1883 Environmental Toxicology and Chemistry, Vol. 15, No. 11, pp. 1883–1893, 1996 Printed in the USA 0730-7268/96 $6.00 + .00 EFFECTS OF SUBSTRATE MINERALOGY ON THE BIODEGRADABILITY OF FUEL COMPONENTS SABINE E. APITZ*² and K ATHLEEN J. MEYERS-SCHULTE‡ ²Remediation Research Laboratory, Naval Command, Control and Ocean Surveillance Center, Research Development Testing and Evaluation Division 361, San Diego, California 92152, USA ‡Computer Sciences Corporation, 4045 Hancock Street, San Diego, California 92110, USA (Received 28 September 1995; Accepted 1 April 1996) Abstract—Experiments were carried out to determine the effects of mineralogy on the biodegradability of components of a whole fuel by a soil microbial consortium. Samples of quartz sand (Fischer Sea Sand) and illite clay (API 35) were spiked with marine diesel fuel, aged, slurried, and inoculated, and concentrations of fuel components were monitored over time. To help distinguish biotic from abiotic processes, identical samples were poisoned with mercuric chloride and were run in parallel. While there was chromatographic and biomarker evidence of n-alkane biodegradation in the sand samples, illite samples showed no evidence of biogenic loss of aliphatic components. Polycyclic aromatic hydrocarbons, on the other hand, were lost equivalently on both minerals and in both cases were lost to a much greater extent than were total petroleum hydrocarbons (TPHs). These results suggest that under our experimental conditions, illite inhibited the bioavailability of some TPH components to the soil microbial consortium. Keywords—Biodegradability Mineralogical effects n-Alkanes Polycyclic aromatic hydrocarbons INTRODUCTION Petroleum-contaminated soils are highly complex systems. First, soils can be composed of naturally occurring organic mat- ter and one or many minerals, exhibiting varying properties such as surface chemistry, grain size, and porosity. Second, petroleum products are mixtures of hundreds of aliphatic and aromatic organic compounds, the relative proportions of which vary greatly between fuel type and somewhat between batches of the same type. Each component differs in its reactivity, sol- ubility, volatility, mineral surface affinity, and biodegradability [1–4]. Furthermore, the mode of introduction of the contaminant into the soil or sediment, postdepositional weathering, and di- verse mobility characteristics of the components can drastically alter the composition of the bulk fuel contaminant. However, little work has been done to examine mineral-specific effects on bioavailability, mobility, and degradability of fuel compo- nents, which are important issues from a remediation standpoint. Remediation technology is often developed and demonstrated on soils composed of pure, simple, and relatively inert quartz sand [5–7]. However, the bulk of natural soils are actually com- plex mixtures of materials, including many reactive and high- surface-area components such as clay minerals, organic matter, and metal sulfides and/or oxyhydroxides. A key factor in both the remediation and bioavailability of the most recalcitrant frac- tions of hazardous organic waste is the long-term sorption that can occur between organic molecules and clays or other minerals in soils and sediments [8,9]. Organic-mineral binding mecha- nisms are poorly understood [10], and the fundamental chem- istry controlling these important issues in natural systems must be determined before it will be possible to develop an intelligent approach to remediation of contaminated sites. We have shown, * To whom correspondence may be addressed. Presented at the Department of Defense Tri-Service Workshop on Bioavailability of Organic Contaminants in Soils and Sediments, Mon- terey, California, USA, April 9–12, 1995. for instance, that the rate and nature of interaction and weath- ering [11] of fuels on pure clays and pure sands are funda- mentally different, probably because of different mineral/con- taminant interactions taking place in the clays and sands. The purpose of this experiment was to take a first look at the effects of substrate mineralogy on the biodegradability of fuel components. In order to achieve this, representative ‘‘end-mem- ber’’ minerals, a quartz sand and an illite clay, were chosen. MATERIALS AND METHODS Single-component soils, Fischer Sea Sand and illite clay (API 35 [12–14]; see below for a discussion of clay organic content), were lightly ground and sieved to the same size class, 25 to 140 mesh (710–104 m). Sand and clay were weighed in 5-g aliquots into vials, and then marine diesel fuel (DFM, Navy Fuel Depot, San Diego, CA, USA) was added to yield concen- trations of 0 to 35 mg DFM/g dry substrate. Prepared samples were then placed on a Cole-Parmer Roto-Torque for 2 weeks. The samples were used as standards for a fluorescence exper- iment [11] and were then stored in sealed vials, which were not airtight, for 2 years. After 2 years, individual 5-g samples were combined (all the sands together and all the clays together) and homogenized by mixing in a lapidary tumbler. Homogeneity of the combined samples was confirmed by bulk fluorescent mea- surements of subsamples of the materials. Concentrations after homogenization were 8.5  0.2 mg/g and 11.5  0.2 mg/g for sand and clay, respectively, as determined by gas chromatog- raphy with flame ionization detection (GC–FID) analysis of extracts (see method description below). Clearly, these samples were not fully representative of field- contaminated samples since they were not subjected to envi- ronmental conditions present in natural systems. Yet, there are advantages to using spiked samples for laboratory investiga- tions. Because extractions of well-aged, field-contaminated sam- ples are never quantitative, it is impossible to be certain what was in the samples to start with. With spiked samples, accurate