Deep learning analysis for rapid detection and classification of household plastics based on Raman spectroscopy

The overuse of plastics releases large amounts of microplastics. These tiny and complex pollutants may cause immeasurable damage to human social life. Raman spectroscopy detection technology is widely used in the detection, identification and analysis of microplastics due to its advantages of fast speed, high sensitivity and non-destructive. In this work, we first recorded the Raman spectra of eight common plastics in daily life. By adjusting parameters such as laser wavelength, laser power, and acquisition time, the Raman data under different acquisition conditions were diversified, and the corresponding Raman spectra were obtained, and a database of eight household plastics was established. Combined with deep learning algorithms, an accurate, fast and simple classification and identification method for 8 types of plastics is established. Firstly, the acquired spectral data were preprocessed for baseline correction and noise reduction, Then, four machine learning algorithms, linear discriminant analysis (LDA), decision tree, support vector machine (SVM) and one-dimensional convolutional neural network (1D-CNN), are used to classify and identify the preprocessed data. The results showed that the classification accuracy of the three machine learning models for the Raman spectra of standard plastic samples were 84%, 93% and 93% respectively. The 1D-CNN model has an accuracy rate of up to 97% for Raman spectroscopy. Our study shows that the combination of Raman spectroscopy detection techniques and deep learning algorithms is a very valuable approach for microplastic classification and identification.

Prediction of vibrational spectrum and thermodynamic properties for phosphorus mononitride

In this work, an accurate potential energy curve (PEC) for the ground electronic state of phosphorus mononitride (PN) molecule has been determined from a variationally improved Hulburt-Hirschfelder (VIHH) oscillator model in conjunction with the experimental spectral constants (De,ωe,ωexe,Be,αe,re). We have numerically solved the Schrödinger equation for the VIHH potential using the LEVEL program, obtaining the pure vibrational spectrum that converges to the dissociation limit. In addition, the partition functions of PN molecule are calculated using the full rovibrational energies. Ultimately, thermodynamic properties like molar heat capacity, entropy, enthalpy, and Gibbs free energy were calculated for the PN molecule and show good agreement with those data from the NIST (National Institute of Standards and Technology) database.

Theoretical study on the vibrational spectra and potential energy curves for SiC (X3Π) and SiS (X1Σ+) molecules

In this work, the vibrational constants (ω_e,ω_ex_e) calculated by the variational algebraic method (VAM) and some other molecular constants (D_e,r_e,B_e,α_e) were used to construct the improved Hulburt-Hirschfelder (IHH) analytical potential energy function (APEF). Not only that, but the calculated VAM potential points are used as the 'true' energies to determine the value of the variational parameter λ which is the pivotal fitting parameter in the IHH potential. With limited experimental data, high-precision IHH potential can be achieved by combining the VAM and the IHH APEF. This combination of the VAM and the IHH APEF is referred to be VAIHH APEF, which is employed to study the vibrational energies and potential energy curves (PECs) of SiC (X³Π) and SiS (X¹Σ⁺) molecules, yielding full vibrational spectra and spectroscopic constants. The calculational results indicate that the VAIHH APEFs of SiC (X³Π) and SiS (X¹Σ⁺) molecules are in good agreement with the experimental RKR potential points. Accurate PECs of SiC (X³Π) and SiS (X¹Σ⁺) molecules imply that the VAIHH APEF is of high quality.

Deciphering the structural preference encoded in amide-I vibrations of lysine dipeptide in gas phase and in aqueous solution

Protein's biological function is critically associated with its structural feature, which is encoded in its amino acid sequence. For evaluation of conformational fluctuation and folding mechanism, DFT calculations were performed on the model compound - lysine dipeptide (LYSD) in gas phase to demonstrate the correlation between amide-I vibrations and secondary structure. Molecular dynamics simulations were carried out for the structural dynamics of LYSD in aqueous solution. The results show that LYSD tends form C_7eq, C₅, β, PP_II and α conformations in the gas phase and primarily presented PP_II and α conformations in aqueous solution. The obtained amide-I vibrational frequencies of LYSD conformers were assigned, thus build the correlations between amide-I probes and secondary structure of LYSD. These results provide theoretical insights into the structural feature of LYSD through amide-I vibrations, and would shed light on site specific structural prediction of polypeptides.

