Innovative Biological Destruction of Hazardous Chlorinated and Brominated Volatile Disinfection By-Products Using Biotrickling Filters
A Dissertation submitted to the
Division of Research and Advanced Studies of the University of Cincinnati
in partial fulfillment of the requirements for the degree of
DOCTOR OF PHILOSOPHY
in the Department of Chemical and Environmental Engineering of College of Engineering and Applied Sciences
2017 By
Bineyam Hadgu Mezgebe MCP, University of Cincinnati, 2009 Post Graduate Diploma, Berlin University of Applied Sciences, Berlin, Germany 1999
Committee Dr. George A. Sorial (chair) Dr. Margaret J. Kupferle Dr. David Wendell Dr. Endalkachew Sahle-Demessie Dr. Ashraf Aly Hassan
I | P a g e
Abstract
Disinfection by products (DBPs) resulted from the reactions between the chlorine and natural organic substances which increased the formation of trihalomethanes (THMs). DBPs are carcinogens or have been known to cause health risks. Chloroform (CF) is the most abundant of all THMs with a maximum contaminant level (MCL) of 0.070 mg/L. In addition, CF and other THMs could also originate from sources other than by-products of water disinfection.
Several physical and chemical removal methods are used to treat chloroform, which are expensive and could generate secondary pollutants. Biofiltration is one of the most proven technologies for volatile organic compound (VOC) control as it is environment–friendly, cost effective and releases fewer byproducts. In this study, an integrated technology was proposed.
The integrated technology consists of nitrogen or air stripping followed by anaerobic or aerobic bio-trickling Filter (BTF). This study evaluated first CF only and secondly mixtures of THMs
(CF and dichlorobromomethane (DCBM)). A co metabolite (ethanol) and surfactant (Tomadol
25 – 7) have been used to improve the biodegradation process. In addition, surfactin a bio surfactant was seeded within the BTF and its effectiveness has been investigated. Finally, microbial analysis was conducted to determine the dominant and responsible microbes for the
BTFs performance.
In the anaerobic BTF, a co metabolite and surfactant were utilized to enhance the biodegradation process. Upon the addition of ethanol and Tomadol 25 - 7, the performance of the anaerobic reactor improved from an initial 49% to over 64% removal efficiency of CF. On the other hand, the average removal efficiency of CF for aerobic fungi BTF under acidic condition was 80%. For both cases the initial CF loading rate was constant at 0.27 g /m3.hr. The microbial community analysis suggested that A. oryzae and A. restrica were the dominant and responsible
II | P a g e bacteria species observed in the anaerobic BTF. Fusarium sp. and F. solani were the dominant fungi responsible for higher CF degradation.
In practice, a BTF will be exposed to mixtures of THMs and other VOCs emitted together. In this regard, two other independent BTFs were studied to remove mixtures of THMs
(CF and DCBM). These two independent BTFs, one with co metabolite and another one with bio surfactant were investigated in removing the stripped THMs. Adding co metabolite at highest loading rate resulted in removal efficiencies of 85% for CF and 87% for DCBM.
Whereas at the same THMs loading rate condition, the use of surfactin without co metabolite showed a removal efficiency of 85% and 80% for CF and CBM, respectively. The microbial community analysis suggested that Fusarium sp. was the most dominant fungi for the BTF with co metabolite. Whereas, Fusarium oxysporum was the dominant microbes for the BTF with surfactin.
III | P a g e
IV | P a g e
Acknowledgments
I would like to express my deepest appreciation to my advisor, Dr. George A. Sorial, for his guidance, help, and support throughout my journey at the University of Cincinnati. His continuous pursuit of perfection and demand for scientific excellence was a major driving force behind this work. He has helped me a great deal in reaching my goals as a doctoral candidate and as a future researcher. I would like to thank my mentor Dr. Endalkachew Sahle-Demessie for his help and guidance throughout my doctorial work. I also would like to thank my friend and mentor Dr. Ashraf Aly Hassan for his guidance on my research early-on with my laboratory work and later on finishing my research. Moreover, I would like to give appreciation to my committee members, Dr. Margaret J. Kupferle, and Dr. David Wendell for their support and contribution. I gratefully acknowledge the United States Environmental Protection Agency
(USEPA) for the grant that I received during my research.
I would like to thank all my colleagues and friends at University of Cincinnati:
Ayenachew Tegenu, Hengye Jing, Abderrahman Zehraoui, Hafiz Salih, Shahram Ghasemzadeh and Palanisamy, K for making this journey enjoyable and memorable. I would like to thank all my group members and all my friends for all the needed encouragement and support
Finally, I dedicate this work to my parents, especially for my Mom, who made me believe in my abilities to pursue the PhD, my wife, who has endured with me all the challenges I have faced during the program, my sister, who always provided me with moral support.
V | P a g e
Table of Contents
Abstract ...... II Acknowledgments ...... V Table of Contents ...... VI List of Tables ...... X List Figures ...... XI 1. Introduction ...... 1 1.1 Background ...... 1 1.2 Significance of the study ...... 2 1.2.1 DBPs exposure, related health risks and regulations ...... 2 1.2.2 Current controlling techniques and their challenges ...... 4 1.3 Biofiltration and THMs ...... 5 1.4 Specific project objective and innovative concept of biofiltration ...... 6 1.5 Objectives ...... 7 1.5.1 Gas stripping of trihalomethanes (THMs) from water ...... 7 1.5.2 Batch studies ...... 8 1.5.3 Anaerobic vs aerobic conditions for the degradation of chlorinated compounds ...... 9 1.6 Structure of dissertation ...... 11 1.7 References ...... 16 Chapter 2 The Effectiveness of Aeration of Drinking Water to Control Trihalomethanes from Publicly Water System ...... 18 2.1 Abstract ...... 18 2.2 Introduction ...... 19 2.3 Experimental methods and analysis ...... 22 2.3.1 Materials and methods ...... 22 2.3.2 Bench-scale bubble aeration experiment for THMs...... 23 2.3.3 Thermal stratification in the large water storage tanks ...... 23 2.4 Results and Discussion ...... 25 2.4.1 Result from lab scale bubble aeration experiment ...... 25 2.4.2 Bubble aeration modeling for THMs removal ...... 26 2.4.2.1Bubble aeration mathematical modeling - I ...... 26
VI | P a g e
2.4.2.2 Mathematical modeling of bubble aeration - II ...... 27 2.4.2.3 Model parameter estimations ...... 29 2.4.4 Comparing bench scale bubble aeration study with model prediction ...... 31 2.4.5 Bubble aeration and temperatures stratification in large tank ...... 32 2.4.6 Quantitative risk analysis for THMs ...... 33 2.5 Conclusions ...... 35 2.6 References ...... 49 3. Performance of Anaerobic Biotrickling Filter and its Microbial Diversity for the Removal of Stripped Disinfection Byproducts ...... 51 3.1 Abstract ...... 51 3.2 Introduction ...... 52 3.2 Materials and methods ...... 55 3.2.1Materials ...... 55 3.2.2 Experimental methods ...... 56 3.2.2.1 Biotrickling filter bed experiment ...... 56 3.2.2.2 Analytical methods ...... 57 3.2.2.3 Microbial community molecular analysis ...... 58 3.3. Results and discussion ...... 59 3.3.1 Biotrickling filter performances ...... 59 3.3.2. Dehalogenation pathways and kinetics for the different phases ...... 62 3.3.3 Carbon mass balance ...... 63 3.3.4 Microbial ecological analyses and correlation ...... 65 3.4 Conclusion ...... 67 3.5 References ...... 78 4. Comparative Study on the Performance of Anaerobic and Aerobic Biotrickling Filter for the Removal of Chloroform ...... 82 4.1 Abstract ...... 82 4.2 Introduction ...... 83 4.3 Materials and methods ...... 85 4.3.1 Materials ...... 85 4.3.2 Biotrickling filter (BTF) ...... 86
VII | P a g e
4.3.3. Strategies of biomass control ...... 88 4.3.4 Sampling and analysis ...... 88 4.3.5 Microbial community molecular analysis ...... 90 4.4 Experimental results ...... 91 4.4.1 Anaerobic biotrickling filter performance ...... 91 4.4.2 Aerobic biotrickling filter performance ...... 93 4.5 Discussion of the results ...... 93 4.5.1. Performance comparison for anaerobic and aerobic BTFs ...... 93 4.5.2 Kinetics of CF removal in the BTFs ...... 94 4.5.3 Carbon mass balance ...... 96 4.5.4 Microbial ecological analyses and correlation ...... 97 4.6 Conclusion ...... 100 4.7 References ...... 109 5. Impact of Co metabolite Concentration on the Removal of Trihalomethanes by Biotrickling Filter...... 112 5.1 Abstract ...... 112 5.2 Introduction ...... 113 5.3 Materials and methods ...... 116 5.3.1Materials ...... 116 5.3.2 Biotrickling filter (BTF) ...... 117 5.3.3 Sampling and analysis ...... 118 5.3.4 Microbial community molecular analysis ...... 119 5.4. Results and discussion ...... 120 5.4.1 Biotrickling filter performances ...... 120 5.4.2. Kinetics of the CF and DCBM for the different phases ...... 123 5.4.3 Carbon mass balance ...... 124 5.4.4 Nitrogen utilization and COD reduction...... 125 5.4.5 Microbial ecological analyses and correlation ...... 126 5.6 Conclusion ...... 127 5.7 References ...... 137
VIII | P a g e
6. Effectiveness of Bio surfactant without Co metabolite in the Removal of Trihalomethanes by Biotrickling Filter ...... 140 6.1 Abstract ...... 140 6.2 Introduction ...... 141 6.3 Materials and methods ...... 144 6.3.1Materials ...... 144 6.3.2 Biotrickling filter (BTF) ...... 145 6.3.3 Sampling and analysis ...... 147 6.3.4 Microbial community molecular analysis ...... 148 6.4. Results ...... 149 6.4.1Batch comparison study of synthetic and bio surfactant for the solubility of CF in fungi based acidic nutrient solution ...... 149 6.4.2 BTF Performances ...... 150 6.5 Discussion of the results ...... 152 6.5.1 Performance comparison for BTF–A and B ...... 152 6.5.2 Kinetics CF and DCBM for the different phases ...... 153 6.5.3 Carbon mass balance ...... 154 6.5.4 Nitrogen utilization and COD reduction...... 155 6.5.5 Microbial ecological analyses and correlation ...... 157 6.6 Conclusion ...... 158 6.7 References ...... 170 7. Conclusions and recommendations ...... 173 7.1 Summary ...... 173 7.2 Conclusions ...... 173 7.2.1 Effects of aeration for THMs control ...... 173 7.2.2 Biological study ...... 174 7.3 Recommendations of future work ...... 176
