Biotreatment of waste gas containing pyridine in a biofilter

Industrial waste gas emissions containing pyridine are generated from pyridine manufacturing industries, and in industrial operations where pyridine is used as a solvent, as an intermediate for synthesis and as a catalyst for a variety of applications. Pyridine has unpleasant fishy odor with an odor index of 2390 and waste gaseous emissions containing pyridine require proper treatment prior to discharge. A biofilter, packed with compost and wood chips and inoculated with Pseudomonas pseudoalcaligenes-KPN for enrichment of pyridine-degrading microorganisms, was operated on a continuous feed basis for a period of more than 2 years. The results indicate that the biofilter medium with optimal moisture content of 68% and an effective bed retention time (EBRT) of 28.50s could degrade pyridine effectively (>99%) at a loading of 434 g pyridine m(-3)h(-1). The treated waste gas was also found to be free from pyridine odor.     

Macro-kinetic investigation on phenol uptake from air by biofiltration: Influence of superficial gas flow rate and inlet pollutant concentration

The macro-kinetic behavior of phenol removal from a synthetic exhaust gas was investigated theoretically as well as experimentally by means of two identical continuously operating laboratory-scale biological filter bed columns. A mixture of peat and glass beads was used as filter material. After sterilization it was inoculated with a pure strain of Pseudomonas putida, as employed in previous experimental studies. To determine the influence of the superficial gas flow rate on biofilter performance and to evaluate the phenol concentration profiles along the column, two series of continuous tests were carried out varying either the inlet phenol concentration, up to 1650 mg . m(-3), or the superficial gas flow rate, from 30 to 460 m(3) . m(-2) . h(-1). The elimination capacity of the biofilter is proved by a maximum volumetric phenol removal rate of 0.73 kg . m(-3) . h(-1). The experimental results are consistent with a biofilm model incorporating first-order substrate elimination kinetics. The model may be considered a useful tool in scaling-up a biofiltration system. Furthermore, the deodorization capacity of the biofilter was investigated, at inlet phenol concentrations up to 280 mg . m(-3) and superficial gas flow rates ranging from 30 to 92 m(3) . m(-2) . h(-1). The deodorization of the gas was achieved at a maximum inlet phenol concentration of about 255 mg . m(-3), operating at a superficial gas flow rate of 30 m(3) . m(-2) . h(-1). (c) 1996 John Wiley & Sons, Inc.     

Removal of H2S in down-flow GAC biofiltration using sulfide oxidizing bacteria from concentrated latex wastewater

A biofiltration system with sulfur oxidizing bacteria immobilized on granular activated carbon (GAC) as packing materials had a good potential when used to eliminate H(2)S. The sulfur oxidizing bacteria were stimulated from concentrated latex wastewater with sulfur supplement under aerobic condition. Afterward, it was immobilized on GAC to test the performance of cell-immobilized GAC biofilter. In this study, the effect of inlet H(2)S concentration, H(2)S gas flow rate, air gas flow rate and long-term operation on the H(2)S removal efficiency was investigated. In addition, the comparative performance of sulfide oxidizing bacterium immobilized on GAC (biofilter A) and GAC without cell immobilization (biofilter B) systems was studied. It was found that the efficiency of the H(2)S removal was more than 98% even at high concentrations (200-4000 ppm) and the maximum elimination capacity was about 125 g H(2)S/m(3)of GAC/h in the biofilter A. However, the H(2)S flow rate of 15-35 l/h into both biofilters had little influence on the efficiency of H(2)S removal. Moreover, an air flow rate of 5.86 l/h gave complete removal of H(2)S (100%) in biofilter A. During the long-term operation, the complete H(2)S removal was achieved after 3-days operation in biofilter A and remained stable up to 60-days.     

