S. Kokot et al., DISCRIMINATION OF REACTIVE DYES ON COTTON FABRIC BY RAMAN-SPECTROSCOPY AND CHEMOMETRICS, Applied spectroscopy, 51(3), 1997, pp. 387-395
FT-Raman spectra were obtained from undyed poplin cotton fabric and fr
om the same fabric differently dyed with a bi-functional reactive dye,
Cibacron C (molecular structure unknown); the four series of the dyed
samples each contained the dye in a different form--unfixed, ammonia-
treated/unfixed, fixed, and ammonia-treated/fixed. The spectra were do
minated by the dye, but the different states of the dye were not obvio
usly differentiated. Application of principal component analysis showe
d that the spectral groups of the four different dye states can be dis
criminated from each other and from that of the undyed cotton. Further
, for each series of the dyed fabrics, which contain samples with diff
erent amounts of dye, the individual dye concentration subgroups are d
istinguished. Exploratory quantitative studies suggested that FT-Raman
spectroscopy may be a suitable quantitative method for the prediction
of % concentration of the unfixed dye on cotton fabrics. A comparison
of the FT-Raman results with those from a parallel FT-IR study report
ed elsewhere indicated that similar qualitative conclusions may be rea
ched with both techniques. However, the FT-Raman approach does provide
additional information from the dominating dye spectrum. A comparison
of prediction of % concentration of the unfixed dye on the fabric ind
icated that better partial least-squares (PLS) calibration models may
be obtained from the FT-Raman data, but the validation results from a
small set of samples suggested only a marginal advantage with the use
of the Raman approach.