IX | P a g e
List of Tables Table 2-1 Physical properties of THMs ...... 36 Table 2-2 Comparison of HC values experimental and predicted ...... 37 Table 2–3 Estimated Diffusivity of THMs as a Function of Viscosity of water at different Temperature ...... 38 Table 2-4 Estimated KLa as a Function of Diffusivity for THMs at different Temperature ...... 39 Table 2-5 Estimated Oral refence dose factor (RfD) and Cancer Slope Factor (SF) for Oral Ingestion of THMs ...... 40 Table 3-1 Operating Condition for the BTF Degrading CF under Anaerobic Condition ...... 69 Table 3-2 Summary of the Community Diversity along with their Metabolic Character sticks ... 70 Table 4–1 Operating conditions for anaerobic and aerobic BTFs degrading CF at a loading rate of 0.27 g/m3h...... 102 Table 5-1 Operating conditions of the BTF degrading CF and DCBM under aerobic acidic conditions ...... 129 Table 6-1 Operating conditions for BTF – A, the BTF was degrading CF and DCBM under aerobic and acidic conditions ...... 160 Table 6-2 Operating conditions for BTF – B, the BTF was degrading CF and DCBM under aerobic and acidic conditions ...... 161
X | P a g e
List Figures Figure 1–1 Schematic Diagram of the BTF System a) Anaerobic, and b) Aerobic ...... 13 Figure 1-2 Schematic Process Diagram for Removal and Biodegradation of DBPs ...... 14 Figure 1-3 Schematic Diagram of the Dissertation Structure...... 15 Figure 2-1 Schematic Diagram of Reactor used for Experimental Bubble Aeration Study ...... 41 Figure 2-2 Temperature Gradient with Respect to depth of storage tank ...... 42 Figure 2–3 Liquid Phase Removal Efficiency at different Time a) DCBM, b) DBCM and c) BF ...... 43 Figure 2-4 Model Estimation of Removal Efficiencies of THMs as a Function of Temperature . 44 Figure 2-5 Transfer Parameter (SP) vs Saturation Parameter (SP) of THMs at 10 and 40OC: a) CF, b) DCBM, c)DBCM, and d) BF ...... 45 Figure 2-6 Liquid Phase Concentration of THMs in Storage tank at Different Temperature using Q= 2 L/ M ...... 46 Figure 2–7 Model Estimation of Removal Efficiency of for CF and BF in storage Tank D ...... 47 Figure 2-8 CF Before and After Aeration, Potential Health Risk effect Assessment of Non – Carcinogenic Hazard quotient (HQ) and Carcinogenic Individual Excess Lifetime cancer risk (IELCR) ...... 48 Figure 3-1Schematic Diagram of the bio trickling filter (BTF) ...... 71 Figure 3-2 Performance of the BTF in the three phases. Phase I (44 days) presence of co metabolite, phase II (97 days) presence of co metabolite and surfactant, phase III (95 days) presence of surfactant with no cometabolite ...... 72 Figure 3-3 Sequential and direct reaction pathways for reductive dehalogenation of CF ...... 73 Figure 3-4 Reaction rate constants for chloroform in three phases. Three data sets have been collected per phase and the error bars present the standard deviations for these replicas. Phase I: Chloroform with Co metabolite (Ethanol), Phase II: Chloroform with Co metabolite (Ethanol) and nonionic surfactant and Phase III: Chloroform with nonionic surfactant...... 74
Figure 3-5 Carbon mass balance: Cumulative carbon input and output as CO2 equivalent in mole for the BTF ...... 75 Figure 3-6 Class level compositions of bacteria based on 97% identity of 16S rRNA gene sequences ...... 76
XI | P a g e
Figure 3-7 Bacterial community diversity for the three conditions for samples collected at the top port of the biofilter. Phase I: Chloroform with Co metabolite (Ethanol), Phase II: Chloroform with Co metabolite (Ethanol) and nonionic surfactant and Phase III: Chloroform with nonionic surfactant ...... 77 Figure 4-1 Schematic diagram of BTFs ...... 103 Figure 4-2 Performance of the anaerobic BTF in the four phases with increasing chloroform-to- ethanol, Phase I: (1: 5), Phase II: (1: 10), Phase III: (1: 20), Phase IV: 1: 40 ...... 104 Figure 4-3 Reaction rate constants for chloroform for both anaerobic and aerobic BTFs in four corresponding phases ratio of chloroform-to-ethanol (co metabolite). Phase I: (1: 5), Phase II: (1:10), Phase III: (1: 20), Phase IV: (1: 40)...... 105 Figure 4-4 Carbon mass balance of BTF over 160 operation days where cumulative carbon
input and output are given as CO2 equivalent in mole for the anaerobic process...... 106 Figure 4-5 Bacterial community diversity for the three phases of anaerobic BTF for samples collected at the top port of the biofilter for selected ratio of chloroform-to-ethanol Phase I: (1: 5), Phase II: (1: 10) and Phase III: (1: 20)...... 107 Figure 4-6 Fungi community diversity for the four phases of aerobic BTF for samples collected at the top port of the biofilter for selected chloroform to ethanol. Phase I: (1:5), Phase II: (1:10), Phase III: (1: 20), Phase IV: (1: 40)...... 108 Figure 5-1 Schematic diagram of the BTF ...... 130 Figure 5-2 Performance of the BTF in the four phases. Phase I: 1: 5 ratio of mixtures of THMs (chloroform and dichlorobromomethane) to ethanol (cometabolite), phase II: 1: 10 ratio of chloroform to ethanol, phase III: 1: 20 ratio of chloroform to ethanol and phase IV: 1: 40 ratio of chloroform to ethanol ...... 131 Figure 5-3 Elimination capacities for mixtures in four phases. Phase I: 1: 5 ratio of mixtures of THMs to ethanol (cometabolite), phase II: 1: 10 ratio of mixture of THMs to ethanol, phase III: 1: 20 ratio of mixture of THMs to ethanol and phase IV: 1: 40 ratio of mixture of THMs to ethanol...... 132 Figure 5-4 Reaction rate constants for mixtures of THMs in four phases. Phase I: 1: 5 ratio of mixtures of THMs to ethanol (cometabolite), phase II: 1: 10 ratio of mixture of THMs to ethanol, phase III: 1: 20 ratio of mixture of THMs to ethanol and phase IV: 1: 40 ratio of mixture of THMs to ethanol...... 133
Figure 5-5 Carbon mass balance: Cumulative carbon input and output as CO2 equivalent in mole for the BTF ...... 134
XII | P a g e
Figure 5-6 Ratio of chemical oxygen demand (COD) to nitrogen utilization vs total loading rates of ethanol and mixtures of THMs. Phase I: 1: 5 ratio of mixture of THMs to ethanol (cometabolite), Phase II: 1: 10 ratio of mixture of THMs to ethanol, Phase III: I:20 mixture of THMs to ethanol and Phase IV: 1: 40 ratio of mixture of THMs to ethanol...... 135 Figure 5-7 A) Fungi community diversity for all the four phases of the BTF for samples collected at the top port (port 2) of the biofilter. B) at the bottom port (port 5) of the biofilter Phase I: 1: 5 ratio of mixture of THMs to ethanol (cometabolite), Phase II: 1: 10 ratio of mixture of THMs to ethanol, Phase III: I:20 mixture of THMs to ethanol and Phase IV: 1: 40 ratio of mixture of THMs to ethanol...... 136 Figure 6-1 Schematic diagram of the BTF ...... 162 Figure 6-2 CF concentration in headspace at different time in the presence of Tomadol 25 – 7 (synthetic surfactant) and surfactin (bio – surfactant ...... 163
Figure 6-3 CO2 production of the Head Space Analysis ...... 164 Figure 6-4 Performance of the BTF in the four phases. Phase I: 5 ppmv of mixtures of THMs (CF + DCBM), phase II: 10 ppmv of mixtures of THMs , phase III: 15 ppmv of mixtures of THMs ...... 165 Figure 6-5 Comparison of Reaction rate constants for mixtures of THMs for BTF A and B: a) for CF, & b) for DCBM ...... 166
Figure 6-6 Carbon mass balance: Cumulative carbon input and output as CO2 equivalent in mole for the BTF for BTF B ...... 167 Figure 6-7 Ratio of chemical oxygen demand (COD) to nitrogen utilization vs total loading rates of mixtures of THMs. Phase I: 5 ppmv of mixture of THMs, Phase II: 10 ppmv of mixture of THMs, and Phase III15 ppm of mixture of THMs... 168 Figure 6-7 Fungi community diversity for all the four phases of the BTF- A for samples collected at the top port (port 2) of the BTF. B) Fungi community diversity for all the four phases of the BTF- B for samples collected at the top port (port 2) of the BTF...... 169
XIII | P a g e
1. Introduction
1.1 Background
Disinfection by-products (DBPs) are formed during chemical water treatment process, when organic moieties and minerals in water react with treatment agents. Chlorinated disinfectants used in water distribution systems to destroy pathogenic microbes may react with natural organic matters (fluvic and humic acids) and produce a range of DBPs including trihalomethanes (THMs) [1]. THMs are considered carcinogenic environmental pollutants [1].
The maximum allowable contaminant level in drinking water for total THMs is 0.08 ppm [2]. In this study, commonly found THMs chloroform (CF) and dichlorobromethane (DCBM) have been used as a model DBPs. Both CF and DCBM are hydrophobic volatile organic compound
(VOC) contaminants, that are recalcitrant to biodegradation due to their high Henry’s law constant at 25˚C are 0.0025 and 0.0016 atm.m3/mol, respectively [3, 4].
Biofiltration systems have recently emerged as an attractive and cost-effective option for controlling a wide variety of contaminants from industrial processes: volatile organic compounds
(VOCs) emissions, organic and inorganic air pollution, and odor from gaseous streams [5].
However, a number of challenges face biofiltration technology. Their performance is strongly dependent on source characteristics (VOCs concentration, employed loading rates, VOCs composition, and emission modes), which limits the handling efficiency and application of the biofiltration systems. Typically, most industrial off-gas streams have variable flow rates and contaminants compositions that limit the handling efficiency of the biofiltration system [6, 7].