Biofiltration of ethylbenzene vapours: influence of the packing material

In order to investigate suitable packing materials, a soil amendment composed of granular high mineralized peat (35% organic content) locally available has been evaluated as carrier material for biofiltration of volatile organic compounds in air by comparison with a fibrous peat (95% organic content). Both supports were tested to eliminate ethylbenzene from air streams in laboratory-scale reactors inoculated with a two-month conditioned culture. In pseudo-steady state operation, experiments at various ethylbenzene inlet loads (ILs) were carried out. Maximum elimination capacity of about 120 g m(-3) h(-1) for an IL of 135 g m(-3) h(-1) was obtained for the fibrous peat. The soil amendment reactor achieved a maximum elimination capacity of about 45 g m(-3) h(-1) for an inlet load of 55 g m(-3) h(-1). Ottengraf-van den Oever model was applied to the prediction of the performance of both biofilters. The influence of gas flow rate was also studied: the fibrous peat reactor kept near complete removal efficiency for empty bed residence times greater than 1 min. For the soil amendment reactor, an empty bed residence time greater than 2 min was needed to achieve adequate removal efficiency. Concentration profiles along the biofilter were also compared: elimination occurred in the whole fibrous peat biofilter, while in the soil amendment reactor the biodegradation only occurred in the first 65% part of the biofilter. Results indicated that soil amendment material, previously selected to increase the organic content, would have potential application as biofilter carrier to treat moderate VOC inlet loads.     

Macrokinetic and quantitative microbial investigation on a bench-scale biofilter treating styrene-polluted gaseous streams

We performed a macrokinetic and quantitative microbial investigation of a continuously operating bench-scale biofilter treating styrene-polluted gases. The device was filled with a mixture of peat and glass beads as packing medium and inoculated with the styrene-oxidizing strain, Rhodococcus rhodochrous AL NCIMB 13259. The experimental data of styrene and microbial concentrations, obtained at different biofilter heights, were used to evaluate the pollutant concentration profiles as well as the influence of styrene loading on biomass distribution along the packing medium. Styrene and biomass concentration profiles permitted detection of a linear relationship between the amount of biomass grown in a given section of the biofilter and that of pollutant removed, regardless of the operating conditions tested. Biomass development in the bed appeared to: depend linearly on pollutant concentration at an inlet styrene concentration of <0.10 g m(-3) in the gaseous stream; achieve a maximum value (7. 10(7) colony forming units per gram of packing material) within a wide styrene concentration range (0.10 to 1.0 g m(-3)); and fall sharply beyond this inhibition threshold. The process followed zeroth-order macrokinetics with respect to styrene concentration, which is consistent with zeroth-order microkinetics with either fully active or not fully active biofilm. The maximal volumetric styrene removal rate was found to be 63 g m(packing material) (-3) h(-1) for an influent pollutant concentration of 0.80 g m(-3) and a superficial gas velocity of 245 m h(-1).     

Assessment of the influence of media particle size on the biofiltration of odorous exhaust ventilation air from a piggery facility

Two pilot scale biofiltration systems were constructed and installed at the University College Dublin Research Farm, Lyons Estate. Experimental units consisting of two pens in a 12 pen pig house were sealed off from other pens. Air from each pen was extracted and treated separately in two biofiltration systems. Wood chips larger than 20 mm were selected as the medium for biofiltration system 1, whereas chips of between 10 and 16 mm were used in biofiltration system 2. The moisture content of the media was maintained at 69+/-4% (w.w.b.) using a load cell method. The volumetric loading rates ranged from 769 to 1847 m3 [gas] m(-1) [medium] h(-1) over a 63-day experimental period. Both biofilters reduced odour between 88% and 95%. Ammonia removal efficiencies ranged from 64% to 92% and 69% to 93%, for biofiltration systems 1 and 2, respectively. Sulphur-containing compounds were reduced between 9-66%, and -147-51% across biofiltration systems 1 and 2. The pH of the biofilters' leachate remained between 6 and 8. Pressure drop for biofilter 2 was 16 Pa greater than that of biofilter I at the maximum volumetric loading rate of 1847 m3 [gas] m(-3) [medium] h(-1). It is recommended that a wood chip media particle size greater than 20 mm be used for large scale operation of a biofiltration system on intensive pig production facilities to reduce the development of anaerobic zones and to minimize pressure drop on the system fans.     

