Bioconversion of petroleum hydrocarbons in soil using apple filter cake.

The aim of this study was to investigate the feasibility of using apple filter cake, a fruit-processing waste to enhance the bioremediation of petroleum contaminated soil. A rotating barrel system was used to study the bioconversion of the xenobiotic compound by natural occurring microbial population. The soil had been accidentally polluted with a total petroleum hydrocarbon concentration of 41,000 ppm. Although this global value was maintained during the process, microbial intervention was evidenced through transformation of the petroleum fractions. Thus, fractions that represent a risk for the environment (GRO, Gasoline Range Organics i.e., C6 to C10-12; DRO, Diesel Range Organics i.e., C8-12 to C24-26 and RRO, Residual Range Organics i.e., C25 to C35) were significantly reduced, from 2.95% to 1.39%. On the contrary, heavier weight fraction from C35 plus other organics increased in value from 1.15% to 3.00%. The noticeable diminution of low molecular weight hydrocarbons content and hence environmental risk by the process plus the improvement of the physical characteristics of the soil, are promising results with regard to future application at large scale.

INTRODUCTION

class="Chemical">Oil and class="Chemical">pan class="Chemical">oil distillate spills are frequent contamination events that affect vast land areas, causing damages to the ecology and the economy of rural regions. Thus, the development of ecological, compatible and inexpensive technologies aimed to remediate hydrocarbon contamination, becomes a priority. These techniques have to be designed and tested thoroughly before applying them to a field problem. Usually, hydrocarbon-contaminated soils lack organic matter of botanical origin and have low microbial activity. The addition of a fruit-processing waste can be beneficial since it represents a source of co-substrates, nutrients and microorganisms and ameliorates the structure and water-retaining capacity of the soil in favour of aerobic conditions. The incorporation of exogenous microorganisms might accelerate the degradation of contaminants in soil (27). Aerobic metabolism generates heat, resulting in significant temperature increases caused by changes in the microbial population and physiology in the mixture. Temperature affects not only the physiological reaction rates and population dynamics of microbes, but also most of the physicochemical characteristics of the environment (2,3).

The selection of waste substrates involves the difficulty of finding a residue with the right balance of nutrients to support optimal growth and production (21). For the experiment of this study, we proposed the utilization of class="Species">apple filter cake for the bioremediation of class="Chemical">petroleum contaminated sclass="Chemical">pan class="Chemical">oil because it is an industrial waste of the impacted zone. Although this material may have other industrial uses, large amounts are generated, what makes it easily available. Transportation distances from apple manufactory to soil treatment areas are short and this lowers its cost. Even though addition of organic material to contaminated soil is not a new concept for improving its remediation (20,22), the use of local waste materials in Argentina has not often been tested before. Since the use of local available organic wastes appears to be a promising cheap alternative to accelerate TPH decomposition in contaminated soil, the success of this approach is not only important locally, but it could be applied in other regions and with the use of other wastes. The aim of this study was thus to investigate the feasibility of employing apple filter cake to enhance the bioremediation of a soil accidentally polluted with petroleum using a rotating barrel system. This preliminary test would help to determine the operational requirements of the system in order to attain target end-point concentrations and its practical application on large-scale treatment.

MATERIALS AND METHODS

Experimental Design

A rotating barrel of 200 litres capacity was specially constructed for this experiment. A 0.5 mm (pore size) net set at the bottom of the barrel defined a compartment for possible leachates collection added with a purge valve. The composition of the mixture consisted of 70 kg of contaminated sclass="Chemical">oil and 30 kg of class="Chemical">pan class="Species">apple filter cake. The soil was loamy sand with a residual concentration of 41,000 ppm of TPH (Total Petroleum Hydrocarbons). It was air dried and sieved with a 2 mm grid. Hydrocarbon (HC) fractions extractable with dichloromethane from dry soil were: GRO: 0.13% (Gasoline Range Organics; i.e., C6 to C10–12); DRO: 2.59% (Diesel Range Organics; i.e., C8–12 to C24–26); RRO: 1.38% (Residual Range Organics; comprising 0.23% from C25 up to C35 and 1.15% compounds of more than C35 and other organics). Aromatic compounds as benzene, toluene, ethylbenzene, xilene (BTEX), trimethylbenzene, methylethylbenzene, propylbenzene and naphthalene (aromatics up to C10) were determined individually. All of them were below the detection limit of 0.01ppm. The content of volatile compounds shorter than C10 was 20ppm.