Furthermore, hydrophobic contaminants are not readily bio-available [8]. Thus, an innovative biofiltration technology is needed to expand the application of this technology in emission
1 | P a g e
control. As a solution to this limitation in biofiltration systems, a novel technology was proposed to achieve stable removal efficiencies of CF and DCBM, as a model DBPs, by utilizing a biotrickling filter (BTF) under anaerobic and aerobic conditions. The increased performance was checked by introducing a co metabolite. Additionally, the bio-availability of hydrophobic compounds enhanced by the introduction of nontoxic surfactants that served dual processes, increasing solubility and limiting excess biomass growth.
1.2 Significance of the study
1.2.1 DBPs exposure, related health risks and regulations
The most commonly found DBPs, which include THMs, form when chlorine and bromine interact with natural organic materials in water, e.g chlorinated drinking water and chlorine-treated swimming pools. DBPs fumes can be found in the air during activities such as showering, bathing, dishwashing, and swimming. People are exposed to these DBPs by drinking chlorinated or brominated water and by breathing air containing DBPs [9]. Other sources include: water from treated wastewater, agricultural run-off, pharmaceuticals, antibacterial agents, estrogens, pesticides and textile dyes [10, 11].
Recent epidemiologic studies indicate that the exposure to these DBPs may increase the risk of bladder cancer [10]. Exposure in swimming pools because of chlorination has been linked to respiratory effects, including asthma. When chlorine reacts with urea in the swimming pool trichloramine is formed which has been suspected for the cause of asthma. DBP precursors, skin cells, hair, and lotions/sunscreens are other DBPs precursors which are found in the swimming pool [10, 12]. Due to their health risks and being a probable carcinogen, trihalomethanes (THMs)
2 | P a g e
especially chloroform became regulated in the U.S. since 1979 [17]. A few other DBPs are currently also regulated in the U.S. including five halo acetic acids (HAAs), chlorite, and bromate. THMs and HAAs are formed primarily by chlorine and chloramines; chlorite is a DBP from chlorine dioxide, and bromate is mostly from ozonation [12].
The passage of the Clean Water Act in 1974 gave a way to control the contaminant level in drinking water. This act requires EPA to determine the level of contaminants in drinking water at which no adverse health effects are likely to occur. Hence, EPA sets an enforceable regulation maximum contaminant levels (MCL) based on the best available science to prevent potential health problems. The rule sets the MCL for THMs and five HAAs (HAA5) at 80 µg/L and 60
µg/L, respectively. Similarly, the 1990 Clean Air Act amended the concentration of VOCs in the air must comply with the US EPA's regulations. The amendments had led to the development of more tough regulations, standards, guidelines and codes of VOCs emissions. It also opened the way for the development of processes aimed at reducing hazardous air pollutants. Since these volatile DBPs contribute greatly to air pollution, there must be air pollution control to eliminate the stripped VOCs from the air exiting the process in order to meet the air quality regulations
[13]. For instance, according to the agency for toxic substances and disease registry (ATSDR), the normal expected amount of chloroform to be in the air ranges from 0.004 to 0.01 ppmv .
Whereas, the daily typical average amount of chloroform in the air ranges from 0.40 to 444 ppmv in various areas. Likewise, the estimated amount of chloroform in drinking water ranges from 0.80 to 18 ppmv. However, in some places, chloroform concentrations may be higher than
9 ppmv. As much as 123 ppmv was found in air at a municipal landfill and up to 18 ppmv was found in treated municipal drinking water. Drinking water derived from well water near a hazardous waste site contained 383 ppmv [14].
3 | P a g e
1.2.2 Current controlling techniques and their challenges
Drinking water-treatment plants have difficulties meeting regulatory limits. Many plants have tried different methods of controlling DBPs mechanisms which include change of disinfection sites, removal of DBPs after their formation, removal of precursors prior to disinfection, and use of alternative disinfectants which create fewer DBPs [15]. Removing precursors before reacting with disinfectant is the most effective and economical methods to control DBPs in most treatment plants. Therefore, many treatment facilities have focused on the removal of Natural Organic Matter (NOM) to control DBPs [12, 16]. Another way of controlling
DBPs after formation is the utilization of different physical and chemical techniques which included adsorption by GAC, ion exchange, and coagulation [17-19]. Additionally, alternative disinfectant which include chlorination, chloramination, chlorine dioxide, ozonation, UV, and
UV/H2O2 are also used to remove DBPs[18].
However, some studies reported that the limitations using these DBPs removal techniques. For instance, Krishna et al. [20] reported the drawbacks of using powdered activated carbon (PAC) for the removal of precursor. The adsorption capability of the activated carbon can be inhibited when the THM react with the carbon surface. Therefore, the removal of THM precursor and THMs from drinking water by PAC is variable [20]. Similarly, Shang et al. [18] and Watson et al. [19] reported that the use of ion exchange, coagulation and adsorption for
DBPs removal can give varied results due to the interference from computing ions and the NOM causing lowering of halide adsorption [18, 19]. The study conducted by Shorney et al. [21] on the use of enhanced coagulation and ion exchange on the removal of total organic compound
(TOC) in reducing precursors showed the removal was not sufficient. The main reason is that the process is affected by the change in pH and turbidity of the TOC. Additionally, greater usage of
4 | P a g e
the ionic composition to adjust pH can significantly increase the concentration of ions such as chloride, sulfate, and sodium and limits the removal process [21].
Another way of reducing THMs, from surplus chlorine is the use of alternative disinfectants, which include chlorination, chloramination, chlorine dioxide, ozonation, UV, and
UV/H2O2. The utilization of an alternative disinfectant is usually the most economical option for
DBPs removal, however, these alternative disinfectants can lead to the formation of potentially toxic DBPs [11, 22]. The changes in disinfection practice can lead to other issues and problems.
For example, using ozone can significantly reduce or eliminate the formation of THMs and
HAAs, but it can result in the formation of bromate, especially when elevated levels of bromide salts are present in the source waters [23]. Similarly, chloramines have been found to promote cyanogen chloride and NDMA formation.[24]. The utilization of alternative disinfectant can also increase the formation of other DBPs, including nitrosamines, iodo-acids and iodo-THMs [12].
Recent studies have identified emerging DBPs that may be more toxic than some of the regulated ones (chlorine- and bromine containing THMs and haloacetic acids). Some of these emerging DBPs are associated with reduced quality drinking water supplies. In some cases, alternative primary or secondary disinfectants to chlorine that minimize the formation of some of the regulated THMs may increase the formation of some of the emerging byproducts [25, 26].
1.3 Biofiltration and THMs
Although most studies show successful biodegradation of CF in the liquid phase, there is a limited amount of reported work on the use of biofiltration for the removal of CF and DCBM from gaseous streams. Biofiltration is one of the proven technologies for removing VOCs from
5 | P a g e
high volume stream as it is environmentally friendly, cost effective and releases fewer byproducts [27]. The use of aerobic biofiltration technique has been reported for the biotreatment of chloroform with other mixtures of different VOCs [27, 28]. Yoon et al. have shown the degradation potential of nine VOCs including chloroform and found the highest removal was for toluene and the lowest removal was for CF [27]. Similarly, Balasubramanian et al. evaluated the biodegradation of chloroform along with a mixture of VOCs commonly found in pharmaceutical emissions, using a biotrickling filter. Their study showed that increasing the rate of CF loading significantly reduced the degradation efficiency of the reactor for the mixture of
VOCs [28].
1.4 Specific project objective and innovative concept of biofiltration
The research will be conducted using CF and DCBM as model for DBPs. Both CF and
DCBM are hydrophobic VOC contaminants, that are recalcitrant to biodegradation due to their high Henry’s law constant at 25˚C are 0.0025 and 0.0016 atm.m3/mol, respectively [3, 4]. DBPs were chosen because of their potential health toxicity. Fig. 1-1 shows the schematic process diagram for the biodegradation of DBPs. The experimental work will be conducted on two lab- scale biotrickling filter (BTF) systems (Anaerobic and Aerobic), and details are shown in Fig. 1-
2. The reactor will be packed with palletized diatomaceous earth biological support media to a depth of about 60 cm (Celite® 6 mm R-635 Bio-Catalyst Carrier; Celite Corp., Lompoc, CA).
6 | P a g e
1.5 Objectives
The main objective of this study is enhancing the overall quality and reducing health risks of drinking water as well as ambient air by eliminating hazardous DBPs. The objective of this project is to apply innovative biological removal for hazardous/ carcinogenic disinfection byproducts (DBPs) using biofiltration technique, which operates under different environmental conditions. This research was conducted mainly using first CF only and secondly using mixtures of CF and DCBM as a model for DBPs by utilizing the innovative technology, bio-trickling filter
(BTF) under different conditions. Additionally, bench scale defuse aeration was investigated to evaluate the effectiveness of air stripping in removing THMs.
1.5.1 Gas stripping of trihalomethanes (THMs) from water
In this study, bench-scale laboratory experiments were conducted to evaluate the effectiveness of air stripping for disinfection by products. Parametric studies were performed to evaluate the effect gas flow rate and THMs inlet concentration on the THMs recovery efficiency.
In the study, three 2-liter stripping tanks were filled with DI water and known amounts of THMs standards were spiked into the DI water. The flow of nitrogen gas to each stripping tank was controlled using individual mass flow controllers. Water samples were collected from each tank initially and at ten minutes intervals, and put in 40-mL glass vials with no headspace. Liquid- liquid micro extraction according to EPA Method 551.1 was followed to quantify THMs concentration. Liquid samples were extracted with methyl tert-butyl ether following EPA
Method 551.1and known amounts of internal standard of bromofluorobenzene was spiked in each sample. The extracted THMs were analyzed using GC-MS.
7 | P a g e
1.5.2 Batch studies
Batch study on surfactant screening before introducing to the BTF
Before injecting surfactant to the BTF system, head space analysis was done to screen different types of surfactants. The batch study assisted to know which surfactant perform better at which concentration in a given temperature. Hence, for the study the influences of five surfactants were studied based on the Henry’s constant of THMs. THMs Henry’s constant estimated at selected temperatures. Surfactant concentrations varied at concentration of 0.5, 1 and 2 times of the critical micelle concentration (CMC). For analytical convenience, the three
THMs mixtures of compounds were prepared in an ethanol solution. The gaseous concentrations of the two systems at equilibrium were measured and used to compute the dimensionless Henry’s law constant using the combined mass balance equations. The equilibrium partitioning in closed systems (EPICS) method used to estimate the Henry’s law constant of each THMs at different temperature and CMC concentration.