Resonance Raman scattering-infrared absorption dual-mode immunosensing for carcinoembryonic antigen based on ZnO@SiO2 nanocomposites

Detection of cancer biomarkers is crucial for the diagnosis and monitoring of malignant tumors. However, the accuracy and sensitivity still require sufficient improvement for practically clinical application. In this work, a reliable and sensitive dual-mode immunosensing method is described for carcinoembryonic antigen (CEA) detection using a biofunctional ZnO@SiO₂ nanocomposite as a resonance Raman scattering (RRS)-infrared (IR) absorption nanoprobe. The multiphonon RRS signal originating from the ZnO and the characteristic IR fingerprint signal of the transverse optical and longitudinal optical phonon modes of the asymmetric stretching of Si-O-Si bonds showed no interference with each other. A CEA antibodies-immobilized substrate was fabricated to capture the analyte/nanoprobe complexes. The RRS intensity at 569 cm^‒1 and the IR absorption at 1061 cm^‒1 were used for quantitative analysis. Accurate CEA detection was performed as a result of the strong resistance of the dual-mode nanoprobe to surrounding interference. The limit of detection was 98.0 fg mL^‒1. The detection range was 500 ng mL^‒1 - 50 fg mL^‒1, which is wider than those of single-mode RRS or IR absorption immunosensings. High reproducibility, selectivity and specificity were achieved. The assay performance of human serum samples demonstrated the practicability of the method in clinical cancer diagnosis.

Theoretical investigation on vibrational spectra, first order hyperpolarizability and NBO analysis of 4-Phenylpyridinium hydrogen squarate

The vibrational frequencies of 4-Phenylpyridinium hydrogen squarate (4PHS) in the ground state have been investigated by using B3LYP/6-311++G(d,p) level. The analysis of molecular structure, natural bond orbitals and frontier molecular orbitals was also performed. The IR spectra were obtained and interpreted by means of potential energies distributions (PEDs) using MOLVIB program. NBO analysis proved the presence of C-H⋯O and N-H⋯O hydrogen bonding interactions, which is consistent with the analysis of molecular structure. The dipole moments and first-order hyperpolarizability (βtot) are calculated and are 5.856 D and 4.72×10(-30) esu, respectively. The high βtot value and the low HOMO-LUMO energy gap (4.062eV) are responsible for the optical and electron-transfer properties of 4PHS molecule. The photoresponse-related results indicate that 4PHS molecule is an excellent organic candidate of photon-responsive materials.

Investigation of crystal structure, vibrational characteristics and molecular conductivity of 2,3-dichloro-5,6-dicyno-p-benzoquinone

Molecular geometries and vibrational spectra for the ground state of 2,3-dichloro-5,6-dicyno-p-benzoquinone (DDQ) and its anion (DDQ(-)) were computed using DFT method at the B3LYP level employing 6-311++G(d,p) basis set whereas for the first excited state (DDQ(∗)), these were calculated using TD-DFT at the B3LYP level employing the 6-311++G(d,p) basis set available with the Gaussian 09 package. The spectra have been experimentally investigated and the observed IR and Raman bands have been assigned to different normal modes on the basis of the calculated potential energy distributions (PEDs). XRD of single crystal has been investigated to determine molecular and crystal structures of DDQ. In order to elucidate the transfer of electrons, electronic structure and electronic absorption have been calculated with the TD-DFT method. Vibronic interaction and its role in the appearance of superconductivity in the DDQ, DDQ(-) and DDQ(∗) molecules have been investigated. The present XRD, molecular, electronic and vibronic studies indicate that mainly the ag C=O stretching and ring stretching modes participate in the charge transfer process.

Synchrotron FT-FIR spectroscopy of nitro-derivatives vapors: new spectroscopic signatures of explosive taggants and degradation products

We report on the first successful rovibrational study of gas phase mononitrotoluene and dinitrotoluene in the TeraHertz/Far-Infrared (THz/FIR) spectral domain. Using the AILES beamline of the synchrotron SOLEIL and a Fourier Transform spectrometer connected to multipass cells, the low-energy vibrational cross-sections of the different isomers of mononitrotoluene have been measured and compared to calculated spectra with the density functional theory including the anharmonic contribution. The active FIR modes of 2,4 and 2,6 dinitrotoluene have been assigned to the vibrational bands measured by Fourier Transform FIR spectroscopy of the gas-phase molecular cloud produced in an evaporating/recondensating system. This study highlights the selectivity of gas phase THz/FIR spectroscopy allowing an unambiguous recognition and discrimination of nitro-aromatic compounds used as explosive taggants.

One-pot synthesis, FT-IR and density functional method (DFT) studies on N-benzyl-N-ethyl-N-[5-nitro-2-(1,1,3,3-Tetramethylbutylamino)-1-benzofuran-3-yl]amine

N-benzyl-N-ethyl-N-[5-nitro-2-(1,1,3,3-Tetramethylbutylamino))-1-benzofuran-3-yl]amine, 7, was synthesized and characterized by elemental analysis, FT-IR, (1)H and (13)C NMR spectra. The geometrical structures, vibrational frequencies, and molecular electrostatic potential (MEP) of 7 in the ground state were calculated by using the DFT/B3LYP method with 6-31 G(d) basis set. The optimized structure was compared with analogous compound. The IR spectrum of 7 was interpreted in terms of potential energy distribution (PED) analysis. Comparison of the experimental vibrational spectra of 7 with the theoretical data indicated good agreement.