incubation, the sample was centrifuged for 5 minutes at 13,400 rpm, the supernatant was removed, and SiGNPs resuspended in 500 µL of MQ water.
|
66431a4221291e5d1d33f214
| 13 |
GNPs are diluted to a final concentration of 0,6 OD by suspending 250 uL of GNPs in 250 uL of a 2% w/v solution of copoly azide positive 4% in water; the sample was incubated for 30 minutes at 25°C under stirring. At the end of the incubation, the sample was centrifuged for 2 minutes at 6,720 x g, the supernatant was removed, and GNPs resuspended in 250 µL of MQ water. Centrifugation was repeated three times to wash GNPs. Finally, GNPs were redispersed in 250 µL of MQ water using an immersion sonicator.
|
66431a4221291e5d1d33f214
| 14 |
were centrifuged for 5 minutes at 12,000 x g, the supernatant was removed, and SiGNPs were resuspended in 200 µL of 10 µM DBCO-modified ssDNA in PBS and incubated overnight at 37°C under stirring. At the end of the incubation, the sample was centrifuged for 5 minutes at 12,000 x g, the supernatant was removed, and SiGNPs resuspended in 200 µL of MQ water. Centrifugation was repeated three times to wash GNPs. Finally, SiGNPs were redispersed using an immersion sonicator.
|
66431a4221291e5d1d33f214
| 15 |
To a 200 µL solution of 40 nm GNPs coated with copoly azide positive 4% (prepared as described in Section 2.8) 30 µL of 2 mg/mL DBCO-modified streptavidin, 70 µL of MQ water and 0.2 µL of Tween 20 were added and the obtained solution was incubated overnight at 25°C under stirring. After the incubation, the sample was centrifuged for 2 minutes at 10,000 x g, the supernatant was removed, and GNPs resuspended in 300 µL of 0.1X PBS + 0.05% Tween 20. Centrifugation was repeated three times to wash GNPs.
|
66431a4221291e5d1d33f214
| 16 |
The TEM images were recorded by ZEISS Libra 200 FE 200kV equipped with Omega filter in column. The samples were prepared by dropping the suspension on copper TEM grid and letting it dry in air. The diameter measurements were performed by the iTEM TEM Imaging Platform software (Olympus). During the last incubation, a single image was acquired every 2 minutes. Then, acquired images were processed and particles immobilized on spots counted using ImageJ software.
|
66431a4221291e5d1d33f214
| 17 |
. All size values diverge from those stipulated by the manufacturer. This discrepancy may arise because NTA measures the hydrodynamic radius in solution, which could encompass ions surrounding the gold particles. In all instances, the introduction of a polymeric coating prompts an increase in particle size as discerned by NTA. This implies that the soft coating induces an expansion in the hydrodynamic radii.
|
66431a4221291e5d1d33f214
| 18 |
Stability tests of gold nanoparticles are crucial to understand their behavior in various environments and applications. We conducted a stability test by exposing the SiGNPs (both uncoated and coated with copoly azide 4%) and GNPs (both uncoated and coated with copoly azide positive 4%) to a solution of HCl 0.1 M or NaOH 0.1 M. Solutions properties were analyzed acquiring absorbance spectra between 400 and 1000 nm.
|
66431a4221291e5d1d33f214
| 19 |
In a plethora of biological assay, biomolecules (including DNA, proteins and peptides) are immobilized on AuNPs to provide labeling of analytes within the assay. The choice of the optimal modification strategy strongly influences the overall performance of the entire assay itself in terms of specificity, sensitivity and nonspecific binding (thus impacting on the signal-to-noise ratio and limit of detection). In this work, 80 nm SiGNPs were coated with copoly azide 4%. The polymer contains azide groups that are capable of reacting with DBCO-modified biomolecules. Conjugation occurred through incubating the polymer coated SiGNPs with the biomolecule solution under appropriate conditions (pH, temperature, buffer), resulting in biomolecules binding to azide groups on the SiGNP surface via covalent bonding.
|
66431a4221291e5d1d33f214
| 20 |
Following this experimental scheme SiGNPs decorated with a ssDNA sequence (called Utag) or streptavidin were synthesized. Absorption spectra for functionalized SiGNPs are reported in Figure The performance of the so-obtained functionalized AuNPs was assessed using SP-IRIS prototype that is able to count individual nanoparticles bound to the surface of a microarray chip as described in Sections 2.15 and 2.16. Briefly, chips were functionalized with two sequences of DNA, namely Probe2 (the negative control) and Probe7. The chips were incubated with Utag-Tag7 that binds to Probe7. Then, SiGNPs binds to Utag-Tag7 directly (in the case of Utag-functionalized SiGNPs) or through Utag-Biotin sequence (when streptavidin-functionalized SiGNPs are used). Results (see Figure ) show that both SiGNPs (functionalized with either Utag or streptavidin) can bind selectively to positive spots (i.e. Probe7) with low nonspecific signals on negative spots. It can be also appreciated how streptavidin-functionalized SiGNPs perform better that Utag-functionalized ones (7-fold improvement in the signal). We hypothesize that this divergent performance stems from the necessity for high-affinity interactions to bind bulky objects such as 80 nm AuNPs. For this reason, streptavidin-functionalized SiGNPs are more likely to be immobilized on the surface of positive spots than ssDNAfunctionalized ones.
|
66431a4221291e5d1d33f214
| 21 |
In order to confirm that our immobilization strategy actually offers advantages in biological assays, we performed a negative control using 80 nm SiGNPs coated with copoly azide 4%, further functionalized with unmodified streptavidin. Lacking DBCOmodification, streptavidin can only adsorb on the surface of nanoparticles. We used these AuNPs with the same experimental procedure (see Supplementary material for details).
|
66431a4221291e5d1d33f214
| 22 |
The results underscore two distinct findings: firstly, the binding of SiGNPs on Probe7 spots is considerably lower when using coated SiGNPs where streptavidin is merely adsorbed onto the surface (see Figure and). This suggests that a lesser amount of streptavidin is immobilized on the surface via adsorption compared to covalent bonding.
|
66431a4221291e5d1d33f214
| 23 |
These findings confirm that copoly azide 4% coating effectively mitigates nonspecific protein interactions with SiGNPs even after prolonged incubation times with high protein concentrations. Secondly, AuNPs with adsorbed streptavidin exhibit a higher degree of adherence to regions of the microarray where they are not intended to bind, thereby augmenting the background signal of the assay (see Figure ). This demonstrates how the functionalization outlined in this study can enhance the performance of biological assays by a combination of signal enhancement and noise reduction.
