Spectra Of Mixtures Research Articles

Thermodynamic and spectroscopic analyses of N-Methyl benzylamine-isomeric amyl alcohol binary mixtures: Jouyban-Acree model correlation and experimental validation

ABSTRACT Excess molar volume, excess isentropic compressibility, deviation in viscosity, and excess Gibbs free energy of activation for viscous flow for binary mixtures of N-Methyl benzyl amine (NMBA) with 2-Methyl-2-butanol (2M2B), 2-Methyl-1-butanol (2M1B), and 3-Methyl-1-butanol (3M1B) were determined from measured densities (ρ), viscosities (η) and speeds of sound (u) at 303.15 K to 313.15 K. Influence Structural disparities on excess properties is discussed concerning intermolecular forces. The Prigogine – Flory – Patterson (PFP) theory identifies predominant molecular interactions, and the Jouyban – Acree model correlates density, speed of sound, and viscosity with composition. Thermodynamic activation parameters were calculated using the Eyring equation to study viscous flow thermodynamics. FT-IR and 1H NMR spectra of equimolar binary mixtures were analysed to confirm experimental findings through absorption bands (cm−1) and chemical shifts.

Suppression of cross-interference in the absorption spectra of gas mixtures

In the quantitative analysis of mixed gases by tunable diode laser absorption spectroscopy, the overlapping of absorption spectra and mutual interference of multi-component gases can lead to problems of large measurement errors and low analysis accuracy. In this paper, an improved firefly algorithm is proposed and applied to the support vector machine regression model to solve this problem. The specific method includes introducing an adaptive step size to balance the local and global searches and using the gradient descent method to accelerate the parameter optimization process so as to improve the model’s generalization ability and prediction accuracy. The experimental results show that the maximum errors of the improved algorithm in the prediction of CH4 and CO gas concentrations are no more than 0.0443 % and 2 ppm, with coefficients of determination, R2, of 0.9994 and 0.99815. The promising results obtained by the system provide theoretical support for the realization of high-precision detection of multicomponent gases with a single source of light, and also demonstrate the high efficiency and feasibility of the method in practical detection.

Design and Optimization of Solid Lipid Nanoparticles for Enhanced Therapeutic Activity of Linagliptin.

To augment oral bioavailability, linagliptin (LGP) loaded solid lipid nanoparticles (SLNs) have been formed using the solvent evaporation technique of emulsification. Utilizing the Box Behken design (BBD), examining the relation between independent and dependent variables was possible. Particle size (PS), entrapment efficiency (EE), and in-vitro drug release (DR) studies were considered response factors, whereas the various concentrations of lipid carrier, surfactant, and co-surfactant were considered process variables for optimization. The design expert program was used to obtain the optimized batch. At 24 hours, it was discovered that the optimized batch expected responses for PS, %EE, and %DR were 105.83 nm, 81.19, and 96.69%, respectively. The observed PS, %EE, and %DR responses of the optimized batch at 24 hours were determined to be 101.36 nm, 80.56, and 96.31%, respectively. There is no interaction seen in the basic physical mixing of drug and polymer, according to the FTIR spectra of pure drugs, polymers, and drug and polymer mixtures. The optimized formulation’s average size was found to be 138 nm by the particle size analysis. Similarly, a value of 0.255 was found by the polydispersity index (PDI). This suggests that the SLN in the formulation has a homogeneous size dispersion. With a zeta value of -32.8 mV, stability is indicated. Linagliptin was effectively developed as a drug with a 24-hour sustained release. Wistar rats were used in the in-vivo evaluation of linagliptin SLN formulations. The drug bioavailability increased five times (a highly significant fivefold increase) in LGP-loaded SLNs (AUC0-t ≈ 134.5 μg*h/mL) compared to raw LGP (AUC0-t = 27.23 μg*h/mL).

An automated Peak Group Analysis for vibrational spectra analysis

Peak Group Analysis (PGA) is a multivariate curve resolution technique that attempts to extract single pure component spectra from time series of spectral mixture data. It requires that the mixture spectra consist of relatively sharp peaks, as is typical in IR and Raman spectroscopy. PGA aims to construct from individual peaks the associated pure component spectra in the form of nonnegative linear combinations of the right singular vectors of the spectral data matrix.This work presents an automated PGA (autoPGA) that starts with upstream peak detection applied to time series of spectra, combining different window-based peak detection techniques with balanced peak acceptance criteria and peak grouping to deal with repeated detections. The next step is a single-spectrum-oriented PGA analysis. This is followed by a downstream correlation analysis to identify pure component spectra that occur multiple times. AutoPGA provides a complete pure component factorization of the matrix of measured data. The algorithm is applied to FT-IR data sets on various rhodium carbonyl complexes and from an equilibrium of iridium complexes.