Batch study on chemical and bio - surfactant
The goal of this batch study was to compare the widely used synthetic and bio surfactant on CF’s degradation. The experiment was performed by placing 200 mL of acidic nutrient with fungi in a 250 mL amber vials. Two kinds of surfactants namely, tomadol 25-7 (synthetic surfactant) and surfactin (bio surfactant) were used for this study. Additionally, nutrient solution with no surfactant added mixed with CF was used as a control. The vials were capped with a 24 mm replacement mini-inert valve (Supelco, PA). The vials were then continuously mixed in a tumbler. The vials were maintained at a constant room temperature of 22±1 oC. By taking sample from head space each vial analysis was performed.
8 | P a g e
1.5.3 Anaerobic vs aerobic conditions for the degradation of chlorinated compounds
The difference in micro-organism growth for anaerobic and aerobic environment depends on the liquid feed for nitrogen source. These organisms consume organic matter to support life functions. Anaerobic microorganisms follow anaerobic digestion, which require little or no oxygen to live, whereas aerobic microorganism use oxygen from air and nitrate to produce nitrogen gas, carbon dioxide. The nutrient feed will be varied between nitrate and ammonia to study the influence of the electron donor on the biodegradation process. Due to these different conditions, it is expected to have different removal efficiencies and operation kinetics between both conditions. The pH and temperature both have impacts on the microbial growth. In practice, a BTF will be exposed to dynamic varying loading conditions, e.g. variable THMs and co metabolite flow rate and contaminant concentration or periodical shut down of contaminants.
The presence of an easily degradable compound in the feed gas has the potential to enhance the biodegradation of chlorinated compounds. Therefore, knowledge about the BTF performance in the absence and presence of a hydrophilic co metabolite and non-toxic surfactant are needed for design and development. Changing the operating environment from anaerobic to aerobic may help to increase the performance of the BTF. In this part of the study, by changing the operating environment from anaerobic to aerobic, the BTF will be set up to operate under nutrient flow of pH 4 to study the aerobic environment change in addition with acidic conditions on microorganisms’ growth. This acidic pH supports fungi growth. The pH has proven in several studies to have a significant impact on BTF performance. Finally, a comparison was conducted for the different environments (anaerobic and aerobic) conditions on the performance of the BTF
9 | P a g e
a. To study the effect of a co metabolite on the biodegradation of DBPs – (single solute -
CF)
The performance and removal efficiency of the BTF affected by different loading rates of the CF and co metabolite (ethanol) feed concentration. By combining different ratio (CF to ethanol), several experiments conducted. Due to the different ratio, the obtained results had different removal efficiency levels as expected.
b. Effect of surfactant on the bio-availability of DBPs and the biomass accumulation-
(single solute - CF)
The introduction of surfactant helped improving the degradation process by increasing solubility of the contaminant. The addition of these non-toxic surfactant to the BTF system encouraged the growth of anaerobic microbes by consuming the contaminant. Thus, the bio- availability of hydrophobic compounds can be enhanced by the use of nontoxic surfactants.
Several batch experiments conducted to choose the very best surfactant for the system. The equilibrium partitioning in closed systems, (EPICS) analysis for headspace used to determine micellar partitioning of DBPs for several DBPs in the presence of surfactants over a range of temperatures and micellar concentration levels.
c. Study of mixtures of THMs
The performance of BTF could be affected when adding mixtures of THMs (CF +
DCBM) instead of single solute (CF). Two independent BTFs were running continuously to degrade gaseous mixtures of CF and DCBM. One system was running in the presence of co metabolite at different loading rate. Another system was running in the presence of surfactin (bio surfactant). Finally, comparison was conducted on the performance of between the two BTF’s,
10 | P a g e
d. Identification of different organisms
The main aim of this study was to isolate the microbial strains responsible for the degradation under the different conditions. The identification of the species would enable the researches to pursue more fundamental research to increase the process efficiency and possibly gain more understanding of natural processes. This part of the study would be pursued by sending biofilm samples to interagency microbiology laboratories for identification using the Mo
Bio PowerSoil DNA (M Bio Lab, Inc., Carlsbad, CA) Kit, which was done by Molecular
Research LP (MR DNA, Shallowater, TX).
1.6 Structure of dissertation
The dissertation is divided into five main topics: Lab scale air stripping experiment,
Anaerobic/ Aerobic condition in the presence and absence of co metabolite and surfactant, mixtures of CF and DCBM with co metabolite, mixtures of CF and DCBM with surfactin (bio surfactant) and microbial structure analysis. The first one was experimentally conducted to determine the aeration application to remove THMs. The second and third topics dealt with for the degradation of single solute (CF). The fourth and fifth dealt with the degradation of mixtures of THMs. The microbial structure analysis was performed at the end of each condition specified.
The dissertation research plan is presented in the figure 1-3.
The single solute (chloroform) study is covered in chapters 3 and 4. Chapter 3 focuses on biodegradation of chloroform in the presence and absence of surfactant. Chapter 4 explains the comparative study of the anaerobic and aerobic BTF for the removal of chloroform at different co metabolite loading rate.
11 | P a g e
The mixtures of THMs study is presented in Chapters 5 and 6. Chapter 5 investigates the biodegradation of mixtures of THMs under acidic condition for the different loadings of co metabolite in fungi based BTF. In chapter 6, surfactin was introduced to the BTF for the removal of mixtures of THMs under acidic condition. In this chapter, investigation has been conducted on the use of bio – surfactant for the removal of THMs. Finally, chapter seven provides the key conclusions obtained from this work and gives future recommendations.
12 | P a g e
Figure 1-1 Schematic process diagram for air stripping and biodegradation of DBPs
13 | P a g e
Figure 1-2 Schematic diagram of BTF systems: a) Anaerobic; b) Aerobic
14 | P a g e
Figure 1-3 Schematic diagram of the dissertation structure
15 | P a g e
1.7 References 1. Villanueva, C.M., et al., Disinfection byproducts and bladder cancer: a pooled analysis. Epidemiology, 2004. 15(3): p. 357-367. 2. Richardson, S.D., Disinfection by-products and other emerging contaminants in drinking water. TrAC Trends in Analytical Chemistry, 2003. 22(10): p. 666-684. 3. Sorial, G. and A. Bagtzoglou, Innovative Remediation Technologies for Pollution Abatement. Water, Air & Soil Pollution: Focus, 2008. 8(3): p. 253-255. 4. Aly Hassan, A. and G. Sorial, Biofiltration of n-hexane in the presence of different levels of benzene vapors, in Biotechniques for Air Pollution Control, J. Bartacek, C. Kennes, and P. Lens, Editors. 2010, CRC Press 5. Cox, H.H. and M.A. Deshusses, Effect of starvation on the performance and re- acclimation of biotrickling filters for air pollution control. Environmental science & technology, 2002. 36(14): p. 3069-3073. 6. Fitch, M.W., E. England, and B. Zhang, 1-Butanol removal from a contaminated airstream under continuous and diurnal loading conditions. Journal of the Air & Waste Management, 2002. 52(11): p. 1288-1297. 7. Mohseni, M. and D. Allen, Transient performance of biofilters treating mixtures of hydrophilic and hydrophobic volatile organic compounds. Journal of the Air & Waste Management Association 1999. 49(12): p. 1434-1441. 8. Aly Hassan, A. and G.A. Sorial, A comparative study for destruction of n-hexane in trickle bed air biofilters. Chemical Engineering Journal, 2010. 162(1): p. 227-233. 9. Prevention, C.f.D.C.a. Disinfection By-products (DBPs). 2014 [cited 2014 February 2]. 10. Barceló, D., Emerging Organic Contaminants and Human Health. Vol. 20. 2012, Barcelona, Spain: Springer Heidelberg New York Dordrecht London. 11. Richardson, S.D., et al., Occurrence, genotoxicity, and carcinogenicity of regulated and emerging disinfection by-products in drinking water: a review and roadmap for research. Mutation Research/Reviews in Mutation Research, 2007. 636(1): p. 178-242. 12. Aly Hassan, A. and G.A. Sorial, A comparative study for destruction of n-hexane in Trickle Bed Air Biofilters. Chemical Engineering Journal, 2010. 162(1): p. 227-233. 13. Calabrese, E.J., Health Effects of Drinking Water Contaminants. 1989: CRC Press. 14. Registry, A.f.T.S.a.D. Toxic Substances Portal: Chloroform. 2014 [cited 2014 03/15/14]; Available from: http://www.atsdr.cdc.gov/substances/toxsubstance.asp?toxid=16. 15. Singer, P.C., Reckhow, D.A., 1999.. , Chemical oxidation In:. Water Quality and Treatment, ed. R.D.E. Letterman. 1999, New York: McGraw-Hill Inc. pp. 12.38-12.44. 16. Bond, T., et al., Treatment of disinfection by‐product precursors. Environmental technology, 2011. 32(1): p. 1-25. 17. Pavoni, B., et al., Assessment of organic chlorinated compound removal from aqueous matrices by adsorption on activated carbon. Water Research, 2006. 40(19): p. 3571- 3579. 18. Xie, L. and C. Shang, A review on bromate occurrence and removal strategies in water supply. Water Science & Technology: Water Supply, 2006. 6(6). 19. Watson, K., M. Farré, and N. Knight, Strategies for the removal of halides from drinking water sources, and their applicability in disinfection by-product minimisation: A critical review. Journal of Environmental Management, 2012. 110: p. 276-298. 20. Gopal, K., et al., Chlorination byproducts, their toxicodynamics and removal from drinking water. Journal of hazardous materials, 2007. 140(1): p. 1-6.
16 | P a g e
21. Shorney, H.L., P. Scanlan, and J. Neemann. Use of Enhanced Coagulation to Reduce DBPs. in Environmental and Pipeline Engineering 2000. 2000. ASCE. 22. IJpelaar, G.F., et al., UV disinfection and UV/H2O2 oxidation: by-product. Environ. Eng. Sci, 2007. 4: p. S51-S56. 23. Corwin, C.J. and R.S. Summers, Adsorption and desorption of trace organic contaminants from granular activated carbon adsorbers after intermittent loading and throughout backwash cycles. Water Research, 2011. 45(2): p. 417-426. 24. Bond, T., M.R. Templeton, and N. Graham, Precursors of nitrogenous disinfection by- products in drinking water––A critical review and analysis. Journal of Hazardous Materials, 2012. 235: p. 1-16. 25. Krasner, S.W., The formation and control of emerging disinfection by-products of health concern. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 2009. 367(1904): p. 4077-4095. 26. Bond, T., et al., Occurrence and control of nitrogenous disinfection by-products in drinking water–a review. Water Research, 2011. 45(15): p. 4341-4354. 27. Yoon, I.-K., C.-N. Kim, and C.-H. Park, Optimum operating conditions for the removal of volatile organic compounds in a compost-packed biofilter. Korean Journal of Chemical Engineering, 2002. 19(6): p. 954-959. 28. Balasubramanian, P., L. Philip, and S.M. Bhallamudi, Biotrickling filtration of complex pharmaceutical VOC emissions along with chloroform. Bioresource technology, 2012. 114: p. 149-159.
