Machine learning and design of experiments with an application to product innovation in the chemical industry.

Industrial statistics plays a major role in the areas of both quality management and innovation. However, existing methodologies must be integrated with the latest tools from the field of Artificial Intelligence. To this end, a background on the joint application of Design of Experiments (DOE) and Machine Learning (ML) methodologies in industrial settings is presented here, along with a case study from the chemical industry. A DOE study is used to collect data, and two ML models are applied to predict responses which performance show an advantage over the traditional modeling approach. Emphasis is placed on causal investigation and quantification of prediction uncertainty, as these are crucial for an assessment of the goodness and robustness of the models developed. Within the scope of the case study, the models learned can be implemented in a semi-automatic system that can assist practitioners who are inexperienced in data analysis in the process of new product development.

Preparation of gold nanoparticles on silica substrate by radio frequency sputtering.

Radio frequency sputtering of gold on amorphous silica substrates was used for the preparation of Au nanoparticles on SiO2. Deposition experiments were carried out in Ar plasmas under mild conditions (RF power = 5/10 W, total pressure = 0.38 mbar, substrate temperature < or = 210 degrees C), focusing in particular on the effect of sputtering time (5/30 min) and substrate temperature on gold nucleation and coalescence, with the aim of obtaining SiO2-supported Au nanoparticles characterized by precise structural and morphological features. To this aim, several analytical techniques were employed for a thorough characterization of the systems properties, including glancing incidence X-ray diffraction, transmission electron microscopy, X-ray photoelectron spectroscopy, atomic force microscopy, and UV-Vis absorption spectroscopy. In particular, the evolution of optical spectra, i.e., of the surface plasmon resonance peak, was used as a probe for the structural features and was related to the results obtained by other characterization techniques. Gold nanoclusters (phi approximately 4/10 nm) dispersed uniformly on silica matrices were obtained under soft conditions, with morphology ranging from island to cluster like. The obtained results make possible a careful modulation of substrate coverage and gold nanoparticle size.