|
66431a4221291e5d1d33f214
| 24 |
Gold nanoparticles exhibit remarkable versatility in size, ranging from 2 to 250 nm, allowing for precise control over their dimensions. Despite being derived from a material known for its costliness, GNPs are economically viable due to their stability over extended periods and their effective utilization at low concentrations. Overall, the stability of gold nanoparticles is essential for ensuring their performance, reliability, and safety across various applications. Stable nanoparticles have a longer shelf life and can be transported over long distances without significant changes in their properties. This is particularly important for commercial applications where consistent product quality is essential.
|
66431a4221291e5d1d33f214
| 25 |
Instability can lead to aggregation, degradation, or loss of functionality over time, affecting the reliability and reproducibility of experimental results . This aggregation process can occur due to electrostatic interactions. When functionalized with charged ligands or molecules, AuNPs may experience electrostatic repulsion or attraction depending on the charge of the functional groups . Another problem is non-specific binding to unintended targets or surfaces, leading to background signals and reduced assay specificity. Minimizing non-specific binding is crucial for improving assay sensitivity and accuracy, particularly in complex biological samples .
|
66431a4221291e5d1d33f214
| 26 |
Consequently, only a limited number of DNA molecules attach to the AuNPs through simple mixing, and they do so in an uncontrolled manner. This haphazard attachment prevents DNA molecules from effectively hybridizing with complementary DNA (cDNA) due to the strong interactions between DNA bases and AuNPs . Several approaches have been developed to overcome this problem, however, the procedures devised require time consuming stepwise addition of salt (around 10-50 mM each time), addition of surfactants, sonication, and freezing . Our work aimed at developing a fast, reliable, and scalable coating process capable of stabilizing and functionalizing AuNP in minutes.
|
66431a4221291e5d1d33f214
| 27 |
Various natural and synthetic polymers exhibit the capability to either physically adsorb or covalently bind to gold nanoparticles, depending on their molecular structure and functional groups. The polymers used in this work share a common backbone of polydimethylacrylamide (poly-DMA) which carries residues of propyl silane and polyethyleneoxide azide (see Figure ). The functional group responsible for forming covalent bonds with the silica layer coating the gold surface is the organosilane. We As a further confirmation, TEM analysis was performed on 40 nm SiGNPs (Figure ).
|
66431a4221291e5d1d33f214
| 28 |
Comparing the uncoated SiGNPs with the same particles after polymer coating and functionalization with polyT ssDNA, it is evident that the size and the shape of the NPs is not influenced, and the samples do not show any large aggregates, proving that the functionalization does not affect the SiGNPs stability. Moreover, the slightly different aggregation state of the polymer coated SiGNPs and the ssDNA-functionalized SiGNPs (Figure and 3c, respectively) suggests that the presence of the oligonucleotide improves the dispersion and the availability of the SiGNPs in solution.
|
60c7536eee301cfe6fc7add0
| 0 |
Detailed chemical kinetic modeling continues to gain interest as an approach to study reactive chemical systems, ranging in application from combustion and pyrolysis of fuels to degradation of active pharmaceutical ingredients. This growth can be attributed to a combination of demand for studying increasingly complex chemistries and supply of computational power and quantum chemistry capabilities. By taking advantage of these computational resources, automatic mechanism generation tools are able to systematically enumerate and evaluate potential chemical pathways, reducing the chance of human error. This is largely a data-driven task, requiring good estimation algorithms for thermochemical and rate parameters, which in turn rely on accurate training data from experiments or quantum chemistry calculations.
|
60c7536eee301cfe6fc7add0
| 1 |
The Reaction Mechanism Generator (RMG) project has been in development for over a decade, with the current Python version having begun development in 2008. RMG v1.0 was described in 2016. Here, we are excited to present RMG v3.0, which brings many new features including Python 3 compatibility, heterogeneous catalysis modeling, and new parameter estimation algorithms. With these and other improvements, the codebase has doubled to over 120,000 lines of Python code. Many developments have been focused on improving nitrogen, sulfur, and aromatic chemistry to better model combustion emissions and refining processes. RMG has recently been used successfully to model ethylamine pyrolysis, di-tert-butyl sulfide pyrolysis, hexylbenzene pyrolysis, effect of substituted phenols on ignition delay, PAH formation in methane oxidation, and catalytic combustion of methane. The structure and concept behind RMG has been described previously, 7 so only a brief overview will be given here. RMG is a tool for automatically constructing detailed chemical mechanisms which is largely comprised of three components:
|
60c7536eee301cfe6fc7add0
| 2 |
The latest release of RMG includes updates across all three components to expand modeling capabilities and improve accuracy, robustness, and performance. RMG uses a core/edge reaction model during mechanism generation, where the core contains species and reactions which have already been identified as being important, and the edge contains species and reactions which are under consideration. To reduce the model truncation error, in each iteration, RMG identifies one or more species to move from the edge to the core based on the species' total formation rate in a homogeneous batch reactor simulation. It then generates new reactions between the newly added species and other species in the core. The model is considered converged when no edge species exceeds the user-specified tolerance for selection.
|
60c7536eee301cfe6fc7add0
| 3 |
The transition for RMG included many steps. The first step was ensuring that Python 3 versions of all of our dependencies were available. This was straightforward for widely-used packages since all of them already supported Python 3. However, some packages developed specifically for RMG also had to be updated with Python 3 support, namely PyDAS and PyDQED. The second step of modifying RMG for Python 3 compatibility was facilitated by automatic tools like python-future, although substantial manual intervention was still required. In the final step, we used this opportunity to standardize function names throughout our API to comply with PEP-8 recommendations, effectively the official Python style guide. In total, transition tasks took approximately 500 developer hours to complete.
|
60c7536eee301cfe6fc7add0
| 4 |
Molecule representations have been extended to include catalyst sites, which are represented as a generic "X" element. New bond types have been implemented to represent the metal-adsorbate bond, including van der Waals bonds (internally represented with a bond order of 0) and quadruple bonds (e.g., for adsorption of a carbon atom). These extend the existing single, double, triple, and benzene bond orders.
