Gurden S. P., Ferreira M. M. C., “Analysis of video images used to study gas-liquid transfer”. Seattle, WA, The United States, 22-26/09/2002: Eighth International Conference on Chemometrics in Analytical Chemistry (CAC VIII, CAC-2002). Program and Abstracts Book, 95 (2002). Poster 33.

       Title:   Analysis of video images used to study gas-liquid transfer

  Authors:    Stephen P. Gurden, Euler M. Lage, Cristiano G. de Faria, Inés Joekes,
                   Márcia M. C. Ferreira
                   Instituto de Química - UNICAMP; Universidade Estadual de Campinas;
                   Campinas, São Paulo 13081-970; Brazil

Keywords:  chemical imaging, gas-liquid transfer, carbon dioxide exchange, PCA, PARAFAC

Presenter:  Stephen Gurden, spgurden@iqm.unicamp.br

            The use of themical imaging is a developing area that has potential benefits for chemical
systems where spatial distribution is important. Examples include processes in which homogeneity
is critical, such as polymerizations, pharmaceutical power blending and surface catalysis, and
dynamic processes such as the study of diffusion rates or the transport of environmental pollutants.
The advent of high-resolution spectroscopic imaging instrumentation along with the continual
increase in data storage and processing power suggest that chemical imaging will become an
important tool in the future.

            The exchange of CO₂ between the air and sea is an important process in terms of the global
mass cycling sytem. In this work, we present a study of images taken from an experiment in which
the exchange of CO₂ from air to water is investigated under controlled temperature and salinity
conditions. The presence of a pH indicator in the water produces a color change that enables the
update of CO₂ to be followed dynamically using a standard video camera.

            Prior to statistical analysis on the data, it is necessary to perform a reconciliation step
(image rotation, cropping and resizing) that yields congruent image arrays. The analysis of single
multivariate images using the PARAFAC model is then described, and contrasted with PCA, as a
form of understanding the relationships between the spatical and wavelength directions. The use of
other single image transformations is also described, including the use of histograms (particularly
meaningful here) and pH imaging. The analysis of multiple images related in time (i.e. movies)
using both PCA and PARAFAC is then described, along with different ways of preprocessing movie
arrays. The model components found are used to understand factors common to all the CO₂/water
exchange experiments, as well as factors specific to particular experimental runs.

             This work is financed by the State of São Paulo Research Foundation (FAPESP).
 
 
 
 

                                                                                95