ANALYSIS OF OILS USING FLUORESCENCE.
19
The ratio of the 535-750 nm flux to the 430-535 nm flux of fluorescence emission
spectra using 365 nm excitation was also used as a parameter to characterise crude oils
but no assessment of its quantitative accuracy was provided.
28, 70 In general, there are
several practical disadvantages in using fluorescence intensity based measurements for
oil characterisation particularly for remote sensing applications. The main problem lies
in the difficulty in the accurate, reproducible measurement of absolute fluorescence
emission intensity.
† These measurements can be affected by instabilities in the excitation
source, detector electronics, sample turbidity, and photobleaching.
61 ,
62
4.2. Time-Resolved
Time-resolved fluorescence measurements are largely insensitive to the negative
factors that affect steady-state measurements and are more easily referenced, making
sensing applications more robust.
63 Time resolved fluorescence techniques,
64 which have
also been employed for characterization of petroleum products, are not only largely free
of these artefacts,
22, 25
but in addition contain information that is lost in the timeaveraging process inherent in steady-state methods. In the case of crude oils, which are
complex mixtures of fluorophores, time-resolved fluorescence measurements offer the
best approach for fully revealing the influence of quenching and energy transfer
processes on fluorescence behaviour.
In 1987, the fluorescence lifetimes measured at a range of different emission
wavelengths for a selection of Kuwaiti and Australian oils were measured using a 337.1
nm excitation source.
11 Regrettably, no chemical compositional data was supplied for the
oils, and the reported lifetime data is incomplete (no fractional intensities supplied). A
photophysical model for the fluorescence lifetime behaviour of crude petroleum oils
detailed the complex balance between energy transfer and quenching that governs the
fluorescence lifetimes of crude oils was derived from dilution studies on several oils.
25
This study used a variety of different excitation sources but did not try to correlate
chemical composition with fluorescence lifetime data.
More recently, we have revisited the use of intensity average fluorescence lifetime
for the characterisation of crude petroleum oils and have sought to correlate fluorescence
lifetime data with physical characteristics (API gravity) and compositional factors such
alkane, aromatic, polar, and sulphur concentration.
21, 22 Figure 14 shows the plot of
lifetime versus alkane concentration (calculated 3 different ways) for a series of 23 oils.
The oils studied are topped oils, where the light hydrocarbon fraction was removed, and
they encompassed a wide range of sources and compositions.
10, 32
Using a longer
excitation wavelength (405 nm) resulted in a slightly better correlation for both corrected
alkane (Figure 15) and polar concentration (Figure 16).
32 In each case, however, there is
still considerable scatter about the best-fit line, preventing the development of accurate,
quantitative models.
† Fluorescence ratio methods like the red-green quotient and the QFT-II avoid the problem of absolute
emission intensity measurements.
19
The ratio of the 535-750 nm flux to the 430-535 nm flux of fluorescence emission
spectra using 365 nm excitation was also used as a parameter to characterise crude oils
but no assessment of its quantitative accuracy was provided.
28, 70 In general, there are
several practical disadvantages in using fluorescence intensity based measurements for
oil characterisation particularly for remote sensing applications. The main problem lies
in the difficulty in the accurate, reproducible measurement of absolute fluorescence
emission intensity.
† These measurements can be affected by instabilities in the excitation
source, detector electronics, sample turbidity, and photobleaching.
61 ,
62
4.2. Time-Resolved
Time-resolved fluorescence measurements are largely insensitive to the negative
factors that affect steady-state measurements and are more easily referenced, making
sensing applications more robust.
63 Time resolved fluorescence techniques,
64 which have
also been employed for characterization of petroleum products, are not only largely free
of these artefacts,
22, 25
but in addition contain information that is lost in the timeaveraging process inherent in steady-state methods. In the case of crude oils, which are
complex mixtures of fluorophores, time-resolved fluorescence measurements offer the
best approach for fully revealing the influence of quenching and energy transfer
processes on fluorescence behaviour.
In 1987, the fluorescence lifetimes measured at a range of different emission
wavelengths for a selection of Kuwaiti and Australian oils were measured using a 337.1
nm excitation source.
11 Regrettably, no chemical compositional data was supplied for the
oils, and the reported lifetime data is incomplete (no fractional intensities supplied). A
photophysical model for the fluorescence lifetime behaviour of crude petroleum oils
detailed the complex balance between energy transfer and quenching that governs the
fluorescence lifetimes of crude oils was derived from dilution studies on several oils.
25
This study used a variety of different excitation sources but did not try to correlate
chemical composition with fluorescence lifetime data.
More recently, we have revisited the use of intensity average fluorescence lifetime
for the characterisation of crude petroleum oils and have sought to correlate fluorescence
lifetime data with physical characteristics (API gravity) and compositional factors such
alkane, aromatic, polar, and sulphur concentration.
21, 22 Figure 14 shows the plot of
lifetime versus alkane concentration (calculated 3 different ways) for a series of 23 oils.
The oils studied are topped oils, where the light hydrocarbon fraction was removed, and
they encompassed a wide range of sources and compositions.
10, 32
Using a longer
excitation wavelength (405 nm) resulted in a slightly better correlation for both corrected
alkane (Figure 15) and polar concentration (Figure 16).
32 In each case, however, there is
still considerable scatter about the best-fit line, preventing the development of accurate,
quantitative models.
† Fluorescence ratio methods like the red-green quotient and the QFT-II avoid the problem of absolute
emission intensity measurements.