|
60c7536eee301cfe6fc7add0
| 5 |
Thermochemistry estimation has been expanded to estimate parameters for surface species by applying adsorption corrections. For a given surface species, the metal is first removed to obtain an estimate for the gas-phase species using existing methods (e.g., group additivity or libraries), then an adsorption correction is determined from a group additivity tree and added to the gas-phase value. Thermochemistry libraries are also supported for surface species. The RMG database currently contains a thermochemistry library with 21 adsorbates on nickel and a more recent library that has 69 H/C/O/N-containing adsorbates on platinum. By default, RMG uses binding energies for Pt(111), but energies for an arbitrary catalyst can be specified in the input file (Figure ). Adsorption corrections are then scaled appropriately based on the specified binding energies. Surface simulations require use of the new SurfaceReactor class. This module performs the reactor simulations necessary for the flux-based algorithm for model growth. It is modeled as a zero-dimensional, isothermal, isochoric batch reactor which tracks surface coverage in addition to gas-phase mole fractions. User specification of surface area to volume ratio and surface site density are required. For surface mechanism generation jobs, RMG will output separate gas-and surface-phase Chemkin mechanism files along with a single Cantera mechanism file.
|
60c7536eee301cfe6fc7add0
| 6 |
A recent case study in methane catalytic combustion on platinum demonstrates the heterogeneous catalysis functionality. It addresses extensive updates to the original release of RMG-Cat. Among those is the new platinum thermochemistry database which is larger and more accurate than the original nickel database and has the advantage of including nitrogen-containing adsorbates. Another important new feature is the ability to explore heterogeneous and gas-phase reactions simultaneously, as with that the resulting microkinetic models can provide an implication of when catalytic surfaces lead to radical chemistry in the gas phase.
|
60c7536eee301cfe6fc7add0
| 7 |
Previously, local first-order sensitivity analysis was available to calculate sensitivities of species concentrations to thermochemistry and rate constants. New methods for both local and global uncertainty analysis have been implemented in RMG. Local uncertainty analysis builds on those first-order sensitivity by incorporating estimated uncertainties for thermochemical and rate parameters to obtain uncertainties for species concentrations. Global uncertainty analysis uses the MIT Uncertainty Quantification Library (MUQ 2) 20 to construct polynomial chaos expansions (PCEs) based on reactor simulations at random points within the uncertainty space of the input thermochemical and rate parameters. Reactor simulations are performed using Cantera. A key feature of the RMG uncertainty module is the ability to track correlated uncertainties in model input parameters, such as correlations arising from group additivity estimates for thermochemistry and rate rule estimates for rate coefficients. This can have significant effects on uncertainty propagation and the resulting uncertainties on output parameters.
|
60c7536eee301cfe6fc7add0
| 8 |
Uncertainty analysis can be requested via the RMG input file, which will lead to it being performed upon completion of model generation. Using uncertainty analysis does require that sensitivity analysis settings also be provided, since sensitivity analysis is required part of local uncertainty analysis. Local uncertainty analysis is also used to determine the parameters to vary for global uncertainty analysis, in order to minimize computational cost. For global analysis, PCE fitting can be controlled by specifying either a maximum runtime, error tolerance, or maximum number of model evaluations. These methods can also be applied to already-generated models via standalone scripts and interactive Jupyter notebooks, with the limitation that the same RMG version must be used for both model generation and analysis. These new tools can provide insights beyond first-order sensitivity analysis to aid in the model development process.
|
60c7536eee301cfe6fc7add0
| 9 |
In RMG, the reaction conditions of interest (i.e., temperature (T ), pressure (P ), initial composition (X 0 )) are provided by defining reactors in the input file. RMG supports three reactor types for mechanism generation, distinguished by the phases involved: SimpleReactor for gas phase, LiquidReactor for liquid phase, and the new SurfaceReactor.
|
60c7536eee301cfe6fc7add0
| 10 |
Because RMG uses a flux-based algorithm for identifying important species and reactions, the reactor conditions used to generate a model directly affect the conditions at which the model is applicable. Previously, the recommended approach for building a model applicable at a range of conditions was to define multiple reactors spanning the space of conditions of interest. For example, if the goal was to develop a model valid for temperatures from 1000 K to 2000 K and pressures from 1 bar to 10 bar, the user may need to define a dozen reactors with all combinations of T = {1000, 1200, 1400, 1600, 1800, 2000} K and P = {1, 10} bar. This can be bothersome to the user, and risks missing important chemistry which may occur in between the chosen points.
|
60c7536eee301cfe6fc7add0
| 11 |
Ranged reactors are a new feature in RMG v3.0 to simplify the task of specifying a range of initial conditions. With the new feature, ranges for T, P, and X 0 can be directly specified for a single reactor block. Internally, RMG will automatically select points within the space of conditions for each iteration, using a weighted stochastic grid sampling algorithm. On each iteration, a coarse grid with 20 points in each dimension is constructed, and the desirability of each point is evaluated based on the number of iterations since it was last chosen. The desirability values are normalized to form probabilities, and a random point is chosen using those probabilities. The algorithm then takes a random step from the chosen point, with a maximum distance of √ 2/2 times the distance between grid points. That point is then used for the simulation. The algorithm continues iterating through the grid points considering the probabilities described above. A simplified example of the algorithm for two dimensions is shown in Figure .
|
60c7536eee301cfe6fc7add0
| 12 |
RMG can now generate isotopically labeled reaction mechanisms via a post-processing algorithm. After a normal RMG job is completed, the isotopes module can generate all combinations of isotopically labeled species and reactions (Figure ). To obtain consistent thermodynamics, RMG modifies species' entropy based on changes to molecular symmetry and modifies kinetic Arrhenius factors based on reaction path degeneracy. Classical, mass-dependent kinetic isotope effects (KIE) are also available.
|
60c7536eee301cfe6fc7add0
| 13 |
One challenge with this approach is that isotopically-labeled mechanisms grow exponentially with the number of atoms that can be isotopically labeled. For example, a single asymmetric molecule with six carbons would be represented by 64 different species with various carbon atoms enriched. Despite the combinatorial complexity, this method is still very useful for generating detailed isotopic mechanisms, and has been shown to provide good agreement and insight into position-specific isotope analysis experiments. Currently, the algorithm is limited to generation of isotopic mechanisms for C, though the framework is easily extensible to other isotopes.