Computational, molecular docking and spectroscopic insights of drug–drug interaction between Ibuprofen and Paracetamol

The combination of existing drugs or drug–drug interaction enthralls researcher because of their potential applications in multiple diseases and better efficacy. The present work is undertaken to study the interaction between two drugs, Ibuprofen and Paracetamol, using Density Functional Theory and spectroscopic techniques (Infrared and Raman spectroscopy). The molecular modeling of the two drugs indicates their interaction through intermolecular hydrogen bonds (O34-H53-O2 and (O1-H33-O34), which is observed in Natural Bond Orbital and atoms in molecules analysis of the compound (Ibuprofen + Paracetamol). The experimental Raman, Surface Enhanced Raman Spectroscopy and Infrared spectra of the physical mixture (Ibuprofen + Paracetamol) are found close to their computed wavenumbers. The interacting state, that is, Ibuprofen + Paracetamol is optimized using B3LYP/6-31G ++ (d, p) model under density functional theory framework and different computed quantum chemical parameters such as, dipole moment, ionization energy, electron affinity, electrophilicity index, chemical potential, hardness and the Highest Occupied and Lowest Unoccupied Molecular Orbital energies and energy gaps are calculated and compared to the monomer state of the drugs. The lowering of energy gap and increased in dipole moment of combined drug molecules show potency as an effective hybrid drug. The molecular docking study of the combined drug molecule against the 4PH9 receptor shows a better binding affinity with its residues through hydrogen bonding and hydrophobic interaction.

Investigation of Neighboring Group Participation in 3,4-Diacetylated Glycosyl Donors in the Gas Phase.

A key challenge in oligosaccharide synthesis is the stereoselective installation of glycosidic bonds. Each glycosidic linkage has one of two possible stereo-chemical geometries, α/β or 1,2-cis/trans. An established approach to install 1,2-trans glycosidic bonds is neighboring group participation (NGP), mediated by a 2-O-acyl group. Extension of this intramolecular stabilization to nucleophilic groups located at more remote positions has also been suggested, but remains poorly understood. Previously, we employed infrared ion spectroscopy to characterize the molecular ions of monoacetylated sugar donors and showed how the strength of the stabilizing effect depends on the position of the participating ester group on the glycosyl donor ring as well as on its relative stereochemistry. In this work, we investigated glycosyl donors carrying two acyl groups. Using isotope labelling and isomer population analysis we were able to resolving spectra of isomeric mixtures and establish the relative contribution of individual species. We conclude that 3,4-diacetyl mannosyl donors exclusively form a dioxanium ion as a result of C-3 acyl stabilization. In contrast, the glucosyl and galactosyl cations form mixtures of C-3 and C-4 acyl participation products. Hence, the combination of isotope labeling and population analysis allows for the study of increasingly complex glycosyl cations.

Optimized infrared spectrum of mixtures.

Using density functional theory at D3-B3LYP/aug-cc-pVDZ level combined with the conductor-like polarizable continuum model (CPCM) solvent model, a study of the IR spectrum of :HCN mixtures is reported. The CPCM solvent effect notably enhances the accuracy of the IR spectra compared to gas-phase calculations, while the dielectric constant value has minimum impact on the final spectrum. An optimized methodology is suggested that effectively minimizes the root mean square deviation between theoretical and experimental data. This novel approach not only enhances the quality of the final IR spectra but also captures relevant spectral features, highlighting its potential to decipher molecular interactions in such intricate mixtures.

Rotational spectra of ten new fluoroethylene/CO2 clusters, (C2H3F)x(CO2)y: Application of data-centered methods to the assignment of spectra in complex mixtures

Chirped-pulse Fourier-transform microwave (CP-FTMW) spectroscopy has been used to measure the spectra of ten previously unobserved fluoroethylene (FE)/CO2 clusters, (FE)x(CO2)y, with x from 1 to 4 and y from 0 to 4. Multiple spectra were recorded with varying concentrations of CO2 in the sample, and at least 400,000 free induction decays averaged per data set. This allowed implementation of data-centered approaches, using intensity variation, to identify subsets of transitions belonging to the same cluster species or those of similar composition, simplifying the assignment process for the complex mixture of clusters present. All spectra were fitted to Watson A-reduction Hamiltonians, including quartic distortion constants, with very few species requiring higher order distortion constants for satisfactory fits to be obtained. Computational data at MP2/6-311++G(2d,2p) and ωB97X-D/6-31+G(d,p) levels suggested FE:CO2 ratios and likely structural arrangements of each cluster based on comparisons with experimental rotational constants.