17 | P a g e
Chapter 2 The Effectiveness of Aeration of Drinking Water to Control Trihalomethanes from Publicly Water System
2.1 Abstract
Post-treatment aeration is one of the methods used to control trihalomethanes (THMs) following long water aging. The effectiveness of bench-scale bubble diffuse aeration to reduce three THMs, dichlorobromomethane (DCBM), dibromochloromethane (DBCM) and bromoform
(BF) was studied. In the study, THMs concentration was held constant while investigating for different gas flow rates. Differences in the percent THMs removals were assessed, including impacts on their Henry’s law constants. The removal efficiencies of these THMs from the initial concentration of 100 µg/L for air flow rate of 2 L/min were 99%, 97% and 88%, respectively.
The estimated Henry’s law constant values obtained from this experiment for DCBM, DBCM and BF were 0.0883 ± 0.008, 0.0502 ± 0.003 and 0.0291 ± 0.003, respectively. Computational methods were used to understand the effects of temperature, flowrate, initial THM concentration and Henry’s law constant. Large storage tank data were collected from known storage facilities for the temperature gradient effect study. Long term temperature monitoring of large water storage tanks revealed that seasonal variation creates temperature stratification which widens as ambient temperature increased. Mathematical model estimated the effects of the temperature stratification on the removal efficiency of THMs using diffuse aeration in large containers.
Finally, quantitative health risk analysis for chloroform (CF) from different streams also investigated. The reduction of CF from aeration was assessed based non-carcinogenic substances hazard quotient (HQ) and individual excess lifetime cancer risk (IELCR) estimates confirmed that aeration could reduce risk to an acceptable range.
18 | P a g e
2.2 Introduction
The formation of Disinfection byproducts (DBPs) from chlorination have been known to cause health risks [1]. Chlorination is most economical and feasible disinfectant in the United
State, but it generates various DBPs especially halogenated organic by-products such as THMs which are potential carcinogens. When chlorine treated finished water stored for long time,
THMs are formed at a rate that is proportional to the chlorine consumed over the storage period
[2]. Water aging and temperature are some of the factors affecting THMs formation within a storage tank. When water enters the tank it will almost certainly have a different temperature to the water already in the tank [3]. The exchange of air between the top of a water tank and the surrounding atmosphere is very little. Also, the air above the water surface in a tank will quickly reach saturation and no further heat transfer will then take place by this means. Therefore, the water temperature will be related to a more general heat balance and it could be higher than the surrounding air mainly due to absorption of solar radiation. [3], The seasonal temperature difference in the storage tanks also accounts to the difference in chemical reaction rates and thus higher temperatures in the warm season accelerated the rate of THM production than in cold season in the distribution system [4]. The influences of seasonal variations of tank temperatures have not been well studied. Recent studies have shown that tank aeration can reduce THMs formed as the result of high water age. Thus, the objective of this study is to evaluate aeration as a method to remove THMs from drinking water in storage tanks. Large-scale, post-treatment
THMs removal strategies within a distribution system have generally been focused on aeration technologies [5, 6]
Aeration is one of the techniques used to remove THMs by utilizing the liquid-gas concentration gradient as means for mass transport of volatile compounds from the liquid to gas
19 | P a g e
phase. It is a mass transfer process for removing contaminant from liquid to gas phase [7]. It uses relatively clean air to remove volatile contaminants dissolved in water and transfers the contaminants into the gaseous phase [8]. This technique has shown to be effective for removing compounds such as chlorinated benzenes, simple halogenated organic compounds, aromatic hydrocarbons, pesticides and select trihalomethanes (THMs) from contaminated water [9, 10].
Air stripping showed acceptable performance for CF removal from drinking water storage with the removal efficiency of 97.6% [11]. Another study also showed a 95% removal of CF by aeration [12]. Similarly, Sherant et al [5] investigated the removal of THMs for different air to water ratio of 0.8:1 and 53:1 and find out the efficiency ranges from 92 – 99%. Tarquin et al. reported on their research that CF was the easiest to remove while bromoform was the most difficult and they gave the range of removal efficiency for chloroform was 69% to 96% whereas for bromoform was 32% to 87% [13].
There are many techniques for air striping including diffused aeration, packed tower air stripping, mechanical surface aeration and spray aeration [14]. Diffuse or bubble aeration uses compressed air to the bottom of the water tank which allows air bubbles to travel to the top of the tank while transferring oxygen from gas to aqueous phase and stripping volatile compounds [15]
Air is introduced through a bubbler or a nozzle into the water stream[16]. Bubble aeration has been recognized for long time as a means of removing THMs and used as a cheaper way of reducing THMs in a storage tanks [6, 17-19]. The main factors influencing the removal of THMs are the concentrations of THMs, the air-to-water ratio, hydraulic load, total height of packing and the Henry’s law constant of each compound. For most stripping applications, the equilibrium between air and water can be described using Henry’s Law constant. It is the most important
20 | P a g e
parameter that affects the extent of removal of VOCs and given by the following equation [18,
20]
Cg HC CL …………………………………………..………………………………………….. (1)
where Cg and CL are gas and liquid phase THMs and HC is Henry’s constant,
In this study, lab scale experiments were conducted to evaluate the effectiveness of diffused gas (nitrogen (N2)) stripping on the removal of THMs using a bench-scale bubble stripping unit. In the experimental work, THMs concentrations were held constant on the bench- scale diffuse aerator while investigating for three selected flow rates. The percent removal of the three THMs, the influent concentrations and the impacts on the Henry’s constant of THMs assessed using a model. Most systems have CF as the predominant THM species thereby making it the primary target of aeration reduction. However, CF, with the greatest Henry’s constant, is the easiest THM to remove from water. CF was easily stripped, and sample reproducibility was low due to losses during storage after the tests. This study evaluates whether properly designed aeration systems can successfully remove the brominated compounds as well. Hence, for this study dichlorobromomethane (DCBM), dibromochloromethane (DBCM) and Bromoform (BF) were selected. Similarly, the application of aeration for the use of public water system storage tanks were studied. Also, the temperature gradient data were collected from known storage tank for the temperature gradient effect study. Finally, investigation has been conducted to determine the quantitative health risk of chloroform from different streams. The overall objective of this study was to assess and develop mathematical model to calculate the capability of bubble aeration to be part of an adaptation strategy to maintain satisfactory THM levels within a distribution tank.
21 | P a g e
2.3 Experimental methods and analysis
2.3.1 Materials and methods
All chemicals used were analytical grade (99.9% purity), and were procured from Fisher
Scientific (Fair Lawn, NJ, USA) and Sigma Aldrich (USA). All experiments were performed with deionized (DI) water from U.S. Environmental Protection Agency, Cincinnati. The batch of water used in an experiment allowed to reach room temperature overnight prior to running an experiment. All glassware was cleaned accordingly before each experiment. Samples were collected from the reactor at initial time, to determine the initial THMs concentrations. After the initial samples were taken, the gas supply was operated to achieve the desired air flow rate for the specific experimental run. At predefined sampling times, the aeration was stopped, and samples were taken. The vials which were used in the sampling filled and cap with zero headspace. These experimental sampling steps were repeated at subsequent sampling times until the final sampling step was reached.
Data was collected for tank temperature stratification analysis. The water storage tank for this study is located in southwest Pennsylvania. The maximum capacity of this tank is 260,000 gallons, it the most it typically contains at any given time is about 251,000 gallons. This tank contains a single inlet/outlet (i.e., one pipe goes into the tank). Based on the operational data collected from the time of the study in 2013, the average flow was 30.1 GPM and peak flow was
545 GPM. The water source for this tank is drawn from the Allegheny River. The water is then treated using conventional surface water treatment that uses chloramines as a secondary disinfectant.
22 | P a g e
2.3.2 Bench-scale bubble aeration experiment for THMs
Three 2-liter stripping tanks were filled DI water, and known amounts of THMs standards were spiked into the DI water. Fig. 2-1 presents the schematic diagram of the container used for this bench scale bubble aeration study. The flow of nitrogen gas to each stripping tank was controlled using individual mass flow controllers (SmartTrack 100, Sierra Instruments, and
Monterey, CA). Water samples were collected from each container initially and at ten-minute intervals, and put in 40-mL glass vials with no headspace. Liquid-liquid microextraction according to EPA Method 551.1 was followed to quantify THMs concentration. Liquid samples were extracted with methyl tert-butyl ether following EPA method and known amounts of internal standard of bromofluorobenzene was spiked in each sample. The extracted THMs were analyzed using GC-MS (Agilent Technologies 6890 GS-Ms of 5973). The oven temperature was initially held at 350C for 16.80 minutes, then, increased by 25 °C/min up to 300 °C and stayed for 2 minutes. The sample then went through a column (DB5, 30.0 µm X 320 µm 0.25µm) with splitless inlet mode. Column’s flow rate was 1mL/min with an average velocity of 26 ml/sec.
Table 2-1, provided with lists of physical properties of the THMs including Henry’s constants, with the boiling points and molecular weights.
2.3.3 Thermal stratification in the large water storage tanks
Water tanks can suffer from thermal stratification inside water storage tanks which could result in serious water quality problems such as loss of residual disinfectant, taste, and odor complaints. Water entering the tank during the fill cycle is typically much colder than water already sitting in the tank, which has been warmed by the sun. The cold water will remain on the bottom of the tank while the warmer, more buoyant water floats to the top of the tank for many
23 | P a g e
days. The formation of THMs, in large storage tanks with aging water are likely to be higher. For this reason, storage tank bubble or spray aeration can be an effective treatment option for
THM removal [5, 6]. Active tank mixing can eliminate thermal stratification in storage tanks by ensuring that water entering the tank is completely circulated with the rest of the tank’s volume. Aeration could result lower headspace temperatures and reduce corrosion rates inside the tank.
Additionally, post-treatment DBP removal strategies within a distribution system have generally been focused on aeration technologies [5, 6]. The temperature stratification data for this tank was collected using temperature loggers. The general procedure includes that the loggers should be evenly spaced along the wire rope on the temperature monitor based on the tank information. The temperature logger on the float will move freely along the wire rope throughout the tank’s typical operating range. The first temperature logger that is secured on the wire rope (i.e., below the float) should be fastened so that it will be slightly below the typical minimum operating level. Doing so should ensure that the loggers are continuously submerged in water and not recording air temperature. The last temperature logger should be fastened on the wire rope, just above the anchor. The remaining loggers should be fastened at evenly distributed distances between the first and last loggers. In this study, four temperature ranges were chosen 6, 10, 16 and 23 0C by optimizing the average temperatures with in the tank at different levels as shown on Fig. 2-2. The complete temperatures data with respect depth used for the analysis was collected from April 4, 2014 from 8:00 am to May 14, 2014 noon for more than 40 days. The result indicates that as depth increased towards the top of the storage tank, the temperature is increasing.