|
60c7536eee301cfe6fc7add0
| 14 |
Atom types in RMG are a set of atom descriptors that describe the local environment around an atom. They can accelerate graph isomorphism (by using specific types) and improve flexibility when defining reactions (by using more generic types). The set of available atom types has been revised and expanded to improve representation of heteroatoms. Particular focus has been placed on expanding atom type descriptors for the various bonding configurations of nitrogen and sulfur, for which the full list of updated atom types has been recently reported. New carbon and oxygen atom types for representing formal charges and varying numbers of lone pairs have been added, along with additional halogen atom types. For surface chemistry, atom types representing generic surface sites have been added, along with quadruple bonds for carbon and silicon. A list of these new atom types is shown in Table .
|
60c7536eee301cfe6fc7add0
| 15 |
Oxygen with one lone pair and two benzene bonds pathways are now recognized by RMG. To address the increase in computational requirements for handling additional resonance pathways and structures, as well as to identify the representative localized structures, a heuristic-based filtration algorithm will identify representative resonance structures on-the-fly. This approach was shown to correspond well to quantum calculations.
|
60c7536eee301cfe6fc7add0
| 16 |
Parameter estimation is possibly the most important step in mechanism generation, especially for flux-based algorithms like the one used in RMG. Because the criteria for selecting species to add into the model depends on the calculated reaction flux to those species, thermochemistry and rate constant predictions must not only be accurate for important species, but they must be reasonably correct for unimportant species, so that they can be properly neglected. The estimation algorithms rely on data which have been collected and stored in the RMG-database. This release of RMG includes both newly added data and new estimation algorithms.
|
60c7536eee301cfe6fc7add0
| 17 |
For thermochemistry estimation, RMG relies primarily on group additivity, where the thermochemistry for a molecule is derived from the sum of contributions from each heavy atom. However, a major limitation of standard group additivity is that only local fea-tures are captured; longer range effects such as steric interactions and ring strain must be treated separately.
|
60c7536eee301cfe6fc7add0
| 18 |
For steric interactions, RMG previously included a limited set of gauche (i.e. 1,4) and 1,5-interactions. In the current version, these corrections have been re-organized into cyclic and non-cyclic non-nearest-neighbor interactions, following the addition of a new set of group additivity values for ring substituents by Ince et al. Ring strain can substantially affect the thermochemistry of many cyclic and polycyclic species. A previous limitation of the group additivity algorithm was that ring strain corrections would only be applied if there was an exact match to the molecule. To address this, a new estimation algorithm was developed to provide an estimate for the ring strain of any molecule based on a heuristic algorithm which decomposes the molecule into mono-and bicyclic substructures. Furthermore, the group additivity estimator in RMG has been significantly expanded for sulfur compounds, with the addition of 200 new group values for various C/H/O/S groups. These values were fitted from a collection of thermochemical data derived from quantum chemistry calculations.
|
60c7536eee301cfe6fc7add0
| 19 |
Going beyond group additivity, RMG v3.0 also includes an updated neural network based thermochemistry estimator, developed using the chemprop package for molecular property prediction. Many molecular property prediction models are based on DFT data, including the previous version of the RMG thermochemistry estimator, because they are readily available in large databases or can be calculated with low computational cost. Since RMG strongly benefits from more accurate predictions, the new thermochemistry estimator was designed using a transfer learning approach that is able to learn accurate models from small high-quality data sets composed of experimental and coupled cluster calculations. As described in the chemprop publication, the deep-learning models use a message passing neural network (MPNN) to encode molecular graphs into fixed-length feature vectors which are passed through additional fully-connected neural network layers to make the thermochemistry predictions. Instead of using the featurization for atoms and bonds implemented by chemprop, we removed features that depend on resonance structure and added ring membership features, which we have shown to be beneficial. Two separate models were trained, one to predict enthalpies of formation and one to predict entropy and heat capacities simultaneously.
|
60c7536eee301cfe6fc7add0
| 20 |
New kinetics families have been implemented in RMG to allow automatic enumeration of new reaction pathways. All of the new families which have been added since RMG v1.0.0 are shown in Table . A complete list of all families can be found on GitHub. These new kinetics families include reactions involved in the propargyl recombination pathway to benzene formation, peroxide reactions relevant in liquid phase oxidation chemistry, surface reaction types for heterogeneous catalysis simulations, and a few other reactions types which have been found to be important for various systems. The solution which is being introduced in RMG v3.0 is the capability of automatically generating the tree. The new method uses machine learning approaches to automatically generate a decision tree based on the available training reactions. Starting with a generic reaction template, new groups are generated based on pre-defined types of extensions, e.g., adding an atom, adding a bond, specifying an element, etc. An optimal extension is chosen at each level of the tree by determining information gain based on the reduction in reaction rate variance. More details of the algorithm will be described in a separate publication.
|
60c7536eee301cfe6fc7add0
| 21 |
To address this challenge, a software package like RMG should make use of up-to-date computational hardware to improve its performance without sacrificing accuracy of the generated mechanisms. Computational hardware development, more specifically, new chip designs allow for the addition of several cores to a single processor. Furthermore, each core might allocate a number of threads that can execute parts of a software in parallel and therefore, reduce execution time.
|
60c7536eee301cfe6fc7add0
| 22 |
In the case of RMG, parallelization is challenging to implement since the core algorithm described in detail by Gao et al. is iterative in nature, i.e., tasks must be performed in order because they rely on the results or prior tasks. However, there are certain portions of the algorithm which are more amenable to parallelization. In RMG v3.0, parallelization has been completely revamped using the built-in multiprocessing module in Python, providing parallel processing support for reaction generation and quantum calculations for the QMTP (Quantum Mechanics for Thermochemical Properties) module.
|
60c7536eee301cfe6fc7add0
| 23 |
One task which can require substantial computing time in RMG is molecule comparison, which is done to identify if two molecules in RMG are the same chemical species. Part of the challenge is because RMG uses localized resonance structures to represent molecules, so simply comparing two structures may not be sufficient to determine whether or not they are the same. Instead, all of the resonance structures must be compared. Therefore, the standard approach to comparing molecules was to generate all resonance structures for the two species and comparing them to each other using graph isomorphism. To confirm that two molecules are the same, the comparison can return as soon as a matching pair of resonance structures is found. However, to confirm that two molecules are different, all combinations of resonance structures must be checked.