Toluene biofiltration by the fungus Scedosporium apiospermum TB1

The performance of biofilters inoculated with the fungus Scedosporium apiospermum was evaluated. This fungus was isolated from a biofilter which operated with toluene for more than 6 months. The experiments were performed in a 2.9 L reactor packed with vermiculite or with vermiculite-granular activated carbon as packing material. The initial moisture content of the support and the inlet concentration of toluene were 70% and 6 g/m3, respectively. As the pressure drop increased from 5-40 mm H2O a strong initial growth was observed. Stable operation was maintained for 20 days with a moisture content of 55% and a biomass of 33 mg biomass/g dry support. These conditions were achieved with intermittent addition of culture medium, which permitted a stable elimination capacity (EC) of 100 g/m3(reactor)h without clogging. Pressure drop across the bed and CO2 production were related to toluene elimination. Measurement of toluene, at different levels of the biofilter, showed that the system attained higher local EC (200 g/m3(r)h) at the reactor outlet. These conditions were related to local humidity conditions. When the mineral medium was added periodically before the EC decreases, EC of approximately 258 g/m3(r)h were maintained with removal efficiencies of 98%. Under these conditions the average moisture content was 60% and 41 mg biomass/g dry support was produced. No sporulation was observed. Evaluation of bacterial content and activities showed that the toluene elimination was only due to S. apiospermum catabolism.     

Start-up and the effect of gaseous ammonia additions on a biofilter for the elimination of toluene vapors

Biotechnological techniques, including biofilters and biotrickling filters are increasingly used to treat air polluted with VOCs (Volatile Organic Compounds). In this work, the start-up, the effect of the gaseous ammonia addition on the toluene removal rate, and the problems of the heat accumulation on the performance of a laboratory scale biofilter were studied. The packing material was sterilized peat enriched with a mineral medium and inoculated with an adapted consortium (two yeast and five bacteria). Start-up showed a short adaptation period and an increased toluene elimination capacity (EC) up to a maximum of 190 g/m3/h. This was related to increased CO2 outlet concentration and temperature gradients between the packed bed and the inlet (Tm-Tin). These events were associated with the growth of the microbial population. The biofilter EC decreased thereafter, to attain a steady state of 8 g/m3/h. At this point, gaseous ammonia was added. EC increased up to 80 g/m3/h, with simultaneous increases on the CO2 concentration and (Tm-Tin). Two weeks after the ammonia addition, the new steady state was 30 g/m3/h. In a second ammonia addition, the maximum EC attained was 40 g/m3/h, and the biofilter was in steady state at 25 g/m3/h. Carbon, heat, and water balances were made through 88 d of biofilter operation. Emitted CO2 was about 44.5% of the theoretical value relative to the total toluene oxidation, but accumulated carbon was found as biomass, easily biodegradable material, and carbonates. Heat and water balances showed strong variations depending on EC. For 88 d the total metabolic heat was -181.2 x 10(3) Kcal/m3, and water evaporation was found to be 56.5 kg/m3. Evidence of nitrogen limitation, drying, and heterogeneities were found in this study.     

Parameters affecting performance and modeling of biofilters treating alkylbenzene-polluted air

Both short-term and long-term biofiltration experiments were undertaken with a biofilter inoculated with a defined microbial consortium and treating an alkylbenzene mixture. The results obtained with such a biofilter in short-term experiments were very similar to those obtained with a biofilter inoculated with a non-defined mixed culture, in terms of maximum elimination capacities (70-72 g m(-3) h(-1)) and the corresponding removal efficiencies (>95%). However, in long-term experiments, a better performance was reached, with a maximum elimination capacity of 120 g m(-3) h(-1), corresponding to a removal efficiency >99% after 2 years of operation. Inoculation proved to be useful for shortening the start-up period. In the long term, it appeared that biomass distribution was not homogenous along the biofilter, which in some cases resulted in a bad fit between simple model equations and experimental data.     

Toluene gas phase biofiltration by Paecilomyces lilacinus and isolation and identification of a hydrophobin protein produced thereof

Paecilomyces lilacinus consumed toluene as the sole carbon source in a gas-phase biofilter packed with perlite obtaining an average elimination capacity of 50 g m(-3) h(-1), a removal efficiency of 53%, and a final biomass of 31.6 mg biomass g dry support(-1). Hydrophobin proteins from the mycelium produced in the biofilter were purified by formic acid extraction and precipitated by electrobubbling, and the molecular weight was found to be 10.6 +/- 0.3 kDa. The peptide mass fingerprinting analysis of the purified hydrophobin by matrix-assisted laser desorption/ionization time-of-flight resulted in the identification of two peptides that presented high homology with sequences of class I hydrophobin proteins from other ascomycetous fungi when compared against the National Center for Biotechnology Information database. The yield of hydrophobin (PLHYD) from P. lilacinus was 1.1 mg PLHYD g biomass(-1). These proteins modified the hydrophobicity of Teflon by lowering the contact angle from 130.1 (+/-2) degrees to 57.0 (+/-5) degrees supporting hot sodium dodecyl sulfate washing. This work is the first report about biodegradation of toluene by the nematophagous fungus P. lilacinus in a gas-phase biofilter and the identification of its hydrophobin protein.     

