PHASE 2 | 2021

Duration: 12 months

Summary report 

The specific objectives of the second phase of the INFRA-ART project were: to carry out a technical study for the identification of diagnostic spectral features, chemical patterns, and fingerprints within historical pigments; and to develop an automatic algorithm (based on multivariate analysis of spectroscopic data) for pigment classification and identification. According to the working plan in the second phase of the INFRA-ART project five tasks were carried out:

·       Task 2.1 Preparation of reference paint samples. Data collection (Part II)

·       Task 2.2 Development of integrated spectral database (Part II)

·       Task 2.3 Workflow for material identification/discrimination

·       Task 2.4 Algorithm design for rapid classification and identification of paint materials

·       Task 2.5 Testing and validation of developed algorithm on complex mixtures and artworks (Part I)

Tasks 2.1 and 2.2, started in the previous stage, were continued and completed. Around 80 demonstrative paint samples were prepared – mixtures of pigments as well as chromatic overlaps. A new set of reference materials was investigated by complementary spectroscopic techniques (220 reference materials). The registered FTIR and XRF data were integrated in the spectral library developed within the project. The database was implemented and optimized during this year at the web address https://infraart.inoe.ro. The INFRA-ART Spectral Database is an Open Access resource, which currently contains 334 reference materials to which 646 spectra are associated.

Within Task 2.3 a technical study was performed on the characterization and discrimination of pigments by combined FTIR and XRF analysis. A set of 30 natural earth pigments from various geological sources was selected and investigated. Diagnostic spectral features and chemical patterns (fingerprints) were identified and discussed for each particular source.

In Task 2.4 a multivariate statistical treatment of the FTIR data set obtained on the group of earth pigments was tested in order to gain an in-depth understanding of the variance within the sample set. Principal Component Analysis (PCA) was used to identify patterns and calculate probabilities for sample grouping. Several IR regions were selected to carry out the PCA analysis, in  accordance with the characteristic absorption patterns of the main chromophores present in the samples. The PCA of the data matrix from samples containing a similar type of chromophore showed a pattern distribution with respect to the geological origin (provenance). Within Task 2.5 the existing PCA model was applied on a series of earth pigments coming from Roman wall painting fragments that were previously investigated by complementary non-destructive techniques.

Dissemination of the scientific results obtained was achieved through a communication at a prestigious international conference (11th International Conference on Advanced Vibrational Spectroscopy), respectively by publishing one ISI article in the journal Applied Sciences.

Materials included within the INFRA-ART database

In this stage 220 reference materials have been investigated via combined FTIR and XRF analysis.  An index of the investigated materials can be consulted in the Appendix below. FTIR and XRF spectra registered (including data registered within stage 1) were included in the INFRA-ART Spectral Database, available online: https://infraart.inoe.ro/.

Appendix

Case studies

>> Roman wall painting fragmnets (2nd–3rd century A.D, found at Alburnus Maior and Ulpia Traiana Sarmizegetusa) from the collection of The National History Museum of Romania