|
60c7536eee301cfe6fc7add0
| 24 |
The previous approach was very time-consuming, especially when considering resonance structure generation. A timing comparison of various methods for comparing molecules is shown in Figure for five test cases of comparing identical or different molecules. The first (blue) bar shows timing for resonance structure generation followed by graph isomorphism, and it's clear that the resonance structure generation task increases the total time by over an order-of-magnitude, even for species without resonance. A newly implemented isomorphism method, referred to here as "loose isomorphism," relies on ignoring electron-related features, such as radicals, lone pairs, and bond orders. Multiplicity is considered in order to distinguish electronic states. Charge is not yet considered, since RMG does not currently support ions. The purpose is to have an isomorphism approach which is independent of resonance structures and only focuses on the atom arrangement.
|
60c7536eee301cfe6fc7add0
| 25 |
This eliminates the need to generate resonance structures, and both positive and negative results can be determined by comparing a single pair of structures. This new approach can identify identical molecules more reliably than strict isomorphism because it avoids any limitations in the resonance generation algorithm. For example, prior to the implementation of an algorithm for benzyne resonance, loose isomorphism could correctly identify the two resonance structures of benzene as being the same molecule while strict isomorphism could not. The timing for this method is shown by the second (orange) bar in Figure . We see that this method is almost identical in performance to normal isomorphism, with the main difference being faster identification of different molecules with many resonance structures, as demonstrated by Test 5. Additionally, there is a guaranteed performance improvement because resonance structure generation is avoided.
|
60c7536eee301cfe6fc7add0
| 26 |
A third option which also has significant potential is to compare the International Chemical Identifier (InChI) for various molecules. An InChI is a string identifier for a molecule which is designed to be independent of resonance structures and would therefore give the same result as the loose isomorphism method, although multiplicity would still need to be considered separately because InChI does not account for electronic states. Additionally, string comparison is extremely fast. Unfortunately, InChI generation time is non-trivial.
|
60c7536eee301cfe6fc7add0
| 27 |
It is important to note that these timings are not completely representative of actual operation. Importantly, resonance structures and InChI strings can be cached, such that they only need to be generated once. Then subsequent comparisons would require much less time. However, a large portion of comparisons in RMG are with newly generated molecules, where the data would always need to be generated. As a result, the true cost of these comparisons would be in between the total time and just the comparison time.
|
60c7536eee301cfe6fc7add0
| 28 |
With continued growth of the RMG development team and user-base, good software development practices have become increasingly important. In recent years, additional emphasis has been placed on implementing best-practices for open-source software development. All RMG source code is publicly available on GitHub. Code review and continuous integration testing are emphasized as part of the development workflow, which has been formalized via official contributor guidelines. Elements of git-flow and semantic versioning have also been implemented into the development workflow to improve version release planning.
|
60c7536eee301cfe6fc7add0
| 29 |
RMG v3.0 is now available, and we recommend existing users to update their installations to take advantage of new features. Linux and MacOS are supported natively, and Windows is supported via the Windows Subsystem for Linux (WSL). Compared to RMG v1.0.0, there are many new features and substantial improvements across all aspects of the software.
|
60c7536eee301cfe6fc7add0
| 30 |
Python 3 support ensures that RMG is up to date with the latest scientific packages and will be for the foreseeable future. New chemistry features like surface mechanism generation and isotopic mechanism generation enable application of RMG to more systems than ever before. Uncertainty analysis provides new ways to analyze models to quantify the overall uncertainty in a model and identify the parameters which contribute most to that uncertainty.
|
60c7536eee301cfe6fc7add0
| 31 |
Fundamental improvements to molecular representation in the form of new atom types and resonance transformations work together to improve the the accuracy of the localized molecular representations. Parameter estimation, as the key to generating good models, has been improved via expansion of the database as well as addition of new algorithms like the neural network thermochemistry estimator. Finally, performance improvement is always an ongoing focus, and the recent implementation of parallel computing and improved molecule isomorphism comparison are steps towards faster model generation.
|
60c7536eee301cfe6fc7add0
| 32 |
All of the developments mentioned here, and countless others which can be explored in the detailed RMG release notes, 44 have greatly improved the accuracy, robustness, and applicability of RMG to modeling various chemical systems. RMG development is more active than it has been at any point in the past, which promises to continue bringing new and exciting improvements. For example, ongoing development of automated high-throughput quantum calculations for both thermochemistry and kinetics, leading-edge machine learning methods for parameter prediction, and new model expansion algorithms to complement species selection by flux are leading toward construction of even more accurate models using automatic mechanism generation.
|
677ec32dfa469535b936c4d7
| 0 |
buffer, achieving mean absolute error (MAE) and root mean squared error (RMSE) values within computational accuracy. Analysis of key atomic descriptors and qualitative trends in our dataset informed the design of novel NCB candidates we propose with optimized stability. This work enables researchers to predict the relative stability of NCBs before synthesis, thereby streamlining the process to make CFB more affordable and viable at industry scales.
|
677ec32dfa469535b936c4d7
| 1 |
With rising demand for stereo-, chemo-, and regioselective chiral molecule synthesis, interest in biocatalysis has also grown. In particular, cell-free biosynthesis (CFB), the in vitro use of enzymes in chemical synthesis, has gained popularity in recent years due to its higher yields and specificity and lower environmental impact than current synthetic techniques. This method also offers improved yields and resistance to toxicity relative to in vivo biosynthesis by removing cellular metabolic requirements and burdens. However, the cost of CFB is still a significant drawback, with enzyme cofactors representing a significant fraction of the reaction setup price tag.
|
677ec32dfa469535b936c4d7
| 2 |
Approximately 50% of known enzyme reactions require cofactors, the most popular being nicotinamide cofactors, which can be prohibitively expensive and unstable outside of cells. Pathways can be designed to recycle these redox cofactors, and sacrificial substrates can be included where this is not possible, allowing for their inclusion in catalytic rather than stoichiometric quantities, but this is not always enough to offset costs and make enzymatic syntheses economical. As a result, the development and implementation of synthetic nicotinamide cofactor biomimetics (NCBs) is an active area of research that aims to facilitate CFB in industrial settings and at scale.
|
677ec32dfa469535b936c4d7
| 3 |
NCBs are small molecules designed to mimic the redox role of natural nicotinamide cofactors, while providing advantages such as ease and low cost of manufacturing, and tunable enzyme specificity to enable bioorthogonal redox cascades. Simple NCBs contain a central pyridine ring to maintain the hydride transfer ability of their natural counterparts (Figure ), and various ring substituents, especially at the nitrogen position, are introduced to tune chemical properties.