24 | P a g e
2.4 Results and Discussion
2.4.1 Result from lab scale bubble aeration experiment
In the previous studies of diffused aeration for THMs removal, researchers have used synthetically dosed waters. In contrast, the current research used DI water. In this study, the effectiveness of bubble aeration on THMs removal investigated. To examine the feasible treatment strategy for drinking water system, gas stripping testing was conducted at a different
N2 gas flow rate and THMs concentrations. The experiment was conducted at a constant temperature of 22.50C. The initial concentration was 100μg/L for each THM, and the gas flow rates were varying 0.5 to 2 L/min. Fig. 2-3(a-c) shows the remaining liquid phase concentration of THMs (DCBM, DBCM and BF) after bubble aeration. As a result, the remaining DCBM for flow rates 0.5, 1 and 2 L/min were 19, 6 and 0 µg/L, respectively. For DBCM for the same flow rates; the remaining liquid phase concentration was 41, 20 and 3 µg/L respectively. Similarly, for
BF the residual concentrations were 58, 38 and 12 µg/L which was the lowest compared to the others. The figure also shows the prediction values fitted with experimental ones. The removal efficiency of THMs was in direct proportion to the air-to-water ratio [21]. Hence, the increase in flow rate increases efficiency. The reason behind this incidence is that increased air flow rate increases the mass transfer by increasing interfacial area. This will lead to decreasing gas phase resistance and hence increasing the efficiency.
25 | P a g e
2.4.2 Bubble aeration modeling for THMs removal
2.4.2.1Bubble aeration mathematical modeling - I
A mathematical modeling of the removal of THM using bubble aeration was developed to investigate time dependent effects of influent flow rate and Henry's constant for the THM are examined. The model combines mass balance of solute between the liquid and gas phases, and empirical relationships to estimate removals in drinking water treatment tanks by bubble aeration systems. The model also requires the input of various physical and water quality parameters related to the specific storage tank system being modeled. In this case, the distribution equilibrium of THMs in the aqueous and gas phases can be characterized by the dimensionless
Henry’s constant provided in equation (Eqn) 1. To predict the aeration efficiency of THMs concentrations using the dimensionless storage tank Henry’s constant, the following expression is applied using the appropriate expression that was derived from eqn. 1
푉푤 퐶 = 퐻푐 푄푔푎푠…………………………....………………………...…………….…. (2) where Vw is the volume of water, C is the chloroform concentration, HcC is gas phase THMs concentration where Hc is Henry’s constant,
푉푤 푑퐶 = 퐶푚푄 푑푡 − 퐶푄 푑푡 − 퐻푐 푄푔푎푠 푑푡………………………………………………….… (3)
Eqn. 3 shows the mass balance of a tank that was used to derive the mathematical modeling for flowing tanks where Cin is the influent concentration of THMs, Q is the flow of water, and t is time. Equation 3 was then rearranged and simplified to form:
푉푤푑퐶 = −[퐶퐿(푄 + 퐻푐 푄푔푎푠) − 퐶푖푛푄] 푑푡 ……………………………………..……..……… (4)
Solving for CL, eqn.
26 | P a g e
퐶퐿표 (푄+ 퐻푐푄푔푎푠)− 퐶푖푛 푄 퐶푖푛 푄 퐶 = 푡 + ( )…………………………………………….. (5) 퐿 (푄+ 퐻 푄 )/ ) 푐 푔푎푠 푉 푄+퐻푐푄푔푎푠 퐶퐿표 (푄+ 퐻푐푄푔푎푠)푒 푤
For batch systems when the water flow, Q, is zero.
퐻 푄 퐿푛 ( 퐶퐿 ) = − ( 푐 푔푎푠) . 푡 …………………………………………………………………… (6) 퐶퐿표 푉푤 and Simplified to
퐻푐푄푔푎푠 −( ) 푡 푉 퐶퐿 = 퐶퐿푒 푤 ………………………………………………………………...………. (7)
This model was derived based on the mass balance of solutes between the water and air phases for diffuse aeration studies that is similar to previous studies by Sherant et al [5] and Etan et al.
[6].
2.4.2.2 Mathematical modeling of bubble aeration - II
Another approach of determining of the relationship of influent and effluent liquid phase concentration by utilizing the overall mass transfer coefficient for that particular contaminant is stated by Montgomery et al [19]. Exchange of gases between aqueous and gaseous phases is an essential element of many environmental processes. Gas transfer can be used to remove unwanted volatile THMs [22] . Exchange of a dissolved compound with the atmosphere is controlled by the extent of mixing in the aqueous and gaseous phase, the surface area of the interface, the concentration of the compound in the two phases, and the equilibrium distribution of the compound.
The mass transfer from bubbles rising in a well-mixed vessel is model using Eqn 8 [19].
퐶푔 Φ 퐹 = 푄푔푎푠퐻푐 ( − 퐶퐿) (1 − 푒 )……………….………………………….. (8) 퐻푐
27 | P a g e
Based on the principles of gas-transfer applied to bubble aeration, the gas-phase and liquid-phase concentrations of for a rising bubble are related by Equation 9 [8] [16]:
Φ 퐶푔 = 퐶퐿퐻푐 (1 − 푒 )…………………………………………………………...... …….…..… (9)
Where the saturation parameter Ф is defined by:
Φ = − 퐾퐿푎푉푤 ……………………………………………………………………….……….... (10) 푄푔푎푠퐻푐
Where Cg, CL is the gas-phase and liquid-phase concentrations (µg/L), respectively; HC is the
-l dimensionless Henry’s constant, KLa = overall mass-transfer coefficient (min ); Vw = liquid volume (L); Qgas = gas flow rate (L/min). For a particular process conditions the saturation parameter Ф is unique for each compound.
For large values of KLa and small values of HC, the bubble exit concentration approaches saturation. The value of KLa estimation was provided and obtained from Montgomery et al. [19].
퐾퐿푎(푎푖푟) 퐾퐿푎 = − 퐷(푇퐻푀푠)…………………………………………….... ……………. (11) 퐷(푎푖푟)
Where KLa (air) is the mass transfer coefficient of air, DTHMs, and Dair is estimated diffusion coefficient of THMs and air:
푡 −Φ 푡 퐶퐺 = 퐶퐿퐻푐 (1 − 푒 푟)……………………………………………………………..………. (12)
Rearranging Equation (12) for t = tr, the degree of saturation, Sd, of the THM in the bubble:
퐶퐺 −Φ 푆푑 = = 1 − 푒 ……………………………………………………………...………… (13) 퐶퐿퐻푐
Using the liquid-phase mass balance to describe the transfer of THMs from the liquid phase into the gas phase and derive an equation for the change of liquid-phase concentration from initial concentration, CLo to CL with change of time (Δt) as follows:
퐶퐿 퐻 푄푔푎푠 퐿푛 ( ) = − ( ) 푆푑 Δ푡 ……………………………………………..…………..…….. (14) 퐶퐿표 푉푊
28 | P a g e
where fKLa,(the transfer parameter ) = [- ln(1-Sd)]/Sd. Matter-Muller et al. [8] and Roberts et al.
[16] related the mass-transfer rate to three ranges of the saturation parameter. Using a similar approach Fig. 2-4 shows the estimates of Eqn 15 that are simplified over three ranges of degree- of-saturation relating fKLa (transfer parameter) and Φ (saturation parameter) as functions of the degree of saturation (Sd). The three cases are described. Case 1: The transfer parameter fKLa, is equal to 1.05 for Sd = 0.1 and approaches 1 as Sd approaches zero. Eqn (14) can be simplified to the following:
퐶퐿 퐿푛 ( ) = − 퐾퐿푎 Δ푡 …………………………..……………………………………………. (15) 퐶퐿표
This case is true for compounds with high Hc, with high air-flow rates and it is similar to VOC stripping in surface-aeration systems in which the continuous and rapid renewal of fresh air above the water surface is provided, and saturation is negligible [19]. Case 2: For cases where Sd value is close to 1.0, Eqn (14) can be simplified. This is the case in which the exit gas is saturated >99% with respect to the VOC being stripped. This may occur because of low values of Hc or long bubble-retention times. Case 3: For cases where 0.1 < Sd < 0.99, is applied for most VOCs with the exit gas is partially saturated between 10% and 99% of saturation. Eqn (14) must be used to describe this situation. The mass-transfer rate for the VOC is a function of the mass-transfer coefficient as well as the degree of saturation of the VOC in the exit air.
2.4.2.3 Model parameter estimations
A. Henry’s Constant estimation
The equilibrium concentration between gas and liquid phase can be described by Henry’s
[18]. The temperature correction factor provided by Staudinger & Roberts [11] was used to
29 | P a g e
calculate the dimensionless Henry’s constant for the THMs. Eqn 15 shows the temperature correction factor modeled by Nicholson et al. [20].
1 1 퐵 ( − ) 퐻푐(푇) = 퐻푐(20퐶) [퐿표푔 10 푇 293 ]……………………………………………………..16) where Hc(T) is the temperature corrected Henry’s constant, and B is the temperature dependent relationship constant which was experimentally determined [20]. Table 2-2 shows the corrected
o HC values for the THMs at 22.5 C. Comparing experimental data from bench scale equilibrium tests with theoretical values of Hc from Eqn. 16 shows the values differ by less than 1%.
B Mass transfer coefficient
To remove contaminants from water using air stripping, the concentration gradient that exists at the gas liquid interface is the driving force that supports the mass transfer, KLa, [9].
These similarities and differences could be predicted from the relative magnitudes of the diffusivities of THM in water [10], using diffusion coefficient estimation from literature [3]. The overall mass transfer coefficient was then computed using Eqn 15. The diffusion time of a droplet is simply the drop terminal axial velocity times the falling distance. Mass transfer coefficients for constituents were computed from the ratio of the diffusivity of oxygen to the compound of interest. The KLa of volatile constituents were found to be related to (KLa) Oxygen by
Eqn 14 provided by Roberts et al. [10]. The diffusivity estimation for any compound and the viscosity of water at different temperature are given in Eqns 17 and 18 following Warren et al.