Chemical analysis indicated that total class="Chemical">nitrogen content of the sclass="Chemical">pan class="Chemical">oil was 300 ppm, phosphorous content was 6.2 ppm, 6.58% (wt/wt) of moisture, saturation point 35% and pH = 7.2. The concentration of heavy metals was under the guideline values (13).

The class="Species">apple filter cake was obtained from a cider factory of Tunuyán, Mendoza, Argentina. Its humidity content was 85%. The initial class="Chemical">pH value of 4 was modified by the addition of 3.33% (wt./wt.) of class="Chemical">pan class="Chemical">Ca(OH)2 to reach a value of pH = 7. To estimate the HC-like contribution to the mixture made by the apple filter cake, it was treated identically as the soil for TPH extraction. The extractable compounds with dichloromethane from the dry material expressed as hydrocarbon fractions were: GRO: 0.00%; DRO: 16.00%; RRO: 43.00% and 41.00% of C35 and other organic compounds.

The extractable compounds with class="Chemical">dichloromethane from the mixture exclass="Chemical">pressed as class="Chemical">pan class="Chemical">TPH were 40,000 ppm (4% wt/wt.). The expression as hydrocarbon fractions was: GRO: 0.04%; DRO: 1.97%; RRO: 0.32% C25 up to C35 and 1.67% C35 and other organic compounds.

class="Chemical">BTEX and class="Chemical">pan class="Chemical">hydrocarbons shorter than C10 were under detection limit of 0.01ppm. An initial C:N:P ratio of 100:5:1 was selected based on previous studies (9). For this purpose, 12g of urea dissolved in distilled water were added in three fractions every ten days during the first 30 days. The fractionated addition of urea was done to avoid partial loss of nitrogen and fungicidal effects of ammonia gas product of the microorganisms urease conversion over urea (11,12). Control reactor consisted of 70 kg of 100% contaminated soil placed in a bin. The aeration by rotation of the barrel was daily during the first 45 days of treatment to assure oxygen provision and then on days 75, 90 and 210. Subsequently, the mixture was allowed to mature without rotation up to day 360.

Systeclass="Disease">ms and environmental temclass="Chemical">perature, humidity content, class="Chemical">pH and Aerobic class="Chemical">pan class="Chemical">hydrocarbon degrader microorganisms were measured every two days up to day 14, on a weekly regime up to day 45 and then on days 135, 210 and 360 of the process.

pan class="Chemical">Petroleum hydrocarbon analysis were class="Chemical">performed on samclass="Chemical">ples of the initial comclass="Chemical">ponents individually and on both systeclass="Chemical">pan class="Disease">ms taken on days 0, 45, 75, 90, 210 and 360 of the process.

An average sample of 200 g was obtained by mixing five aliquots of approximately 40 g extracted from different points of the reactor. Collected and mixed, samples were dried at room temperature for 24 h and then sieved (2 mm diameter openings). All analytical determinations were performed in duplicate.

ANALYTICAL METHODS

For every sample, a 10 cc aliquot was mixed with 50 ml of distilled class="Chemical">water for five minutes before the class="Chemical">pH was measured with an electrode. Aclass="Chemical">pclass="Chemical">proximately 10g of samclass="Chemical">ple was class="Chemical">placed in a class="Chemical">preweighed class="Chemical">porcelain caclass="Chemical">psule and dried at 105ºC for 24h. The caclass="Chemical">psule with samclass="Chemical">ple was weighed before and after drying to obtain the class="Chemical">percent humidity mass. Distilled class="Chemical">pan class="Chemical">water was stirred with soil and brought to saturation. Subsequently, the saturation point was determined as percent humidity mass. Total nitrogen was determined following a micro-Kjeldahl analysis (5) with minor modification for soils. Absorbed phosphorous was determined according to Bray & Kurtz (8). Total HC were extracted with dichloromethane by Soxhlet to conduct further detailed chemical analyses. A 20g sample was mixed with 5 g Na2SO4 and 10 g MgSO4. Once all the residual moisture had been absorbed by the MgSO4, the mixture was pulverized and placed in a paper extraction thimble. The top of the thimble was filled with glass wool. The HC were extracted at a rate of 20 cycles h-1 for 4.5 h. Fifteen grams of silica gel were added to the extraction solvent in a stoppered container and was mixed continuously for 10 min. The solvent was then filtered through glass wool, placed in a clean pre-weighed porcelain capsule and evaporated. The capsule was dried at room temperature for 24 h and weighed to obtain the mass of total extract by gravimetry (GR). A simulated distillation of the extract was performed with the aim of determining the amount of residue at 320ºC. This fraction is comprised of those compounds that remain in the chromatographic injector at 320ºC. Subsequently, gas-liquid chromatography analyses with flame ionization detection GC-FID were performed using a Clarus 500 Perkin Elmer chromatograph fitted with a moderately polar 30 m column (Perkin Elmer). The column temperature was programmed at an initial temperature of 45ºC for 5 min, then increased at 10ºC min-1 to 300ºC and held there for 40 min. Carlo Erba brand calibrators were used as markers for linear hydrocarbons from C6 to C20. ASTM D7169–5 norm (7) was followed as reference.

