Application of Machine Learning Algorithms to Classify Peruvian Pisco Varieties Using an Electronic Nose.

Pisco is an alcoholic beverage obtained from grape juice distillation. Considered the flagship drink of Peru, it is produced following strict and specific quality standards. In this work, sensing results for volatile compounds in pisco, obtained with an electronic nose, were analyzed through the application of machine learning algorithms for the differentiation of pisco varieties. This differentiation aids in verifying beverage quality, considering the parameters established in its Designation of Origin". For signal processing, neural networks, multiclass support vector machines and random forest machine learning algorithms were implemented in MATLAB. In addition, data augmentation was performed using a proposed procedure based on interpolation-extrapolation. All algorithms trained with augmented data showed an increase in performance and more reliable predictions compared to those trained with raw data. From the comparison of these results, it was found that the best performance was achieved with neural networks.

Biosorción de plomo (II) en solución acuosa con biomasa de los cladodios de la tuna (Opuntia ficus indica) / Biosorption of lead (II) in aqueous solution with biomass of prickly pear cladodes (Opuntia ficus indica) / Biossorção de chumbo (II) em solução aquosa com biomassa de cladódios de pera espinhosa (Opuntia ficus indica)

Resumen Se investigó la remoción de plomo de soluciones acuosas por biomasa de los cladodios de la tuna (BCT), una especie abundante en la sierra peruana. Las propiedades físicas y químicas de la BCT se determinaron por técnicas de FTIR y SEM/ EDX. El análisis FTIR confirmó la presencia de grupos -OH, CO-C, C=O y -NH2, los cuales interactuarían con el metal. La micrografía SEM reveló que la morfología superficial de la BCT tiene porosidad adecuada para la biosorción. En un sistema discontinuo se obtuvo la más alta capacidad de biosorción (qe) con una relación masa BCT/volumen de 4 g/L, pH 4,5 y tiempo de contacto de 1 h. La biosorción siguió un modelo cinético de pseudo-segundo orden y el proceso de difusión intraparticular fue el factor principal de control de la velocidad. Los datos en el equilibrio fueron correlacionados con cinco modelos (Langmuir, Freundlich, Temkin, D-R y Redlich-Peterson) y se ajustaron mejor al modelo de Langmuir, lo que indica que el proceso se realiza en centros activos energéticamente homogéneos, el valor de qe máxima fue 50,25 mg/g. Los resultados obtenidos demuestran que la BCT puede ser usada como biosorbente eficiente para el tratamiento de aguas contaminadas con Pb (II).

Abstract The removal of lead from aqueous solutions by the biomass of prickly pear cladodes (BCT), a species found in abundance in the Sierra region in Peru, was investigated. The physical and chemical properties of BCT were determined by FTIR and SEM/EDX techniques. The FTIR analysis confirmed the presence of -OH, C-O-C, C=O and -NH2 groups which would interact with the metal. The SEM micrograph revealed that the surface morphology of the BCT shows adequate porosity for biosorption. In a discontinuous system, the highest biosorption capacity (qe) was obtained with a BCT mass/volume ratio of 4 g/L, pH 4.5 and 1 h of contact time. Biosorption followed a pseudo-second order kinetic model and the intraparticle diffusion process was the main factor controlling speed. Data at equilibrium were correlated using five models (Langmuir, Freundlich, Temkin, DR, and Redlich-Peterson) and were better fitted to the Langmuir model, which would indicate that the process is carried out in energetically homogeneous active centers, the value of qe maximum was 50.25 mg/g. The results obtained demonstrate that BCT can be used as an efficient biosorbent for the treatment of waters contaminated with Pb (II).

Resumo Investigou-se a remoção de chumbo de soluções aquosas pela biomassa de cladódios de pera espinhosa (BCT), espécie encontrada em abundância na região da Sierra, no Peru, e as propriedades físicas e químicas do BCT foram determinadas por técnicas de FTIR y SEM/EDX. A análise por FTIR confirma a presença dos grupos -OH, C-O-C, C=O e -NH2 que interagem com o metal; a micrografia SEM revela que a morfologia da superfície do BCT mostra porosidade adequada para biossorção. Em um sistema descontínuo, a maior capacidade de biossorção (qe) foi obtida com uma relação massa/volume de BCT de 4 g/L, pH 4,5 e tempo de contato 1 h. A biossorção seguiu um modelo cinético de pseudo-segunda ordem e o processo de difusão intrapartículas foi o principal fator de controle da velocidade. Os dados em equilíbrio foram correlacionados usando cinco modelos (Langmuir, Freundlich, Temkin, DR e Redlich-Peterson) e foram melhor ajustados ao modelo de Langmuir, o que indicaria que o processo é realizado em centros ativos energeticamente homogêneos, o valor de qe o máximo foi de 50,25 mg/g. Os resultados obtidos demonstram que o BCT pode ser usado como um biossorvente eficiente para o tratamento de águas contaminadas com Pb (II).

Comparative study of the removal of phenolic compounds by biological and non-biological adsorbents.

The ability of biological and non-biological adsorbents to remove 2-nitrophenol (2-NP) and 2-chlorophenol (2-CP) from aqueous solutions in batch experiments at room temperature was compared. The marine seaweeds Macrocystis integrifolia Bory (S1) and Lessonia nigrescens Bory (S2) were cross-linked with CaCl(2) to enhance their mechanical properties. Natural bentonite was chemically exchanged with hexadecyltrimethylammonium bromide (B1) and bencyltriethylammonium chloride (B2) to increase their affinity towards organic compounds as well. The adsorption capacity of all of the adsorbents strongly depends on solution pH, whereas equilibrium assays showed a mixed mechanism according to the Langmuir and Freundlich isotherms. The maximum adsorption capacity of 2-NP follows the trend: S1>S2>B2>B1 within the range of 97.37 and 18.64 mg g(-1) whereas for 2-CP, it ranged between 24.18 and 9.95 mg g(-1) with the trend: S1>S2>B2 approximately B1. The importance of the octanol-water partition coefficient as the main factor on the adsorption of these compounds on two different kinds of adsorbents is discussed.

Effect of pH on phenol biosorption by marine seaweeds.

Biosorption of phenol by the marine seaweeds Lessonia nigrescens Bory and Macrocystis integrifolia Bory was investigated as a function of initial solution pH, showing a higher adsorption percentage at pH 10 with values between 10% and 35%, respectively. The apparent ionization constants of the algae were determined by means of the Katchalsky's theory, obtaining values close to 3.0 for both seaweeds. From the results, a strong adsorption dependence on pH is proposed at the level of the phenol's aqueous chemistry driven purely by a polar mechanism that involves the formation of hydrogen bonds with the hydroxyl groups that are spatially arranged in the polysaccharic chains that form the seaweed structure. This study shows that both marine algae are potential biosorbents in their application for the removal of phenol and derivatives from residual waters.