Abstract
An ultraviolet visible (UV-Vis) spectrophotometric and partial least squares (PLS) chemometric method was developed for the simultaneous determination of erythrosine B (red), Brilliant Blue, and tartrazine (yellow) dyes. A training set (n = 64) was generated using a full factorial design and its accuracy was tested in a test set (n = 13) using a Box–Behnken design. The test set garnered a root mean square error (RMSE) of 1.79 × 10–7 for blue, 4.59 × 10–7 for red, and 1.13 × 10–6 for yellow dyes. The relatively small RMSE suggests only a small difference between predicted versus measured concentrations, demonstrating the accuracy of our model. The relative error of prediction (REP) for the test set were 11.73%, 19.52%, 19.38%, for blue, red, and yellow dyes, respectively. A comparable overlay between the actual candy samples and their replicated synthetic spectra were also obtained indicating the model as a potentially accurate method for determining concentrations of dyes in food samples.
| Original language | English |
|---|---|
| Pages (from-to) | 1633-1639 |
| Number of pages | 7 |
| Journal | Applied Spectroscopy |
| Volume | 71 |
| Issue number | 7 |
| DOIs | |
| State | Published - Jul 1 2017 |
Keywords
- Brilliant Blue
- chemometrics
- erythrosine B
- food dye
- partial least squares
- PLS
- Tartrazine
Fingerprint
Dive into the research topics of 'Development of a Direct Spectrophotometric and Chemometric Method for Determining Food Dye Concentrations'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver