168
A. Pérez-Gálvez and J. Fontecha
must be paired with other physicochemical characteristics observed from the chromatographic analysis (retention time) and from the analysis of other spectral features
(mass spectrometry or nuclear magnetic resonance).
One of the main applications of the UV-visible detection coupled to LC is the
quantification of the pigment profile once the identification and acquisition of the
adequate peak isolation have been secured. Quantification is made according to the
external calibration procedure by the analysis of the corresponding standards that are
present in the pigment profile samples. Calibration curves are built for each carotenoid
as the spectrum features, extinction coefficient, and response in the chromatographic
conditions are different, and this is a common source of error in the quantification
methodology. It is frequent that some standards are not commercially available, or
with expensive procedures for the isolation at the lab, so that quantification is made
by using a calibration curve made for a different available carotenoid but resembling
most of the spectral features of the unavailable one. Indeed, carotenoid standards
are not necessarily pure and may contain isomers or impurities. Therefore, they
should be conveniently checked before calibration and even re-purified by thin-layer
chromatography or open-column chromatography (Kimura et al. 2007). The reader
should refer to some laboratory guidelines for validation of the analytical procedures
applied for calibration, quantification, and analytical methodology (U.S. Department
of Health and Human Services 2001; ICH Expert Working Group 2005) and that we
use in our lab. The calibration curves should be linear in the concentration range,
which depends on the intended application, but must include the expected concentrations in the sample, while the correlation coefficients, y-intercepts, slopes, and
residual sum of squares should be noted. The reader must consider that the concentration range of different carotenoids occurring in the same sample is generally wide,
and calibration curves should be built according to this feature, while the statistical
analysis must provide the corresponding significance power for each pigment. The
limit of detection (LOD) is established on the signal-to-noise approach with LOD
between 3 or 2:1 generally considered acceptable, while the limit of quantification
(LOQ) is measured by the same approach and considering a 10:1 LOQ ration as
acceptable.
As it was noted above, some carotenoid standards are commercially available
and even can be obtained upon by-request if the targeted pigment does not appear
in the catalogue of the dealer. Basically, the common carotenoids amarouciaxanthin
A, astaxanthin, canthaxanthin, capsanthin, α-carotene, β-carotene, γ-carotene, βcryptoxanthin, β-apo-8’-carotenal, crocetin, diadinoxanthin, diatoxanthin, fucoxanthin, isofucoxanthinol, lutein, lycopene, neoxanthin, violaxanthin, zeaxanthin are
affordable as they occur in common sources or are reasonably produced by synthesis.
Other less-frequent carotenoids are also available in specialized dealers (www.carote
nature.com), including cis-trans isomers, apocarotenoids, stereoisomers, and esters.
Additionally, some reference materials of mixed pigments are commercially available (www.c14.dhigroup.com) for controlling the performance of the HPLC/UHPLC
method, for identification purposes, and validation of the technique.
A. Pérez-Gálvez and J. Fontecha
must be paired with other physicochemical characteristics observed from the chromatographic analysis (retention time) and from the analysis of other spectral features
(mass spectrometry or nuclear magnetic resonance).
One of the main applications of the UV-visible detection coupled to LC is the
quantification of the pigment profile once the identification and acquisition of the
adequate peak isolation have been secured. Quantification is made according to the
external calibration procedure by the analysis of the corresponding standards that are
present in the pigment profile samples. Calibration curves are built for each carotenoid
as the spectrum features, extinction coefficient, and response in the chromatographic
conditions are different, and this is a common source of error in the quantification
methodology. It is frequent that some standards are not commercially available, or
with expensive procedures for the isolation at the lab, so that quantification is made
by using a calibration curve made for a different available carotenoid but resembling
most of the spectral features of the unavailable one. Indeed, carotenoid standards
are not necessarily pure and may contain isomers or impurities. Therefore, they
should be conveniently checked before calibration and even re-purified by thin-layer
chromatography or open-column chromatography (Kimura et al. 2007). The reader
should refer to some laboratory guidelines for validation of the analytical procedures
applied for calibration, quantification, and analytical methodology (U.S. Department
of Health and Human Services 2001; ICH Expert Working Group 2005) and that we
use in our lab. The calibration curves should be linear in the concentration range,
which depends on the intended application, but must include the expected concentrations in the sample, while the correlation coefficients, y-intercepts, slopes, and
residual sum of squares should be noted. The reader must consider that the concentration range of different carotenoids occurring in the same sample is generally wide,
and calibration curves should be built according to this feature, while the statistical
analysis must provide the corresponding significance power for each pigment. The
limit of detection (LOD) is established on the signal-to-noise approach with LOD
between 3 or 2:1 generally considered acceptable, while the limit of quantification
(LOQ) is measured by the same approach and considering a 10:1 LOQ ration as
acceptable.
As it was noted above, some carotenoid standards are commercially available
and even can be obtained upon by-request if the targeted pigment does not appear
in the catalogue of the dealer. Basically, the common carotenoids amarouciaxanthin
A, astaxanthin, canthaxanthin, capsanthin, α-carotene, β-carotene, γ-carotene, βcryptoxanthin, β-apo-8’-carotenal, crocetin, diadinoxanthin, diatoxanthin, fucoxanthin, isofucoxanthinol, lutein, lycopene, neoxanthin, violaxanthin, zeaxanthin are
affordable as they occur in common sources or are reasonably produced by synthesis.
Other less-frequent carotenoids are also available in specialized dealers (www.carote
nature.com), including cis-trans isomers, apocarotenoids, stereoisomers, and esters.
Additionally, some reference materials of mixed pigments are commercially available (www.c14.dhigroup.com) for controlling the performance of the HPLC/UHPLC
method, for identification purposes, and validation of the technique.