|
677ec32dfa469535b936c4d7
| 4 |
Figure displays sixteen well-characterized cofactor mimics found in the literature, three of which contain variations in the R3 substituent such that they are not nicotinamide structures, however, they are still classified as NCBs for their ability to replace NAD(P)H. Many enereductases and some oxidoreductase enzymes have been shown to readily accept NCBs as redox cofactors. Other types of enzymes do not always naturally accept NCBs, either due to poor binding or incompatible reduction potentials, but activity can be engineered for these synthetic cofactors. Figure . A) Reduction of NAD + . B) Selected NCBs previously seen in literature. The 1,4dihydropyridine in both NAD(P)H and NCBs, shown in teal, is featured as the one constant design component throughout this work.
|
677ec32dfa469535b936c4d7
| 5 |
The use of NCBs makes CFB more industrially feasible due to their reduced cost compared to the natural nicotinamide cofactors. However, many NCBs tested to date are also at risk of decomposing at the nicotinamide moiety in buffers used in CFB. The proposed mechanism of nicotinamide decomposition by Alivisatos et al. is through addition across the 5,6-double bond in the 1,4-dihydropyridine ring, depicted in Scheme S1. This decomposition occurs most quickly in a potassium phosphate buffer, which often yields high enzyme activity. To further promote the applicability of NCBs in CFB, we propose their stability should be included as a feature to optimize during engineering campaigns.
|
677ec32dfa469535b936c4d7
| 6 |
Another opportunity for improvement in the NCB field is expansion of the chemical space. Within the realm of simple NCBs, described by Black et al. as structures which do not contain nucleotide components, there are vast possibilities for the R1 and R2 substituents (N-substituents and C3substituents, respectively). However, only a small number of NCBs have been synthesized and characterized in vitro. Tan et al. investigated chemical and physical properties of NCBs with a methyl R3 substituent (C5-substituent), pointing out the lack of research into modifications at the R3 position. They proposed adding an R3 methyl substituent to only two NCB structures of the eleven they report. Additionally, R2 variants have only been examined for NCBs with a benzyl R1 substituent. In our literature review, we found sixteen simple synthetic NCBs out of which eleven R1, four R2, and two R3 substituents are explored (Figure ). In this work, we expand the chemical space of NCB candidates through a combinatorial generation of simple NCBs from the previously evaluated substituents, resulting in a library of 132 NCB candidates (including the sixteen previously proposed) (Figure
|
677ec32dfa469535b936c4d7
| 7 |
Conformational sampling was performed using CREST version 3.0 with the GFN-FF force field and the GFN2-xTB semi-empirical level of theory with xTB version 6.7.0 (Figure Step II). SMILES (Simplified Molecular Input Line Entry System) text representations of each NCB can be found at and in the supplemental information and were converted to xyz-coordinate files using OpenBabel version 3.1.1. For structures in our library, CENSO version 1.2.0 31 was used to further refine the ensemble before clustering into 10 or fewer representative structures. Transition state conformers were generated using constrained CREST conformational sampling without CENSO to refine the ensembles. More details can be found in section 4 of the supplemental information. Density functional theory (DFT) calculations were performed using Gaussian 16, Revision C.01 (Figure Step III). All geometry optimizations used PBE0-D3(BJ)/6-31+G(d)/SMD(water) as the level of theory. Vibrational frequency analyses at the geometry optimization level of theory confirmed the nature of transition structures and ground state structures based on the presence of only one imaginary normal mode or none at all, respectively. Intrinsic Reaction Coordinate (IRC) calculations also used this level of theory to ensure the nature of transition structures during the mechanistic study. Additional single-point energy corrections with larger basis sets were used to obtain more accurate energies. For the calculation of all barrier heights, we used ωB97M-V/def2-TZVP/SMD(water) to obtain these energies. Calculations which used the ωB97M-V functional were done using Orca version 5.0.3 with support from libXC version 5.1.0. Energy corrections for structures in our library and the newly-designed NCBs were performed with PBE0-D3(BJ)/def2-TZVP/SMD(water). Free energies were calculated and potential energy surfaces constructed using Goodvibes version 3.2. Natural bonding orbital (NBO) program version 7.0.5 was used to obtain atomic partial charges.
|
677ec32dfa469535b936c4d7
| 8 |
NBO partial charges at ten key atoms were gathered from the ground state structure of each reduced cofactor. Additionally, nucleophilic and electrophilic condensed Fukui indices were calculated using NBO partial charges for each after the subtraction and addition of an electron, respectively (Figure , Step IV). Predictive modeling was done using ROBERT software version 1.0.6 (Figure , Step V). Our final model was trained using DFT-level descriptors after testing numerous model architectures (Figure ) before settling on the highest-performing one, a multivariate linear regression model with a 101:18:13 train:validation:test split. Default settings were used except permutation feature importance analysis (PFI). PFI scores were calculated in ROBERT and tabulated, then used to isolate the 8 most important descriptors to train the model. More details about the use of ROBERT software for predictive modeling can be found in section 7 of the supplemental information.