XGBoost algorithm assisted multi-component quantitative analysis with Raman spectroscopy

To improve prediction performance and reduce artifacts in Raman spectra, we developed an eXtreme Gradient Boosting (XGBoost) preprocessing method to preprocess the Raman spectra of glucose, glycerol and ethanol mixtures. To ensure the robustness and reliability of the XGBoost preprocessing method, an X-LR model (which combined XGBoost preprocessing and a linear regression (LR) model) and a X-MLP model (which combined XGBoost preprocessing and a multilayer perceptron (MLP) model) were developed. These two models were used to quantitatively analyze the concentrations of glucose, glycerol and ethanol in the Raman spectra of mixed solutions. The proportion map of hyperparameters was firstly used to narrow down the search range of hyperparameters in the X-LR and the X-MLP models. Then the correlation coefficients (R2), root mean square of calibration (RMSEC), and root mean square error of prediction (RMSEP) were used to evaluate the models’ performance. Experimental results indicated that the XGBoost preprocessing method achieved higher accuracy and generalization capability, with better performance than those of other preprocessing methods for both LR and MLP models.

Improving Quantitative Analysis with Cross Instrument-Sparse Bayesian Learning (CI-SBL) Raman Spectroscopy Analysis Algorithm.

Qualitative and quantitative analysis of Raman spectroscopy is a widely used nondestructive analytical technique in many fields. It utilizes the Raman scattering effect of lasers to obtain molecular vibration information on samples. By comparison with the Raman spectra of standard substances, qualitative and quantitative analyses can be achieved on unknown samples. However, current Raman spectroscopy analysis algorithms still have many drawbacks. They struggled to handle quantitative analysis between different instruments. Their prediction accuracy for concentration is generally low, with poor robustness. Therefore, this study addresses these deficiencies by designing the cross instrument-sparse Bayesian learning (CI-SBL) Raman spectroscopy analysis algorithm. CI-SBL can facilitate spectroscopic analysis between different instruments through the cross instrument module. CI-SBL converts data from portable instruments into data from scientific instruments, with high similarity between the converted spectrum and the spectrum from the scientific instruments reaching 98.6%. The similarity between the raw portable instrument spectrum and the scientific instrument spectrum is often lower than 90%. The cross instrument effect of the CI-SBL is remarkable. Moreover, CI-SBL employs sparse Bayesian learning (SBL) as the core module for analysis. Through multiple iterations, the SBL algorithm effectively identified various components within mixtures. In experiments, CI-SBL can achieve a qualitative accuracy of 100% for the majority of binary and multicomponent mixtures. On the other hand, the previous Raman spectroscopy analysis algorithms predominantly yield a qualitative accuracy below 80% for the same data. Additionally, CI-SBL incorporates a quantitative module to calculate the concentration of each component within the mixed samples. In the experiment, the quantification error for all substances was below 3%, with the majority of the substances exhibiting an error of approximately 1%. These experimental results illustrate that CI-SBL significantly enhances the accuracy of qualitative judgment of mixture spectra and the prediction of mixture concentrations compared with previous Raman spectroscopy analysis algorithms. Furthermore, the cross instrument module of CI-SBL allows for a flexible handling of data acquired from different instruments.

Mid-Infrared Dispersion Spectroscopy as a Tool for Monitoring Time-Resolved Chemical Reactions on the Examples of Enzyme Kinetics and Mutarotation of Sugars.

