Soil Type Effects on Petroleum Contamination Characterization Using Ultraviolet Induced Fluorescence Excitation-Emission Matrices (EEMs) and Parallel Factor Analysis (PARAFAC)

Article excerpt

doi:10.1139/S08-037

Introduction

Advances in optical and computational technology have encouraged the development of fluorescence sensors designed to detect petroleum contaminants using their optical properties. When aromatic hydrocarbon molecules, present in petroleum products, are excited with ultraviolet (UV) light, they emit fluorescence. This is termed ultraviolet induced fluorescence (UVIF). For a particular petroleum product, the emitted fluorescence wavelengths vary uniquely in intensity and wavelength as a function of the excitation wavelengths and elapsed time after excitation (Schulman 1977; Vo-Dinh 1982; O'Connor and Phillips 1984). This fluorescing nature of petroleum hydrocarbons has been the focus of numerous research efforts, and reliable devices now exist to perform UVIF contaminant analyses in situ (Lieberman et al. 1991; St. Germain et al. 1993; Lin et al. 1995; Biggar et al. 2003). The major advantage of UVIF analysis is that it does not require any preparation or sampling prior to analysis, which makes it attractive for in situ petroleum contamination screening and characterization.

Previous studies indicated that fluorescence spectroscopy was successful in identifying and estimating the concentrations of different petroleum contaminants such as gasoline, diesel, and jet fuel in soils with limits of detection in a range relevant for the surveillance of regulatory limits or clean-up decisions. However, the soil matrix has a significant effect on the calibration of fluorescence measurements. That effect is dependent on soil properties such as grain size, humidity, and grains optical reflectance (Aptiz et al. 1992, 1993; Knowles and Lieberman 1995; Moise et al. 1995; Nielsen et al. 1995; Roch et al. 1995; Hart et al. 1996; Kotzick and Niessner 1996; Ralston et al. 1996; Schade and Bublitz 1996).

Despite the demonstrated efficiency of in situ fluorescence measurements in characterizing subsurface petroleum contaminants, the capabilities of the method are generally restricted to "presence/absence" screening applications, with a rough qualitative and quantitative interpretation. The limited use of the technology is believed to be due to the lack of an appropriate calibration procedure to give qualitative and quantitative results to in situ fluorescence measurements.

Previous studies suggested using generic correction factors or functions for calibration purposes (Lohmannsrobena and Rocha 1996; Kenny et al. 2000). However, to fully utilize in situ fluorescence measurements, it is essential to establish a reliable calibration procedure using well-characterized laboratory reference petroleum products and soils to model as closely as possible the field conditions. Using the standard reference material will allow examining the influence of soil matrix on the measured fluorescence signal.

This paper presents a new, integrated analytical framework that utilizes fluorescence measurements to characterize and semi-quantify petroleum contaminants in soils. The new method considers the effect of soil properties on fluorescence measurements; in particular soil grain size distribution, porosity as well as mineralogy, and identify the relationship that describes that effect.

Fluorescence spectroscopy

The fluorescing nature of refined petroleum products and crude oil is related to the molecular electron structure of their aromatic compound constituents that allows electrons, when illuminated with sufficient energy (UV light), to absorb the energy (light photons) and be promoted to higher energy levels. The excited electrons could return to the ground state from the higher energy levels through various non-radiative and radiative mechanisms. Fluorescence is a prevalent radiative de-excitation mechanism in aromatic hydrocarbons due to the nature of their electron bonding. The fluorescence signal or spectrum produced by an aromatic hydrocarbon compound is unique because it reflects its electron structure and can be used as a "fingerprint" to identify it. …