[3]. Table 2-3 and 2-4 represents the calculated, viscosity, Vw , diffusivity, DBW, and KLa values for the THMs at different temperature.
13.26 (10−5) 퐷퐵푊 = 1.14 ……………………………………………………………………. (17) 휂푤 푉퐵
30 | P a g e
퐵 휂푤 = 퐴 10푇−퐶 …………………………………………...…………………………………… (18)
The constants values obtained from literature, A = 2.414*10-5 Pa.s, B = 24.7 K, and C = 140 K
2.4.4 Comparing bench scale bubble aeration study with model prediction
The experimental and mathematical modeling estimates for the removal efficiencies of
DCBM, DBCM and BF for selected gas flow rates of 0.5, 1, and 2 L/min fed to a 2 L vessel are shown Fig. 2-3 (a-c). For this model the initial THMs concentration of 100 ug/ L. The modeling results matched more closely with the experimental values for stripping of DCBM compared to
DBCM and BF. The predicted residual concentrations of DCBM for the flow rates of 0.5, 1 and
2 L/min after one hour aeration, were 27, 7 and 0.5 µg/L, respectively. For DBCM for the same flow rates the predicted remaining liquid phase concentration were 54, 29 and 8 µg/L respectively. The residual predicted values for BF were 79, 54 and 29 µg/ L were 0.5, 1, and 2
L/min. Fig. 2-5 shows a three-dimensional plot of predicted removal efficiency of THM at temperatures ranging from 100C to 400C and retention time from the initial concentration of 100
μg/L for each THM. The air flow rate was 2 L/min and the retention time was 10 mins. The removal efficiencies were directly proportional to THM volatilities. These differences between
o the THMs narrowed for temperature above 30 C. The predicted degree of saturation, Sd, values for this study were between 0.1 and 0.99.
To investigate THMs volatility effect on the aeration, investigations have been conducted on the relationship between transfer and saturation parameters. Fig. 2-5 represents the relationship between the transfer parameters and the saturation parameters as a function of the degree of saturation (Sd) for temperatures between 10 and 40 oC. From the figure, THMs with
31 | P a g e
higher HC like CF, showed wider gap in between the two parameters as a function of Sd which was influenced by HC and temperature. Their influence of on the removal THMs also proved with Staudinger and Roberts, and Ethan Brooks [6, 23].
2.4.5 Bubble aeration and temperatures stratification in large tank
Temperature differences within a large storage were measured over four months (Fig.2-
2). Differences within storage facility indicate possible stratification as ambient temperatures increase which could create stagnant zone. Stratification is also an early warning sign of potential microbial problem. Bubble aeration helps to reduce temperature stratification and enhance removal of THM. Mathematical modeling was used to estimate bubble aeration for the removal of THM at different temperatures of the storage tank at different retention time. Fig.2-6 represents the liquid phase concentration of CF after aeration at different temperatures for known initial concentration of 100 µg/L of a storage tank. As shown in the figure, it can be concluded that aeration can remove THMs significantly. Fig.2-7 shows a 3D mesh plot which presents the removal efficiencies of CF and BF at different temperature with a given retention time within the tank. As the result the removal efficiency of BF is much lower than that of CF. Sherant et al [5] also stated that when THM species became more bromine substituted, their removal efficiency became lower. In general, increasing the bromide level in water increases the level of brominated THMs. This indicates that a lower removal efficiency for TTHM is expected for water high in bromide [5].
32 | P a g e
2.4.6 Quantitative risk analysis for THMs
The reduction of THM using aeration can be assessed based on the decrease in quantitative risk analysis. The analysis includes hazardous identification, exposure assessment, toxicity assessment, and risk characterization. This risk assessment method could be applied for both carcinogens and non-carcinogens [24] to estimate the risk of chronic exposure to THMs and the reduction as the result of bubble aeration. According to U.S. EPA parameter, THMs exposure to the body was assessed from the eqn.19 [24],
( ) 퐼 = 퐶푅 .(퐸퐹).(퐸퐷) 퐶 …………………………………………..………………………. (19) (퐵푊)(퐴푇)
Where, I is the intake rate of THMs (mg/Kg/day), CR is the rate of contact with the THMs
(L/day), EF is the frequency of exposure to the THMs (day/ year), ED is the duration over which exposure is averaged (day), BW is the average body weight (kg), and C is the average THMs concentration during the exposure period (mg/L). Exposure is defined as human contact with
THM species through ingestion, inhalation and dermal intake pathways evaluated based on chronic daily intake [25] [24].
The potential non-carcinogenic and carcinogenic health risks associated with ingesting can be obtained from Eqns 20 and 21. The exposure assessment procedures are summarized and integrated into quantitative and qualitative expressions of the risk level. For carcinogenic effects, the risk is expressed as the probability that an individual will exhibit dose-response characteristics.
Iing,noncar HQing = …………………………………………………………….………………..(20) 푅푓퐷표푟푎푙
IELCRing = Icarc ∗ 푆퐹 ……………………………………………………………………… (21)
33 | P a g e
Where, HQ is hazard quotient (unit less) and calculated from the non-carcinogenic intake rate
(Iing,noncar) and the reference dose factor (RfD) was estimated based on the daily exposure of the human population to a potential hazard that is likely to be without risk of deleterious effects during a lifetime. IELCRing is the individual excess lifetime cancer risk from the carcinogenic intake, Icar and the cancer slope factor, SF. SF is an estimate of the increased cancer risk from oral exposure to a dose of 1 mg/kg-day for a lifetime. In addition, USEPA [24] has recommended various values for substitution in these formulas such as CR 2 L/ Day, EF = 350 day/ y, ED = 30 years, BW = 70 kg, and AT = 365 days/ year *(70 years). Estimated oral RfD and SF for THMs are given Table A. These estimates are obtained from www.epa.gov/iris [26].
The estimate values for RfD and SF for THMs are given in Table 2-5.
The potential for non-cacogenic and carcinogenic health risks of chloroform were calculated for different concentration obtained based on Agency for Toxic Substances and
Disease Registry (ASTDR) [27]. As stated in ASTDR, the amount of chloroform ranges from 2 to 44 µg / L in treated drinking water. However, in some places, chloroform concentrations may be higher. As much as 88 µg / L was found in treated municipal drinking water [27]. Drinking water derived from well water near a hazardous waste site contained 1,900 µg /L. Surface water containing 394 µg / L has also been found near a hazardous waste sites [27][32].
Fig. 2- 8 represents the concentrations of chloroform before and after aeration, and the potential health risk effect assessment of non-carcinogenic and carcinogenic. Estimations were calculated before and after aeration. As a result, the potential non carcinogenic health risk effect
(HQ) for chloroform before aeration ranges from 1.2*10-1 to 5.2 and after aeration 1.2*10-3 to
5.2*10-2. After aeration data are in acceptable range. The USEPA considers that HQ < 1 is to be acceptable for regulatory purposes[24]. The values for carcinogenic health risk effects (IELCR),
34 | P a g e
before aeration ranges from 3.15*10-6 to 1.4*10-4. After aeration IELCR decreased in the range of 3.15*10-8 to 1.36*10-6 which is again the after-aeration results are in the acceptable range.
The USEPA considers that the acceptable IELCR is ≤ 10-6- 10-4 [24].
2.5 Conclusions
The objective of this study was to investigate aeration as a method for controlling THMs from drinking water storage tanks. Airflow had the most significant effect on THM removal from water. In the aeration study, as the air flow increased from 0.5 to 2 L/ min in the batch study, the removal of THMs risen from 81% to 99% for DCBM, 59% to 97% for DBCM and
42% to 88% for BF. The tank temperature gradient study showed as aeration can remove the
THMs significantly. The mathematical modeling of this study provides for the assessment of potential impacts of aeration on gas flows. As a result, as the airflow increased, the experimental data points tends to perfectly agree to the predictive outputs. Finally, the quantitative risk analysis for CF revealed that after aeration, the values for potential non-cacogenic/ cacogenic are within the USEPA’s acceptable range. Thus, aeration significantly reduces chloroform concentration to safe drinking water levels.
Acknowledgments
The work conducted was partly supported by the contract number EP11C000147 obtained from the EPA Path Forward Innovation Project from the EPA-University of Cincinnati Grants
Program.
Disclaimer
The views expressed in this article are those of the authors and do not reflect the official policy or position of the Unites State Environmental Protection Agency. Mention of trade names,
35 | P a g e
products, or services does not convey official EPA approval, endorsement, or recommendation.
This manuscript has been subjected to the Agency’s review and has been approved for
publication.