Ongoing technological advancements in the field of mid-infrared (MIR) spectroscopy continuously yield novel sensing modalities, offering capabilities beyond traditional techniques like Fourier transform infrared spectroscopy (FT-IR). One such advancement is MIR dispersion spectroscopy, utilizing a tunable quantum cascade laser and Mach-Zehnder interferometer for liquid-phase analysis. Our study assesses the performance of a custom MIR dispersion spectrometer at its current development stage, benchmarks its performance against FT-IR, and validates its potential for time-resolved chemical reaction monitoring. Unlike conventional methods of IR spectroscopy measuring molecular absorptions using intensity attenuation, our method detects refractive index changes (phase shifts) down to a level of 6.1 × 10-7 refractive index units (RIU). This results in 1.5 times better sensitivity with a sevenfold increase in analytical path length, yielding heightened robustness for the analysis of liquids compared to FT-IR. As a case study, we monitor the catalytic activity of invertase with sucrose, observing the formation of resultant monosaccharides and their progression toward thermodynamic equilibrium. Anomalous refractive index spectra of reaction mixtures, with substrate concentrations ranging from 2.5 to 25 g/L, are recorded, and analyzed at various temperatures, yielding Michaelis-Menten kinetics findings comparable to the literature. Additionally, the first-time application of two-dimensional correlation spectroscopy on the recorded dynamic dispersion spectra correctly identifies the mutarotation of reaction products (glucose and fructose). The results demonstrate high precision and sensitivity in investigating complex time-dependent chemical reactions via broadband refractive index changes.

Synthesis and characterization of Co(II) porphyrin complex supported on chitosan/graphene oxide nanocomposite for efficient green oxidation and removal of Acid Orange 7 dye

Catalytic degradation of Acid Orange 7 (AO7) by hydrogen peroxide in an aqueous solution has been investigated using cobalt(II) complex of 5, 10, 15, 20 Tetrakis [4-(hydroxy)phenyl] porphyrin [Co(II) TPHPP] covalently supported chitosan/Graphene Oxide nanocomposite [Co(II) TPHPP]-Cs/GO, as highly efficient and recoverable heterogeneous catalyst. The structures and properties of [Co(II) TPHPP]-Cs/GO nanocomposite were characterized by techniques such as UV–Vis, FT-IR, SEM, EDX, TEM, and XRD. The oxidation reaction was followed by recording the UV–Vis spectra of the reaction mixture with time at λmax = 485 nm. [Co(II) TPHPP]-Cs/GO nanocomposite demonstrated high catalytic activity and could decompose 94% of AO7 within 60 min. The factors that may influence the oxidation of Acid Orange 7, such as the effect of reaction temperature, pH, concentration of catalyst, Acid Orange 7, and hydrogen peroxide, have been studied. The results of total organic carbon analysis (TOC) showed 50% of dye mineralization under mild reaction conditions of AO7 (1.42 × 10−4M) with H2O2 (8 × 10−2M) in the presence of [Co(II) TPHPP]-Cs/GO nanocomposite (15 × 10−3 g/ml) and pH = 9 at 40 °C. The reuse and stability of the nanocomposite were examined and remarkably, even after six cycles of reuse, there was no significant degradation or deactivation of the recycled catalyst. Residual organic compounds in the reaction mixture were identified by using GC–MS analyses. The radical scavenging measurements and photoluminescence probing technology of disodium salt of terephthalic acid indicated the formation of the hydroxyl radical as the reactive oxygen species in the [Co(II) TPHPP]-Cs/GO nanocomposite/H2O2 system. A mechanism for the oxidation reaction has been discussed.

JWST Reveals CO Ice, Concentrated CO2 Deposits, and Evidence for Carbonates Potentially Sourced from Ariel’s Interior

The Uranian moon Ariel exhibits a diversity of geologically young landforms, with a surface composition rich in CO2 ice. The origin of CO2 and other species, however, remains uncertain. We report observations of Ariel’s leading and trailing hemispheres, collected with NIRSpec (2.87–5.10 μm) on the James Webb Space Telescope. These data shed new light on Ariel's spectral properties, revealing a double-lobed CO2 ice scattering peak centered near 4.20 and 4.25 μm, with the 4.25 μm lobe possibly representing the largest CO2 Fresnel peak yet observed in the solar system. A prominent 4.38 μm 13CO2 ice feature is also present, as is a 4.90 μm band that results from 12CO2 ice. The spectra reveal a 4.67 μm 12CO ice band and a broad 4.02 μm band that might result from carbonate minerals. The data confirm that features associated with CO2 and CO are notably stronger on Ariel’s trailing hemisphere compared to its leading hemisphere. We compared the detected CO2 features to synthetic spectra of CO2 ice and mixtures of CO2 with CO, H2O, and amorphous carbon, finding that CO2 could be concentrated in deposits thicker than ∼10 mm on Ariel’s trailing hemisphere. Comparison to laboratory data indicates that CO is likely mixed with CO2. The evidence for thick CO2 ice deposits and the possible presence of carbonates on both hemispheres suggests that some carbon oxides could be sourced from Ariel’s interior, with their surface distributions modified by charged particle bombardment, sublimation, and seasonal migration of CO and CO2 from high to low latitudes.

