Abstract
Near infrared (NIR) spectroscopy has been investigated as a tool for monitoring emulsion polymerisation reactions using multivariate calibration models for estimating mean particle size and monomer concentration. While the models were promising, the studies used small datasets which do not sufficiently span the reaction profile space having very similar trajectories for all the reactions used to build and test the models. This leads to potential pitfalls in the building of robust calibration models and also in their rigorous assessment.
We investigate the use of total diffuse reflectance measurements for the estimation of mean particle size over the course of a styrene emulsion polymerisation reaction run in batch mode. The region of the electromagnetic spectrum considered ranged from ultraviolet to the NIR (300-1900nm). Seventeen reactions were carried out. The experiments were designed such that the reactions varied significantly in terms of the reaction profile, final conversion and particle size range.
Partial least squares (PLS) regression was used to build models for estimating mean particle size. Different pre-processing methods and wavelength range of measurements were investigated to build the best performing calibration model. This analysis was made using a novel cross-validation approach which ensures that representative samples were used for building models and evaluating their performance. It was found that a combination of the visible-NIR region covering the wavelength range 376-852nm and 1548 nm - 1876 nm provided the best performing models for predicting mean particle size.
We investigate the use of total diffuse reflectance measurements for the estimation of mean particle size over the course of a styrene emulsion polymerisation reaction run in batch mode. The region of the electromagnetic spectrum considered ranged from ultraviolet to the NIR (300-1900nm). Seventeen reactions were carried out. The experiments were designed such that the reactions varied significantly in terms of the reaction profile, final conversion and particle size range.
Partial least squares (PLS) regression was used to build models for estimating mean particle size. Different pre-processing methods and wavelength range of measurements were investigated to build the best performing calibration model. This analysis was made using a novel cross-validation approach which ensures that representative samples were used for building models and evaluating their performance. It was found that a combination of the visible-NIR region covering the wavelength range 376-852nm and 1548 nm - 1876 nm provided the best performing models for predicting mean particle size.
| Original language | English |
|---|---|
| Title of host publication | Asia Pacific confederation of Chemical Engineering Congress 2015 (APCChe 2015) incorporating CHEMECA 2015 |
| Place of Publication | Melbourne |
| Publisher | Engineers Australia |
| Pages | 444-449 |
| Number of pages | 6 |
| ISBN (Print) | 9781922107473 |
| Publication status | Published - 2015 |
| Event | Asia Pacific Confederation of Chemical Engineering Congress (2015) - Melbourne, Australia Duration: 27 Sept 2015 → 1 Oct 2015 Conference number: 2015 |
Conference
| Conference | Asia Pacific Confederation of Chemical Engineering Congress (2015) |
|---|---|
| Country/Territory | Australia |
| City | Melbourne |
| Period | 27/09/15 → 1/10/15 |