Supplément 2.20: Absorbance and Fluorescence of Oils (2/2)

Method of Absorbance Measurements

The absorbance of oils is an important auxiliary parameter in many fluorometric investigations. The absorbance of many chemicals can be easily determined with conventional spectrophotometers using standard 1 cm x 1 cm sample cuvettes. However, crude oils have a comparably much higher absorbance and make it necessary to use cuvettes with an optical path length of 1 to 10 μm.

Alternatively, the sample can be diluted in an organic solvent down to a concentration level which ensures transmission of light over distances of a few millimetres. Experiments have shown that cyclohexane is most suitable for this purpose: other types of solvents are not suitable to bring all compounds of crude oil samples into solution. In this way the use of optical setups for analysis of highly absorbing liquids, that are not generally available, is avoided and absorption measurements can be done with standard cuvettes and laboratory photometers.

To cover the entire wavelength range of optical remote sensing, absorbance measurements were made at wavelengths between 250 and 700 nm with 1 nm resolution. The absorbance of the solvent utilised for diluting the sample is discriminated by a double beam measurement, with the solvent in the reference cuvette

An example: Measuring with the Kontron UVIKON 810 Spectrophotometer

Absorbance data measured with this instrument were obtained by using quartz cuvettes with a 1 cm optical path length. A deuterium lamp is used between 250 and 340 nm, a tungsten halogene lamp at higher wavelengths. Data output is given in Absorbance Units (ABS) which corresponds to

where $I_o$ is the incident light intensity and $I$ is the intensity transmitted through the sample cuvette. The dynamic range is from -0.3 to 4 ABS, with an accuracy specified to 0.002 ABS at 1.0 ABS.

ABS data were converted to absorption coefficients a in Lambert's absorption law

with the cuvette length d, see question What is the thickness of oil films? in supplement 2.19, page 3. Inverting the equation yields:

With the transformation

it follows:

Oil samples were diluted with cyclohexane down to a concentration that yields data within the dynamical range of the instrument. To avoid possible errors due to interactions with the solvent a series of dilutions is made with each sample. From these data an average over the entire linear range is calculated, and absorption coefficients a are derived from this average.

Method of Fluorescence Measurements

Data included in the catalogue were measured with conventional spectrofluorometers, which are a standard equipment in most chemical laboratories. These instruments meet the requirements of providing spectra which are applied in hydrographic research and survey, i.e. outside the 'clean' environment of an optical laboratory. The high spectral resolution of laser spectrometers, which makes these instruments very useful for many purposes, is not required here.

Each substance sample is analysed with the wavelength parameters given in the table below. Excitation wavelengths are selected according to emission lines of spectroscopic lamps or lasers which are prominently applied for such purposes. Emission is registered with a one nanometre wavelength increment.

The geometrical orientation of the sample cuvette with respect to the excitation and emission ray path differs from the conventional 90 degree configuration. Because of the high absorbance of samples, the cuvette is oriented such that fluorescence is excited and registered through a single side of the cuvette, with a cuvette holder termed front surface assembly by some manufacturers. This method allows a direct analysis of highly absorbing liquids, avoiding a dilution of samples with organic solvents that might by themselves contribute to the fluorescence signal.

In this configuration, however, particular attention has to be paid to the quality of the cuvette itself: the best available fused silica is required, otherwise fluorescence contributions from the cuvette window obscure the sample fluorescence. For the same reason cuvettes need to be cleaned very carefully following each filling with highly fluorescent samples.

Basic parameters of fluorescence measurements

Wavelengths are given in nanometres.

A crucial requirement for measuring spectra in a reliable way is to perform a careful calibration of the instrument both on the excitation and the emission side of the optical setup. The calibration procedure for a Perkin Elmer spectrofluorometer, as described in detail below, enables us to derive data that are corrected for the spectral sensitivity of the instrument, and allows signals originating from different substances to be related to each other on a relative scale. In principle, each commercially available spectrofluorometer can be calibrated in this or a similar way, and a procedure following these guidelines is mandatory for spectral data to be included in this catalogue.

Absolute calibration of spectra in cross-sectional units like, e.g., cm²/sr is hampered by an apparent lack of standard reference substances which are applicable for this purpose. Therefore, one of the oil samples investigated here, Nigerian Light Crude Oil, is used as the reference material. Its fluorescence emission, integrated from 320 to 699 nm, and excited with 308 nm wavelength, is arbitrarily set to have an efficiency of one. Spectra of other substances, and also obtained with other excitation wavelengths, are given in ratio to this unit. To enable the user to quantitatively interprete these spectra, and to ensure compatibility of data obtained by different laboratories in future, a sample of Nigerian Light Crude can be made available upon request.

An example: Measuring with the Perkin Elmer Model 650-40 Sectrofluorometer

The front surface assembly equipped with 1 cm standard silica cuvettes was applied with these highly absorbing samples. The instrument was run with the parameters listed in the Table above, and with a 60 nm/min wavelength drive and 1 sec integration time. All measurements were performed at room temperature (20°C). The instrument is equipped with a Xenon DC lamp as the light source. The emission spectrum and the stability of lamp and excitation ray path are controlled daily prior to and following each series of measurements. This is done by means of a fluorescence screen of 0.02 mol/L 1-dimethylaminonaphtalene 5-sulfonate in a solution of NaOH in water. The solution is practically opaque at wavelengths below 350 nm and only slightly absorbing above 450 nm. Hence it follows that, with a constant setting to 500 nm of the detection monochromator, and the excitation monochromator scanning between 250 and 450 nm, the emission of this sample yields a direct measure of the spectral intensity of the excitation path in this wavelength range. Since the solution is highly absorbing, the front surface assembly is required.

The same result can also be obtained with the organic dye Rhodamine B. Its fluorescence band peaking at 600 nm, this material allows an extension of the calibration curve up to the red portion of the spectrum. For this, a concentrated solution of Rhodamine B (3 mg/L in ethylene glycol) is used. The excitation monochromator is scanned from 200 to 600 nm, while the emission monochromator is set to 615 nm.

The stability of the detection system is controlled daily, using a 1 ppm solution of quinine sulfate dihydrate in 0.1 mol/L HClO₄ as a fluorescence standard (Velapoldi & Mielenz, 1980). In this way, the long-term stability of the detector setup is found to be better than 2% over a period of 1 month.

Fluorescence emission of quinine sulfate covers a wavelength range of 400 to 550 nm (~10% intensity points of the maximum intensity at 450 nm), and it can therefore be used as a calibration material in this region of the spectrum. For the purpose of calibrating the emission path of the instrument at wavelengths below and above this range other means must be applied. A broad-band calibration is preferably done with a tungsten standard lamp with known brightness temperature. In these measurements an OSRAM WI 9 standard lamp with T_B=2856 K. A black tubus in the ray path between lamp and detector in front of its entrance aperture serves to eliminate possible stray light contributions from the signal. The optical axis of the detection path of rays is controlled with the output beam of a HeNe laser. Position of the tungsten standard is at 5 m distance from the spectrofluorometer. With the OSRAM WI 9 lamp utilised here a calibration down to 350 nm wavelength can be achieved. A further extension down to the UV can be done with a tungsten halogene lamp because of its higher filament temperature and the bulb made of quartz. The emission spectrum of this lamp, if not calibrated, is determined by comparison with the calibrated tungsten standard. Brightness temperatures of 3200 K typically allow its use as a calibration source down to 300 nm or even lower wavelengths.