Resolution Enhancement of Metabolomic J-Res NMR Spectra Using Deep Learning.

J-Resolved (J-Res) nuclear magnetic resonance (NMR) spectroscopy is pivotal in NMR-based metabolomics, but practitioners face a choice between time-consuming high-resolution (HR) experiments or shorter low-resolution (LR) experiments which exhibit significant peak overlap. Deep learning neural networks have been successfully used in many fields to enhance quality of natural images, especially with regard to resolution, and therefore offer the prospect of improving two-dimensional (2D) NMR data. Here, we introduce the J-RESRGAN, an adapted and modified generative adversarial network (GAN) for image super-resolution (SR), which we trained specifically for metabolomic J-Res spectra to enhance peak resolution. A novel symmetric loss function was introduced, exploiting the inherent vertical symmetry of J-Res NMR spectra. Model training used simulated high-resolution J-Res spectra of complex mixtures, with corresponding low-resolution spectra generated via blurring and down-sampling. Evaluation of peak pair resolvability on J-RESRGAN demonstrated remarkable improvement in resolution across a variety of samples. In simulated plasma data, 100% of peak pairs exhibited enhanced resolution in super-resolution spectra compared to their low-resolution counterparts. Similarly, enhanced resolution was observed in 80.8-100% of peak pairs in experimental plasma, 85.0-96.7% in urine, 94.4-98.9% in full fat milk, and 82.6-91.7% in orange juice. J-RESRGAN is not sample type, spectrometer or field strength dependent and improvements on previously acquired data can be seen in seconds on a standard desktop computer. We believe this demonstrates the promise of deep learning methods to enhance NMR metabolomic data, and in particular, the power of J-RESRGAN to elucidate overlapping peaks, advancing precision in a wide variety of NMR-based metabolomics studies. The model, J-RESRGAN, is openly accessible for download on GitHub at https://github.com/yanyan5420/J-RESRGAN.

A feature selection method for overlapping peaks in vibrational spectroscopy using nonnegatively constrained classical least squares

Chemometric models for multicomponent mixture analysis are reliant upon representative and accurate training data. However, the required amount of training data can increase exponentially with the number of constituents. Shifting spectral baselines and changing process objectives may necessitate recurring calibration data collection. Furthermore, all possible constituents may not be known prior to analysis. Here, we introduce a preprocessing procedure that reduces the burden of collecting extensive calibration datasets by filtering out non-target species (species that are not part of the calibration data) in real-time. The method, nonnegatively constrained classical least squares (NCCLS), utilizes a spectral nonnegativity constraint on non-target species to remove them from mixture spectra. The preprocessing method is physically motivated, does not rely on time-series data, and can operate in real-time. NCCLS is compared to established methods using in silico data and an experimental dataset comprised of Raman spectra and attenuated total reflectance - Fourier transform infrared spectra.

Pressure broadening and narrowing due to oxygen and nitrogen gas mixtures at 1270 nm band: part III

Collisional effects on the absorption spectrum of oxygen and nitrogen gas mixtures at 1270 nm band appear as a broadening or narrowing at the component of Lorentzian width in the Voigt profile. The deconvolution method of spectral lines reveals two kinds of bands, the first centered at 1264 nm, and the second centered at 1268 nm. At high pressures, the Gaussian full width may be fixed, and Voigt's full width at half maximum (FWHM) nearly equals to Lorentzian full width. Partial pressures of each gas were between 1-10 bar at 298 K. New results highlight the effect of nitrogen as a collisional partner on the oxygen spectrum. PACS: 32.30. Bv; 33.20. Ea; 33.20.-t; 34.50.-s; 34.50. Ez; 51.30. +I; 87.64. km.