Mesophilic and thermophilic biotreatment of BTEX-polluted air in reactors

This study compares the removal of a mixture of benzene, toluene, ethylbenzene, and all three xylene isomers (BTEX) in mesophilic and thermophilic (50 degrees C) bioreactors. In the mesophilic reactor fungi became dominant after long-term operation, while bacteria dominated in the thermophilic unit. Microbial acclimation was achieved by exposing the biofilters to initial BTEX loads of 2-15 g m(-3) h(-1), at an empty bed residence time of 96 s. After adaptation, the elimination capacities ranged from 3 to 188 g m(-3) h(-1), depending on the inlet load, for the mesophilic biofilter with removal efficiencies reaching 96%. On the other hand, in the thermophilic reactor the average removal efficiency was 83% with a maximum elimination capacity of 218 g m(-3) h(-1). There was a clear positive relationship between temperature gradients as well as CO(2) production and elimination capacities across the biofilters. The gas phase was sampled at different depths along the reactors observing that the percentage pollutant removal in each section was strongly dependant on the load applied. The fate of individual alkylbenzene compounds was checked, showing the unusually high biodegradation rate of benzene at high loads under thermophilic conditions (100%) compared to its very low removal in the mesophilic reactor at such load (<10%). Such difference was less pronounced for the other pollutants. After 210 days of operation, the dry biomass content for the mesophilic and thermophilic reactors were 0.300 and 0.114 g g(-1) (support), respectively, reaching higher removals under thermophilic conditions with a lower biomass accumulation, that is, lower pressure drop.     

Biological oxidation of hydrogen sulfide under steady and transient state conditions in an immobilized cell biofilter

The removal of hydrogen sulfide (H(2)S) was investigated in a lab scale biofilter packed with biomedia, encapsulated by sodium alginate and polyvinyl alcohol (PVA). The main H(2)S oxidation products were SO(4)(2-), SO(3)(2-), S(2-) and S(0). The immobilized cell biofilter required no separate acclimatization period and showed high removal efficiencies (RE) within the first few days of experiments. The removal efficiencies in the biofilter were consistently greater than 99% even when H(2)S loading was 6 g m(-3)h(-1). The maximum elimination capacity achieved in this study is 8 g H(2)Sm(-3)h(-1) at a loading rate of 13 g H(2)Sm(-3) h(-1). The response of the immobilized cells to fluctuations in inlet concentration and flow rate was determined by subjecting the biofilter to inlet loads of up to 10 g H(2)Sm(-3)h(-1). The biofilter responded effectively to these shock loading conditions and convalesced rapidly within 4-8h. Pressure drop values were consistently less throughout the operational period. The results from this study suggest that an immobilized cell biofilter is effective in treating H(2)S under steady and transient operating conditions.     

Denitrification and nitric oxide reduction in an aerobic toluene-treating biofilter

The presence of significant denitrification activity in an aerobic toluene-treating biofilter was demonstrated under batch and flow-through conditions. N2O concentrations of 9.2 ppmv were produced by denitrifying bacteria in the presence of 15% acetylene, in a flow-through system with a bulk gas phase O2 concentration of >17%. The carbon source for denitrification was not toluene but a byproduct or metabolite of toluene catabolism. Denitrification conditions were successfully used for the reduction of 60 ppmv nitric oxide to 15 ppmv at a flow rate of 3 L min-1 (EBRT of 3 min) in a fully aerated, 17% v/v O2 (superficially aerobic) biofilter. Higher NO removal efficiency (97%) was obtained by increasing the toluene supply to the biofilter.     