|
677ec32dfa469535b936c4d7
| 9 |
To expand the limited space of NCB candidates, we assembled a library of 132 NCBs, over 85% of which have not (to our knowledge) been previously evaluated. Each molecule contains the redox active 1,4-dihydropyridine ring and is uniquely identified by the combination of R1, R2, and R3 substituents (Figure ). Each substituent possibility was combinatorically added to the core 1,4dihydropyridine moiety. Substituents were selected based on previously evaluated NCBs to improve the likelihood that the resultant molecules will maintain some semblance of those that have been experimentally validated. However, we also included two additional R2 substituents which have not appeared in the literature to help examine substituent effects within the library, 4 and 6 in Figure . These substituents were added due to their electron-withdrawing capabilities, which we hypothesize will improve cofactor stability and allow further examination of the electronic effect of R2 substituents on stability while maintaining chemical similarity to precedented mimics. To refer to specific molecules we will adopt a three-character naming convention detailed in Figure , where the first, second and third characters correspond to the R1 (A-K), R2 (1-6), and R3 (a, b) substituent identities, respectively. For example, all NCBs with an octyl substituent at the R1 position are named D**, where the asterisks represent wildcard characters that could be any R2 and R3 substituent, respectively. Similarly, all NCBs which have a methyl group as the R3 substituent are named **b. In addition to the reduced form of each NCB, our automated workflow also generates the oxidized and decomposed species to study substituent effects on stability or redox activity. Generation of this expansive library was automated using an in-house developed Python script that is publicly available at and can be openly used to further expand this library with different substituents. More details about the generation of our NCB library can be found in section 1 the supplemental information. After SMILES (text-based) representations of each NCB in our library were generated, we followed an in-house automated protocol to perform conformational sampling and obtain optimized geometries at the DFT level of theory. This in-house script can be found at . Our conformational sampling procedure, using CREST to generate conformers and CENSO then CREST clustering to refine the ensemble, was used to isolate ten conformers of each NCB which were used for DFT optimizations and the curation of electronic-structure derived descriptors. A description of all calculated descriptors can be found in section 6 of the supplemental information. Principal component analysis (PCA) was applied to visualize the variations within our dataset of NCBs featurized by calculated atomic descriptors (Figure ). The first two principal components account for 71.6% of the total variance in the data, providing a sufficient representation of the diversity in our NCB library. Most of the experimental structures occupy a localized region of the chemical space, alluding to an improvement in molecular diversity. This robust library of structures allows more research into various substituent effects, opening the door for tuning specific properties of NCBs, such as reduction potential and stability in buffer.
|
677ec32dfa469535b936c4d7
| 10 |
Having expanded NCB candidate space, we aimed to understand the stability in terms of the NCB decomposition mechanism. NCBs which have been studied experimentally are noted to have high decomposition rates in phosphate buffer, which is problematic for CFB applications where enzymes often have highest activity in phosphate buffers. Thus, we modeled the decomposition mechanism of NCBs using a negatively charged phosphate molecule to approximate the buffer solution. Referencing a previously-proposed decomposition scheme, we modeled the decomposition of four NCBs which have experimental decomposition rates in phosphate buffer: MNAH (A1a), BNAH (F1a), P2NAH (G1a), and P3NAH (H1a). The activation barriers obtained are consistent with a process that is kinetically feasible at ambient temperatures.
|
677ec32dfa469535b936c4d7
| 11 |
DFT modeling of the decomposition pathway suggests that the process occurs in a stepwise manner (Figure ). The first step involved C5 protonation via dihydrogen phosphate to form Int I, followed by attack of hydrogen phosphate at C6 to form the functionally inactive product. A concerted mechanism could also be envisioned; however, it is not supported at this level of theory (see section 3 of the supplemental information for discussion). In each evaluated NCB, TS I corresponds to the highest energy barrier. As a result, these NCBs follow the stability trend of F1a = G1a > A1a > H1a, which is comparable to the experimental study that found the stability in phosphate buffer of F1a and G1a to be equivalent and higher than that of A1a and H1a. The species product labeled on the reaction scheme in Figure does not necessarily represent the final decomposed species, and has been known to further decompose into another final product. However, because experimental rates of decomposition are measured by the disappearance of the 1,4-dihydropyridine moiety and our calculated values are consistent with experiment, we do not consider further decomposition. BNAH (F1a), P2NAH (G1a), and P3NAH (H1a). Calculations were performed with the level of theory ωB97M-V/def2-TZVP/SMD(water)//PBE0-D3(BJ)/6-31+G(d)/SMD(water). After identifying TS I as a predictor of NCB stability, we sought a simpler stability representation that would be more suitable for high-throughput calculations. Optimization of transition state structures often requires more computationally intensive protocols and manual intervention.
|
677ec32dfa469535b936c4d7
| 12 |
was an appropriate alternative. Calculating ∆G requires only the optimization of minima structures and is therefore computationally cheaper. Such an approach is well grounded in studies of catalytic reactivity, where the thermodynamics of an elementary step is used to quantitatively assess the kinetic feasibility, formalized by the Evans-Polanyi principle and linear (free) energy relationships (LFERs). To ensure that this assumption holds, we optimized transition state structures for a subset of 34
|
677ec32dfa469535b936c4d7
| 13 |
Though quantitative decomposition rates were not available, we achieved qualitative accuracy in ranking the stabilities of B1a, B1b, D1a, F1a, and F1b from Tan et al. with 100 mM NCB concentrations. Due to high correlation between ∆G and ∆G ‡ , as well as qualitative agreement 121.2 kcal/mol (-591.0 to -507.2 kJ/mol). The majority of our library shows increased stability when compared with a common mononucleotide NCB, nicotinamide mononucleotide (NMNH -), which has a stability of -134.1 kcal/mol (-560.9 kJ/mol). These data suggest that we will be able to design NCBs with higher stability than those currently used in CFB.
|
677ec32dfa469535b936c4d7
| 14 |
To train a machine learning model to predict NCB stability in buffer, we had to identify descriptors to effectively describe the NCBs in ways relevant to decomposition. Because decomposition in phosphate buffer occurs within the conserved, redox active 1,4-dihydropyridine ring, this motif was our focus while describing NCBs. DFT-level descriptors were calculated at each atomic position within the 1,4-dihydropyridine ring (N, C2, C3, C4, C5, and C6), as well at the first atom of each substituent (R1, R2, & R3) and the descriptor values of both hydrogens at the C4 position (C4H) were averaged for a total of ten descriptors from eleven atomic loci. Calculated descriptors include partial atomic charges derived from natural population analysis (NPA), as well as the electrophilic and nucleophilic condensed Fukui indices (f(+) and f(-), respectively). Further explanation of descriptors can be found in the supplemental information. Upon examination, these descriptors are able to effectively capture different substituent identities at each position, so no other features were considered in this study (Figure ).