PFAS remediation: Evaluating the infrared spectra of complex gaseous mixtures to determine the efficacy of thermal decomposition of PFAS

Due to their widespread production and known environmental contamination, the need for the detection and remediation of per- and polyfluoroalkyl substances (PFAS) has grown quickly. While destructive thermal treatment of PFAS at low temperatures (e.g., 200–500 °C) is of interest due to lower energy and infrastructure requirements, the range of possible degradation products remains underexplored. To better understand the low temperature decomposition of PFAS species, we have coupled gas-phase infrared spectroscopy with a multivariate curve resolution (MCR) analysis and a database of high-resolution PFAS infrared reference spectra to characterize and quantify a complex mixture resulting from potassium perfluorooctanesulfonate (PFOS–K) decomposition. Beginning at 375 °C, nine prevalent decomposition products (namely smaller perfluorocarbon species) are identified and quantified.

Explainable AI to Facilitate Understanding of Neural Network-Based Metabolite Profiling Using NMR Spectroscopy.

Neural networks (NNs) are emerging as a rapid and scalable method for quantifying metabolites directly from nuclear magnetic resonance (NMR) spectra, but the nonlinear nature of NNs precludes understanding of how a model makes predictions. This study implements an explainable artificial intelligence algorithm called integrated gradients (IG) to elucidate which regions of input spectra are the most important for the quantification of specific analytes. The approach is first validated in simulated mixture spectra of eight aqueous metabolites and then investigated in experimentally acquired lipid spectra of a reference standard mixture and a murine hepatic extract. The IG method revealed that, like a human spectroscopist, NNs recognize and quantify analytes based on an analyte's respective resonance line-shapes, amplitudes, and frequencies. NNs can compensate for peak overlap and prioritize specific resonances most important for concentration determination. Further, we show how modifying a NN training dataset can affect how a model makes decisions, and we provide examples of how this approach can be used to de-bug issues with model performance. Overall, results show that the IG technique facilitates a visual and quantitative understanding of how model inputs relate to model outputs, potentially making NNs a more attractive option for targeted and automated NMR-based metabolomics.

Molecular interaction study using density, viscosity, speed of sound and FT-IR in 2-methyl-2-butanol + alkoxyethanols: Experimental and modelling approach

In this article, the intermolecular interactions in the binary liquid mixtures of 2-methyl-2-butanol with alkoxyethanols viz., 2-ethoxyethanol, 2-propoxyethanol and 2-butoxyethanol have been interpreted using thermophysical properties. Density (ρ), viscosity (η) and speed of sound (u) of pure components and their binary mixtures have been measured over the entire range of composition at 298.15, 303.15 and 308.15 K and at pressure p = 0.10 MPa. The experimental data were used to calculate the thermophysical properties such as excess molar volume (VmE), excess molar isentropic compressibility (Ks,mE), excess speed of sound (uE), deviation in viscosity (Δη) and excess Gibbs free energy of activation for viscous flow (ΔG∗E). The changes in these properties with composition and temperature of the liquid mixtures were discussed in terms of intermolecular interactions. Also, the effect of increase in alkyl chain length of alkoxyethanol molecule on these properties has been studied. The excess functions and their deviations have been correlated using Redlich-Kister type polynomial equation. The experimentally measured values of viscosities have been correlated with various semi-empirical relations such as Grunberg-Nissan, Tamura-Kurata, Hind et al., Katti-Chaudhri, McAllister’s two parameter model, Heric-Brewer and McAllister’s three parameter model. Apart from this, Prigogine-Flory-Patterson (PFP) theory has been used for the estimation of VmE values for studied binary mixtures. FT-IR spectra of these liquid mixtures have also been studied.

Identification of Reaction Network Hypotheses for Complex Feedstocks from Spectroscopic Measurements with Minimal Human Intervention.

In this work, we detail an automated reaction network hypothesis generation protocol for processes involving complex feedstocks where information about the species and reactions involved is unknown. Our methodology is process agnostic and can be utilized in any reactive process with spectroscopic measurements that provide information on the evolution of the components in the mixture. We decompose the mixture spectra to obtain spectroscopic signatures of the individual components and use a 1-D convolutional neural network to automatically identify functional groups indicated by them. We employ atom-atom mapping to automatically recover reaction rules that are applied on candidate molecules identified from chemistry databases through fingerprint similarity. The method is tested on synthetic data and on spectroscopic measurements of lab-scale batch hydrothermal liquefaction (HTL) of biomass to determine the accuracy of prediction across datasets of varying complexities. Our methodology is able to identify reaction network hypotheses containing reaction networks close to the ground truth in the case of synthetic data, and we are also able to recover candidate molecules and reaction networks close to the ones reported in the previous literature studies for biomass pyrolysis.