Continuous deodorization and bacterial community analysis of a biofilter treating nitrogen-containing gases from swine waste storage pits

A biofilter inoculated with Arthrobacter sp. was applied to the simultaneous elimination of trimethylamine (TMA) and ammonia (NH3) from the exhaust air of swine waste storage pits. The results showed that the biofilter achieved average removal efficiencies of 96.8+/-2.5% and 97.2+/-2.3% for TMA and NH3, respectively. A near-neutral pH (7.3-7.4) was maintained due to the accumulation of acid metabolites and the adsorption of alkaline NH3. Low moisture demand, low pressure drop and high biofilm stability in the system were other advantages. After long-term operation, the bacterial community structure showed that at least twenty-five bands were explicitly detected by a denaturing gradient gel electrophoresis (DGGE) method. However, the inoculated Arthrobacter sp. still maintained a dominant population (>50%). Paracoccus denitrificans' presence in the biofilter could play an important role in oxidizing NH3 and reducing nitrite by heterotrophic nitrification and anaerobic denitrification.     

Biofiltration of n-butyric acid for the control of odour

Odour control from pig production facilities is a significant concern due to increased public awareness and the development of more stringent legislation to control production. Although many technologies exist, biofiltration is still the most attractive due to its low maintenance and operating costs. One of the key odour components, n-butyric acid, was selected for a laboratory scale biofilter study. It was examined as a sole carbon substrate in order to investigate the effectiveness of biofiltration in reducing n-butyric acid concentration under different operating conditions using a moist enriched woodchip medium. Three superficial gas velocities; 38.2, 76.4, and 114.6 m x h(-1) were tested for n-butyric acid concentrations ranging from 0.13 to 3.1 g [n-butyric acid] m(-3) [air]. For superficial gas velocities 38.2, 76.4, and 114.6 m x h(-1), maximum elimination capacities (100% removal) of 148, 113 and 34.4 g x m(3) x h(-1), respectively, were achieved. Upon investigation of effective bed height, true elimination capacities (100% removal) of 230, 233 and 103 g x m(-3) x h(-1), respectively, were achieved at these superficial gas velocities. Averaged pressure drops for superficial gas velocities 38.2, 76.4, and 114.6 m x h(-1) were 30, 78 and 120 Pa, respectively. It was concluded that biofiltration is a viable technology for the removal of n-butyric acid from waste exhaust air, but near 100% removal efficiency is required due to the low odour detection threshold for this gaseous compound.     

Analysis of methanotrophic communities in landfill biofilters using diagnostic microarray

Biofilters operated for the microbial oxidation of landfill methane at two sites in Northern Germany were analysed for the composition of their methanotrophic community by means of diagnostic microarray targeting the pmoA gene of methanotrophs. The gas emitted from site Francop (FR) contained the typical principal components (CH4, CO2, N2) only, while the gas at the second site Müggenburger Strasse (MU) was additionally charged with non-methane volatile organic compounds (NMVOCs). Methane oxidation activity measured at 22 degrees C varied between 7 and 103 microg CH4 (g dw)(-1) h(-1) at site FR and between 0.9 and 21 microg CH4 (g dw)(-1) h(-1) at site MU, depending on the depth considered. The calculated size of the active methanotrophic population varied between 3 x 10(9) and 5 x 10(11) cells (g dw)(-1) for biofilter FR and 4 x 10(8) to 1 x 10(10) cells (g dw)(-1) for biofilter MU. The methanotrophic community in both biofilters as well as the methanotrophs present in the landfill gas at site FR was strongly dominated by type II organisms, presumably as a result of high methane loads, low copper concentration and low nitrogen availability. Within each biofilter, community composition differed markedly with depth, reflecting either the different conditions of diffusive oxygen supply or the properties of the two layers of materials used in the filters or both. The two biofilter communities differed significantly. Type I methanotrophs were detected in biofilter FR but not in biofilter MU. The type II community in biofilter FR was dominated by Methylocystis species, whereas the biofilter at site MU hosted a high abundance of Methylosinus species while showing less overall methanotroph diversity. It is speculated that the differing composition of the type II population at site MU is driven by the presence of NMVOCs in the landfill gas fed to the biofilter, selecting for organisms capable of co-oxidative degradation of these compounds.     

