AComDim as a multivariate tool to analyse experimental design application to γ-irradiated and leached ion exchange resins

Abstract

International audienceLeaching MIR spectroscopy Ion exchange resin γ-Irradiations of ion exchange resins were carried out under various experimental conditions, selected using an experimental design, to simulate the ageing of such nuclear wastes. Those resins are a commercial mixed bed, constituted of 75 wt.% of a cationic resin and 25 wt.% of an anionic resin, and both are pure cationic and anionic resins. Then, irradiated samples were leached, at two different temperatures (20 °C and 50 °C). Solid matrices were characterized by a spectroscopic technique and spectral data were analysed by the ANOVA Common Dimensions (AComDim) method to detect and highlight influential factors and interactions. Four factors were studied (absorbed dose under irradiation, atmosphere during irradiation, dose rate, leaching temperature and resin type) with several levels for each. In a first study of the separate resins, the absorbed dose and the irradiation atmosphere appeared as the most influencing factors for the anionic resin. The leaching temperature was significant but only in interaction with other factors, while this parameter was significant for the cationic and mixed bed resins. In a global study of all the spectral data, the resin type appeared as the most influential factor for the radiolysis of resins

AComDim, a multivariate tool to highlighting impact of agroclimatic factors on Moringa oleifera Lam. leaf’s composition from their FTIR-ATR profiles

“…This was achieved by using ANOVA Common Dimension (AComDim) method to prioritize the significant factors and their interactions from Fourier Transform Infrared (FTIR) signatures of MOL leaf samples, harvested over three years and in different agroclimatic zones of Cameroon. This multi-block method has shown its effectiveness in different fields [16][17][18][19] and is especially interesting because it uses all the spectral variance for each level of factors studied. In this study, five factors with different sub-levels were considered such as the collection year and month, the agroclimatic zone of harvest (defined according to Liénou et al [20], the soil nature (defined according to the classification of Segalen [21]) and the maturity of MOL leaves.…”

A tutorial on the analysis of multifactorial designs from one or more data sources using AComDim

“…As a method for the analysis of data with an associated multifactorial design, AComDim was applied to different types of data. 19,20,[35][36][37] The objective of this tutorial is to focus on a dataset produced from multiple sources while describing the same samples so as to demonstrate how the multiblock nature of AComDim can be useful to tackle such complex data structures. To do so, this section will illustrate how AComDim can be applied for the analysis of multifactorial design data acquired from one or more data sources using a dataset with signals of starch-lignin mixtures representing different Time Domain-Nuclear Magnetic Resonance (TD-NMR) relaxation curves.…”

“…As a method for the analysis of data with an associated multifactorial design, AComDim was applied to different types of data 19,20,35‐37 . The objective of this tutorial is to focus on a dataset produced from multiple sources while describing the same samples so as to demonstrate how the multiblock nature of AComDim can be useful to tackle such complex data structures.…”

A synthetic review of some recent extensions of ComDim

“…Since the variability common to all the matrices is the contribution of the residuals, CC1 tends to characterize this noise. Parameters computed during the AComDim procedure enable the computation of F‐values to evaluate whether the variability of each factor + noise block in CC1 is significantly less than that of the noise block 28–32 . This can be done by comparing the saliences of the factor and interaction blocks on CC1 to those of the residual block.…”