|
677ec32dfa469535b936c4d7
| 15 |
To limit computational cost as we moved forward, descriptors were calculated only for the reduced species of each NCB. Using a single species for descriptor calculation also reduces the computational cost of making future predictions with out-of-sample NCBs. We selected the reduced cofactor species for descriptor generation due to its role as the "universal intermediate"
|
677ec32dfa469535b936c4d7
| 16 |
for both decomposition and catalytic reduction, allowing this set of descriptors to be used in a future multi-objective optimization for both stability and reduction potential. Permutation feature importance (PFI) analysis was implemented to isolate and use only the most important descriptors for NCB stability to help prevent model overfitting. PFI begins with a model trained on all descriptors, then systematically permutates one descriptor at a time to determine the impact of that descriptor on the overall model. The worse a model performs when a descriptor is permutated, the more important that descriptor is to making predictions. This analysis is automated in ROBERT, resulting in a PFI score for each descriptor (Table ). The eight descriptors with the highest PFI scores were used in our model, including atomic charges at N1, C2, C3, C5, and C4H and electrophilic condensed Fukui indices at C3, R2, and C4H. We also performed a comparison of these properties calculated from the lowest energy conformer, which was used in model training, with the Boltzmann weighted properties for our representative subset of 34 NCBs. The maximum differences between values for all atomic charges and electrophilic condensed Fukui indices of these NCBs were 0.001 and 0.008, respectively. The differences are low relative to the magnitude of the values, suggesting we do not need to include the full conformational ensemble for our descriptors.
|
677ec32dfa469535b936c4d7
| 17 |
An array of four model types and seven train:validation:test splits were used to train possible DFTlevel models for predicting stability using ROBERT (Figure ). We selected the highestperforming model architecture after default PFI analysis based on the root mean squared error (RMSE) values of the validation set, a multivariate linear regression (MLR) employing a 101:18:13 train:validation:test split (Figure ). PFI scores were then further analyzed to limit model training to only the eight most important physical properties of the NCBs. This number of descriptors was selected because it was the lower limit at which we still had high model performance (Table ). More details regarding model training can be found in section 7 of the supplemental information. This analysis led to a high-performing model, shown by a high correlation (test set R 2 = 0.98), as well as mean absolute error (MAE) and RMSE within DFT accuracy levels (≤ 3 kcal/mol, 12.552 kJ/mol). descriptors. In addition to the standard model assessment performed in ROBERT, we also tested our model performance with a Spearman correlation, which a gave rank-order correlation of 0.974, demonstrating the ability of our model to correctly order NCBs with respect to their relative stabilities (Table ). This feature is ideal for an optimization protocol, where only the most stable NCBs are considered for experimental validation, so getting the correct stability ranking is as effective as predicting the actual value. We also explored a model trained with semi-empirical descriptors in an attempt to reduce the computational cost of making predictions, but the difference in performance relative to the cost of each method was not enough to favor the semi-empirical model (see section 7 of the supplemental information).
|
677ec32dfa469535b936c4d7
| 18 |
After training a regression model to predict NCB stability, we used that information to design NCBs with substituent groups outside the scope of those previously identified. To do this, we further examined the descriptors which were used in training the DFT-level model by examining the relationship between each descriptor and the NCB stability. In Figure , we show a subset of these relationships and label the data according to substituent positions to highlight how certain descriptors can discriminate by substituent identity. Examining the relationship of f(-) at the C4H atoms with stability (Figure ), there are two substituents which yield higher stabilities, corresponding to R1 substituents E and J. Each of these substituents have an aryl substituent bound directly to the N atom in the 1,4-dihydropyridine ring. Furthermore, the two most stable NCBs in our library are E4b and J4b, suggesting that aryl R1 substituents are a favorable design component to enhance NCB stability (Table ). This suggests that alterations to an aryl R1 substituent could be a viable path for further NCB chemical space exploration. Other substituent identities also impact stability, but not to the extent of that seen from R1 aryl substituents. We performed a similar analysis of the R2 substituent (Figure ), showing that the aldehyde moiety (4) tends to have higher stability than the other functional groups represented in our library at this position. The five most stable NCB structures in our library, and fifteen of the twenty most stable, have 4 as the R2 substituent. We hypothesize that this trend is due to the highly electronwithdrawing nature of aldehydes. Though 4 is not a substituent previously evaluated in literature, we added it to our library as an additional substituent to determine if the electron-withdrawing characteristics made a large impact on NCB stability. Based on these data, this is a key design feature in optimizing stability in NCBs. The next most stable substituent, 2, is also highly electronwithdrawing as a nitrile, which has been evaluated previously.
|
677ec32dfa469535b936c4d7
| 19 |
Finally, it is clear when looking at the relationship between C4H partial atomic charge and stability colored by R3 substituent (Figure ) that NCBs with an R3 methyl substituent have higher stability than those with only hydrogen at that position. While examining NCBs with similar C4H partial charges, those with b at R3 demonstrate greater stability. Furthermore, our eleven most stable NCBs have a methyl substituent at the R3 position. Tan et al. hypothesized that this improved stability is due to kinetic effects, and we are still able to replicate this trend even though our analysis is strictly thermodynamic. This observation could be caused by additional steric effects, such as lower flexibility in the 1,4-dihydropyridine ring with greater congestion at the C5 carbon with a methyl substituent rather than a hydrogen.
|
677ec32dfa469535b936c4d7
| 20 |
(1) and R3 hydrogen (a) with a highly electron-withdrawing R1 substituent (R1sub-VI). We wanted to design this NCB to look more similar to current structures to ensure the calculated properties are reasonable because the majority of structures in literature include 1 as the R2 substituent and a and the R3 substituent. In addition to altering the R1 substituent, we also looked at how changing the R2 and R3 substituents impacted stability. To do this, we designed four structures with new R2 substituents. These moieties included a thioaldehyde (R2sub-I), thioketone (R2sub-II), trifluoromethyl (R2sub-III), and nitro (R2sub-IV). Thioamides have literature precedent as dinucleotide biomimetic cofactors, which share most of their structure with NAD(P)H, and are proposed to behave similarly to their amide counterparts but have higher reduction potentials than the natural cofactors. We selected a thioketone rather than a thioamide because ketones demonstrate higher stability than amides in our library. Additionally, the increased stabilities shown by R2 aldehydes led us to also choose a thioaldehyde due to the increased stability shown by aldehydes in our study. Electron-withdrawing groups seem to improve stability at the R2 position, with aldehydes performing so well, so we selected a highly electron-withdrawing substituent, trifluoromethyl, to further investigate this trend. This reasoning also led us to design an NCB with a nitro R2 substituent. Additionally, we designed two structures with a trifluoromethyl R3 substituent (R3sub-I & R3sub-II). We selected this moiety as a possible R3 substituent to test if electron-withdrawing effects at this position would have any influence on stability in addition to the bulkiness of the substituent. Trifluoromethyl substituents have only a slightly larger size than the previously tested methyl substituent but are more electron-withdrawing.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.