Fungal biofiltration of alpha-pinene: effects of temperature, relative humidity, and transient loads

Over the past decade much effort has been made to develop new carrier materials, more performant biocatalysts, and new types of bioreactors for waste gas treatment. In biofilters fungal biocatalysts are more resistant to acid and dry conditions and take up hydrophobic compounds from the gas phase more easily than wet bacterial biofilms. In the present study, a biofilter packed with a mixture of perlite and Pall rings and fed alpha-pinene-polluted air was inoculated with a new fungal isolate identified as Ophiostoma species. alpha-Pinene is a volatile pollutant typically found in waste gases from wood-related industries. The temperature of waste gas streams from pulp and paper industries containing alpha-pinene is usually higher than ambient temperature. Studies were undertaken here on the effect on performance of temperature changes in the range of 15-40 degrees C. The effect of temperature on biodegradation kinetics in continuous reactors was elucidated through equations derived from the Arrhenius formula. Moreover, the effects of the relative humidity (RH) of the inlet gas phase, transient loads (shock or starvation), and the nature of the nitrogen source on alpha-pinene removal were also studied in this research. The results suggest that the fungal biofilter appears to be an effective treatment process for the removal of alpha-pinene. The optimal conditions are: temperature around 30 degrees C, RH of the inlet waste gas stream around 85%, and nitrate as nitrogen source. The fungal biofilter also showed a good potential to withstand shock loads and recovered rapidly its full performance after a 3-7 days starvation period.     

Microbial community and physicochemical analysis of an industrial waste gas biofilter and design of 16S rRNA-targeting oligonucleotide probes

A study was conducted to investigate the microbial community structure, the physicochemical properties, and the relationships between these parameters of a full-scale industrial biofilter used for waste gas abatement in an animal-rendering plant. Fluorescence in situ hybridization (FISH) was successfully combined with digital image analysis to study the composition of the microbial community. Several new nucleic acid probes were designed and established based on published 16S rDNA sequences and on ones retrieved from the biomass of the biofilter under investigation. Bacterial detection rates varied greatly over time and filterbed depth between 27.2% and 88.1% relative to DAPI counts. Overall, members of the Betaproteobacteria followed by Actinobacteria, Alphaproteobacteria, Cytophaga-Flavobacteria, Firmicutes and Gammaproteobacteria were the most abundant groups. Among the groups below phylum level, members of the Alcaligenes/Bordetella lineage were on average the most abundant group accounting for up to 8.5% of DAPI-stained cells. Whereas the community composition generally showed no vertical gradient, the lower 50 cm of the biofilter proved to be the most active part for the degradation of aldehydes such as 2- and 3-methylbutanal, 2-methylpropanal, and hexanal. This zone of the filterbed being operated in up-flow direction degraded about 80% of these compounds. Dimethyldisulphide was the most common reduced sulphur compound. Statistical analysis of microbial versus waste gas parameters generally revealed only weak or non-significant correlations between the two. Possible explanations for this finding are discussed.     

Fungal removal of gaseous hexane in biofilters packed with poly(ethylene carbonate) pine sawdust or peat composites

The removal of volatile organic compounds (VOC) in biofilters packed with organic filter beds, such as peat moss (PM) and pine sawdust (PS), frequently presents drawbacks associated to the collapse of internal structures affecting the long-term operation. Poly(ethylene ether carbonate) (PEEC) groups grafted to these organic carriers cross linked with 4,4'-methylenebis(phenylisocyanate) (MDI) permitted fiber aggregation into specific shapes and with excellent hexane sorption performance. Modified peat moss (IPM) showed very favorable characteristics for rapid microbial development. Water-holding capacity in addition to hexane adsorption almost equal to the dry samples was obtained. Pilot scale hexane biofiltration experiments were performed with the composites after inoculation with the filamentous fungus Fusarium solani. During the operation of the biofilter under non-aseptic conditions, the addition of bacterial antibiotics did not have a relevant effect on hexane removal, confirming the role of fungi in the uptake of hexane and that bacterial growth was intrinsically limited by an adequate performance of the composites. IPM biofilter had a start-up period of 8-13 days with concurrent CO(2) production of approximately 90 g m(-3) h(-1) at day 11. The final pressure drop after 70 days of operation was 5.3 mmH(2)O m(-1) reactor. For modified pine sawdust (IPS) packed biofilter, 5 days were required to develop an EC of about 100 g m(-3) h(-1) with an inlet hexane load of approximately 190 g m(-3) h(-1). Under similar conditions, 12-17 days were required to observe a significant start-up in the reference perlite biofilter to reach gradually an EC of approximately 100 g m(-3) h(-1) at day 32. Under typical biofiltration conditions, the physical-chemical properties of the modified supports maintained a minimum water activity (a(w)) of 0.925 and a pH between 4 and 5.5, which allowed the preferential fungal development and limited bacterial growth.