class="Chemical">BTEX and class="Chemical">pan class="Chemical">benzene isomers were determined using a Clarus 500 Perkin Elmer gas-chromatograph equipped with a mass selective detector (GC/MS) and automatic Headspace sampler. A 30 m moderately capillary column HP5 type was used working isothermically at 50ºC. Specific ions were used to quantify the aromatic compounds up to C9. EPA Method 8260 B SW846 norm (18) was followed as reference.

Culturable class="Chemical">hydrocarbon degrading bacteria were enumerated by standard class="Chemical">plate technique with aclass="Chemical">pclass="Chemical">proclass="Chemical">priate mineral media (12) using 10% (vol/vol.) class="Chemical">pan class="Chemical">diesel fuel as the sole carbon source.

RESULTS AND DISCUSSION

Evolution of the operational variables of the system

The control of temperature was of great concern for our experiment, considering it is a fundamental parameter for control of the activity of the biological process.

Fig. 1 shows the temperature variation of the mixture sclass="Chemical">oil-class="Chemical">pan class="Species">apple filter cake inside the reactor and room temperature during the length of the experiment. An important difference can be appreciated during the first 23 days, which correlated with a change in the physical structure of the mixture, due to the fast degradation of apple filter cake. A strategy of temperature control between 40ºC and 45ºC would seem to be appropriate for the bioremediation of PAH-contaminated wastes since it allows for the maximum microbial diversity to be expressed for both, bacteria, actinomycetes, and fungi (4). On this base and considering the increase in the microbial population, we can accept the temperature range reached of 32 ºC to 36 ºC for eleven days as acceptable for the benefit of our objectives.

Figure 1

Inside reactor and environmental temperature evolution during the biological process.

Fig. 2 shows that in spite of the changes of temperature, the class="Chemical">hydrocarbon degrader microbial class="Chemical">poclass="Chemical">pulation stayed in the order of 108 cfu g-1 sclass="Chemical">pan class="Chemical">oil representing an extension of the active phase of the process up to day 40. On day 210 descended to the order of 106 cfu g-1 soil, which is still a significant quantity of active HC degrader biomass.

Figure 2

Numbers of Hydrocarbon-degrading microorganisms during treatment.

Numbers of pan class="Chemical">Hydrocarbon-degrading microorganisclass="Chemical">pan class="Disease">ms during treatment.

It is known that the content of class="Chemical">cellulose, class="Chemical">pan class="Chemical">hemicellulose, lignin and proteins in crop residues influences their own initial decomposition rate. In our system, the apple filter cake exerted a positive effect that increased the microbial population by two orders of magnitude. The reasons of this phenomenon could be that apple filter cake acts as a nutrient supplier and physical conditioner contributing to increased oxygen diffusion and mineral nutrient availability, as well as carbon source and mechanical support surface for bacterial adsorption. A noticeable change in soil structure, aspect and odour, alike a garden soil was observed after treatment compared to control soil. The improvement of soil physicochemical characteristics probably accelerated microbial adaptation and selection and provided additional microorganisms, which contributed to metabolise or maybe cometabolise contaminant hydrocarbons. Furthermore, the contribution of organic matter to soil is generally accepted as a fundamental factor of fertility because its contribution with essential nutrients and improvement of physical conditions (10,24). The decrease of the biological parameters after day 40 could probably be attributed to a reduction in nutrients and carbon source bioavailability, to hydrocarbons adsorption to apple filter cake material and soil or to toxic effects produced by residual petroleum fractions (1,16,19).

Moisture content of the mixture remained between 16.28% and 17.12% up to day 45, which corresponds to approximately 45% to 50% saturation. The saturation point of the sclass="Chemical">oil was determined to be aclass="Chemical">pclass="Chemical">proximately 35% moisture by weight of sclass="Chemical">pan class="Chemical">oil. This is consistent with the literature which suggests optimum moisture content for biodegradation of hydrocarbons in soil of approximately 40 to 48 percent of soil saturation (23,25). Nevertheless, our data differed from the majority of previous studies that recommended the maintenance from 30% to 80% (3,9,15,19). The difference in optimal moisture content among these studies may be attributed mainly to the scale of work and thus the environmental conditions prevailing in the reactors. Fresh apple filter cake had a moisture level of 86% and the resultant level in the composting mixture with soil, was enough to avoid compaction and leachates formation. Thus, this moisture range let to the maintenance of aerobic conditions. Besides, with the course of the process, apple filter cake physical characteristics changed becoming a good bulking agent, what further improved the aeration of the process.

pan class="Chemical">Hydrocarbon mineralization is favoured by near neutral class="Chemical">pH values and where it has been shifted away by manmade changes, biodegradability is likely to be imclass="Chemical">paired (17). The addition of class="Chemical">pan class="Chemical">Ca(OH)2 was efficient to maintain pH values between 7.2 and 7.6 during the complete process, providing a suitable media for microbial growth.

HYDROCARBON ANALYSIS

In the experimental reactor a slight difference among the total pan class="Chemical">hydrocarbon content along the class="Chemical">process (41,000 class="Chemical">pclass="Chemical">pm in the initial contaminated sclass="Chemical">pan class="Chemical">oil and control reactor, 39,000 ppm for the mixture at day 75 and 42,000 ppm at the end of the process) was observed, which could be attributed to sampling and analytical variations. Under these considerations, 40,000 ppm is considered constant during the process (Table 1).

Table 1

Composition of chromatographic fractions of dichloromethane extractable hydrocarbons.

Components	Dry soil Day 0	Dry mixture Day 45	Dry mixture Day 90	Dry mixture Day 90	Dry mixture Day 90	Control reactor Day 360
GRO (%)	0.13	0.04	0.02	0.01	0.01	0.12
DRO (%)	2.59	1.97	0.96	0.63	0.35	2.59
RRO up to C35 (%)	0.23	0.32	1.16	0.75	0.80	0.25
RRO sup to C35 and others (%)	1.15	1.67	1.86	2.51	3.00	1.14

Composition of chromatographic fractions of pan class="Chemical">dichloromethane extractable class="Chemical">pan class="Chemical">hydrocarbons.

A diminution of the shorter and the appearance of longer chain compounds can be observed. This phenomenon is also evident in the physical aspect of the Soxhlet extracts. Those obtained at initial stages of the process showed aspect and behavior of pan class="Chemical">oil while the extracts obtained at later stages demonstrated a class="Chemical">proclass="Chemical">portional increase in chain length behaving as a hard wax.

Based on these results a transformation of low molecular weight (GRO and DRO) into heavier class="Chemical">paraffin molecules is demonstrated. class="Chemical">pan class="Species">Apple filter cake has light paraffin content of low molecular weight and an important paraffin content of high molecular weight (greater than C28). These paraffins contributed by apple filter cake to the mixture are of botanical origin and with the evolution of the process, its concentration increases to expenses of the natural humification process and of the fractions GRO and DRO that practically disappear. This phenomenon of residues formation could be interpreted as a result of the microbial activity by incorporation of metabolites into the organic soil matrix. With this hypothesis we reinforce that of Kästner et al. (14) who worked with (14C) anthracene and demonstrated that the microbial activity produces covalently bound metabolites that become integral parts of humic matter which cannot be further identified.

Low molecular weight class="Chemical">hydrocarbons are the main health risk factors in class="Chemical">pan class="Chemical">petroleum hydrocarbon mixtures (6). In this experiment, BTEX and other light hydrocarbons up to C10 were under detection limits before and after the process. To evaluate the performance of the process, the analysis was focused on lighter paraffin chromatographic fractions, what showed a diminution of hydrocarbons from 2.95% (GRO, DRO and RRO up to C35) to 1.39% in 90 days of treatment and to 1.16% after the maturation phase. The noticeable diminution of low molecular weight hydrocarbons content by the composting process encourages us to consider the technique as a valuable tool for bioremediation.

The success of the composting process depends not only on decreasing chemical concentrations, but also on the sclass="Chemical">oil´s suitability to suclass="Chemical">pclass="Chemical">port the normal sclass="Chemical">pan class="Chemical">oil fauna and flora (14). In this sence, ecotoxicity tests on the product would be relevant to prove the productivity of the soil after the bioremediation treatment.

As a bioremediation alternative, the utilization of class="Species">apple filter cake was successful considering that the class="Chemical">pan class="Chemical">hydrocarbon fractions of environmental concern were noticeable reduced. Thus, its use of for amelioration of hydrocarbon contaminants promotes soil sustainability and re-use.