Table 2-1: Physical properties of the THMs
Melting Boiling Vapor Henry's const. THM Vap. Pres Specific Solubility MW point point Density (atm.m3/mole) Name ( mmHg) Gravity (mg/L) (OC) (0C) (g/L) @ 20 OC
CF 119.38 -64 62 160 4.12 1.49 7840 3.1*10-3 DCBM 163.8 -57.1 90 50 6.7 1.9771 4500 2.12*10-3 DBCM 208 -22 120 50 8.5 2.451 2700 8.4*10-4 BF 252.77 -4 149 5 8.7 2.89 3130 5.84*10-4 Source Groundwater Chemicals Desk Reference, 3rd Edition
36 | P a g e
Table 2-2: comparison of HC experimental and calculated
Hcc Temperature 0 0 Hc @ 22.5 C given Correction Hc @ 22.5 C experimental % THMs Factor at 20 oC corrected (Average) deference DCBM 0.076 2130 0.0876 0.0883±0.008 0.07% DBCM 0.035 2273 0.0407 0.0502±0.003 0.95% CHBr3 0.0175 2120 0.0201 0.0291±0.003 0.90%
37 | P a g e
Table 2-3: Estimated diffusivity of THM as a function of viscosity of water at different temperature
THMs – Diffusivity in Water (m2/s) T (0C) ᶯw CF DCBM DBCM BF 10 1.307 6.8E-06 6.7E-06 6.6E-06 6.50536E-06 15 1.139 7.95E-06 7.83E-06 7.72E-06 7.61007E-06 20 1.002 9.21E-06 9.07E-06 8.93E-06 8.80717E-06 22.5 0.9469 9.82E-06 9.67E-06 9.53E-06 9.39375E-06 25 0.8904 1.05E-05 1.04E-05 1.02E-05 1.00762E-05 30 0.7975 1.19E-05 1.18E-05 1.16E-05 1.14249E-05 35 0.7171 1.35E-05 1.33E-05 1.31E-05 1.28963E-05 40 0.653 1.5E-05 1.48E-05 1.46E-05 1.43491E-05 45 0.5943 1.67E-05 1.64E-05 1.62E-05 1.59757E-05
38 | P a g e
Table 2-4: Estimated KLa as a function of diffusivity for THMs at different temperature
KLa , 1/min Temp. 0C CF DCBM DBCM BF 10 0.0333 0.0328 0.032320287 0.031859 15 0.038955 0.03837 0.037808751 0.037269 20 0.045083 0.044406 0.043756258 0.043131 22.5 0.048085 0.047364 0.046670525 0.046004 25 0.051579 0.050805 0.050061321 0.049347 30 0.058482 0.057605 0.056761831 0.055951 35 0.066014 0.065024 0.064072032 0.063157 40 0.073451 0.072349 0.071289957 0.070272 45 0.081777 0.08055 0.079371182 0.078238
39 | P a g e
Table 2-5: Estimated oral reference dose factor (RfD) and SF (cancer slope factor) for oral ingestion of THMs
THMs RfD, SF, (mg/kg/day) (mg/kg/day) -1
Chloroform 1*10-2 6.1*10-3
Dichlorobromomethane 2*10-2 6.2*10-2
Dibromochloromethane 2*10-2 8.4*10-2
Bromoform 2*10-2 7.9*10-3
40 | P a g e
Qgas
Figure 2- 1: Schematic diagram of reactor used for the experimental bubble aeration study
41 | P a g e
18 O
8 O
Figure 2-2 Temperature gradient with respect to depth of the storage tank (Over 40 days from 4/4/2014 – 5/14/2014)
42 | P a g e
Figure 2-3 Liquid phase removal efficiency after selected times, a) DCBM, b) DBCM and c) BF
43 | P a g e
Figure 2-4Transfer (TP) vs Saturation parameter (SP) of THMs at 10 and 40 0C : a) CF, b) DCBM, c) DBCM and d) BF
44 | P a g e
Figure 2-5 Model Estimation of removal efficiencies of THMs as function of temperature
45 | P a g e
Figure 2-6: Liquid phase concentration of CF in storage tank at different given temperature using 2 L/min
46 | P a g e
Figure 2-7: Model estimation of removal efficiency for CF and BF in storage tank
47 | P a g e
Figure 2-8 Chloroform before and after aeration, potential health risk effect assessment of non- carcinogenic hazard quotient (HQ) and carcinogenic individual excess lifetime cancer risk (IELCR).
48 | P a g e
2.6 References 1. Durmishi, B.H., et al., Disinfection of Drinking Water and Trihalomethanes: A Review. International J. Advan. Res. Chem. Sci. (IJARCS), 2015. 2(11): p. 45-56. 2. Myerchin, G., D. White, and C. Woolard, Disinfection By-Product Formation during Long-Term Water Storage in Alaska. Journal of Cold Regions Engineering, 2006. 20(4): p. 168-180. 3. Lyman, W.J., et al., Methods for Estimating Physicochemical Properties of Inorganic Chemicals of Environmental Concern. 1984, DTIC Document. 4. Chen, W.J. and C.P. Weisel, Halogenated DBP concentrations in a distribution system. American Water Works Association. Journal, 1998. 90(4): p. 151. 5. Sherant, S.R., Trihalomethane control by aeration. 2008, The Pennsylvania State University. 6. Brooke, E. and M.R. Collins, Posttreatment aeration to reduce THMs. American Water Works Association. Journal, 2011. 103(10): p. 84. 7. Pfafflin, J.R. and E.N. Ziegler, Encyclopedia of environmental science and engineering. Vol. 2. 1992: Taylor & Francis. 8. Matter-Müller, C., W. Gujer, and W. Giger, Transfer of volatile substances from water to the atmosphere. Water Research, 1981. 15(11): p. 1271-1279. 9. Smith, C.T., Trihalomethane Removal and Re-Formation in Spray Aeration Processes Treating Disinfected Groundwater. 2015, University of Central Florida Orlando, Florida. 10. Lang, L., et al., Strategies for Controlling and Removing Trace Organic Compounds Found in Potable Water Supplies at Fixed Army Installations. 1985, DTIC Document. 11. Uyak, V., K. Ozdemir, and I. Toroz, Seasonal variations of disinfection by-product precursors profile and their removal through surface water treatment plants. Science of the total environment, 2008. 390(2): p. 417-424. 12. Ozdemir, C. and S. Dursun, Trihalomethane determination and removals from the main discharge channel of Konya City (Turkey). Environmental technology, 2004. 25(9): p. 1091-1096. 13. Tarquin, A.J., et al., Removal of THMs from Drinking Water Using an Induced Draft Stripping Tower. Presentation at University of Texas at El Paso, 2005. 14. Crittenden, J.C., et al., MWH's Water Treatment: Principles and Design. 2012: John Wiley & Sons. 15. and A.S.o.C.E.A.W.W. Association, Water treatment plant design, ed. r. ed. 1998, New York: McGraw-Hill. x, 806. 16. Roberts, P.V., et al., Volatilization of organic pollutants in wastewater treatment: Model studies. NTIS, SPRINGFIELD, VA(USA). 1984., 1984. 17. Roberts, P.V. and P.G. Daendliker, Mass transfer of volatile organic contaminants from aqueous solution to the atmosphere during surface aeration. Environmental science & technology, 1983. 17(8): p. 484-489. 18. Hand, D.W., D.R. Hokanson, and J.C. Crittenden, Air stripping and aeration. Water Quality and Treatment, a Handbook of Community Water Supplies, American Water Works Association, 1999. 19. Montgomery, J.M., Consulting Engineers, Inc.(1985) Water Treatment Principles and Design. John wiley & sons Inc. USA. 1(1): p. 6-1.
49 | P a g e
20. Nicholson, B., B.P. Maguire, and D.B. Bursill, Henry's law constants for the trihalomethanes: effects of water composition and temperature. Environmental science & technology, 1984. 18(7): p. 518-521. 21. Cecchetti, A.R., H. Roakes, and M.R. Collins, Influence of selected variables on trihalomethane removals by spray aeration. JOURNAL AWWA, 2014. 106: p. 5. 22. Zander, A.K., M.J. Semmens, and R.M. Narbaitz, Removing VOCs by membrane stripping. Journal (American Water Works Association), 1989: p. 76-81. 23. Staudinger, J. and P.V. Roberts, A critical compilation of Henry's law constant temperature dependence relations for organic compounds in dilute aqueous solutions. Chemosphere, 2001. 44(4): p. 561-576. 24. EPA, A., Risk Assessment Guidance for Superfund. Volume I: Human Health Evaluation Manual (Part A). 1989, EPA/540/1-89/002. 25. USEPA, U., Exposure factors handbook. Office of Research and Development, Washington, 1997. 26. Nazaroff, W.W.A.-C., Lisa, Environmental engineering science. 2001, New York: Wiley. 690. 27. ASTDR, Agency for Toxic Substances and Disease Registry. 1997.
50 | P a g e
3. Performance of Anaerobic Biotrickling Filter and its Microbial Diversity for the Removal of Stripped Disinfection Byproducts
3.1 Abstract The objective of this research was to evaluate the biodegradation of chloroform (CF) by using biotrickling filter (BTF) and determining the dominant bacteria responsible for the degradation. The research was conducted in three phases under anaerobic condition, namely, in the presence of co metabolite (Phase I), in the presence of co metabolite and surfactant (Phase II) and in the presence of surfactant but no co metabolite (Phase III). The results showed that the presence of ethanol as a co metabolite provided 49% removal efficiency. The equivalent elimination capacity (EC) was 0.13 g/(m3.hr). The addition of Tomadol 25 - 7 as a surfactant in the nutrient solution increased the removal efficiency of CF to 64% with corresponding EC of
0.17 g/(m3.hr). This research also investigated the overall microbial ecology of the BTF utilizing culture-independent gene sequencing alignment of the 16S rRNA allowing identification of isolated species. Taxonomical composition revealed the abundance of betaproteobacteria and deltaproteobacteria with species level of 97%. A. oryzae (formally dechlorosoma suillum), A. restrica and Geobacter spp. together with other similar groups were the most valuable bacteria for the degradation of CF.
This chapter is based on a publication: Bineyam Mezgebe, George A. Sorial , E. Sahle-Demessie, Ashraf Aly Hassan, and Jingrang Lu. 2017. "Performance of Anaerobic Biotrickling Filter and Its Microbial Diversity for the Removal of Stripped Disinfection By-products." Water, Air, & Soil Pollution: JrnlID 11270_ArtID 3616
51 | P a g e
3.2 Introduction
Disinfection byproducts (DBPs) are formed as the results of the reaction of free chlorine and a series of complex organic precursors [1]. Many DBPs are carcinogens or have been known to cause other health risks [2]. The precursors forming DBPs range from natural occurring humic and fulvic material to anthropological contaminants that persist in treated water
[3]. The highest concentrations of DBPs detected in drinking water constitute trihalomethanes
(THMs) where CF is the major component [2]. The International Agency for Research on Cancer
(IARC) has determined that CF is possibly carcinogenic to humans [4]. Under the Safe Drinking
Water Act (SDWA), the United States Environmental Protection Agency (USEPA) establishes a
Maximum Contaminant Level (MCL) of 70 ppb for CF [5]. Several physical and chemical removal methods such as adsorption, advanced oxidation [6], and air stripping [7] are used to treat chloroform. CF is a volatile organic compound (VOC) that readily evaporates to the atmosphere from treated water surfaces through air stripping [8]. Hence, the need for innovative control technology is becoming more enviable. In this research, biological treatment was used to treat the off-gas CF by bio-trickling filters (BTF).
Most of the research on the biological treatment of CF has been limited to batch liquid phase processes at wastewater treatment plants or hazardous waste disposal sites [9]. Yoon and
Park (2002) studied the effect of gas residence time on the aerobic biodegradation of nine VOCs including CF. In their study they observed that CF showed lower removal efficiency compared to other compounds for longer empty bed residence time of 3 minutes [10]. However, under anaerobic conditions, CF could undergo a reductive biotransformation by pure cultures of methanogens [11, 12], acetogenic bacteria [13], sulfate-reducing bacteria [14] and iron-reducing bacteria [14, 15] producing partial dehalogenation and mineralization [12-15]. Highly
52 | P a g e
chlorinated atoms are good electron targets for anaerobes since chlorine atoms block the activity of oxygenase. Thus, biological techniques have resulted in either partial dechlorination of CF to dichloromethane [14, 16] or complete mineralization to methane and carbon dioxide [14, 17, 18].
Although most studies show successful biodegradation of CF in the liquid phase, there has been limited reported work on the use of biofiltration for the removal of CF from gaseous streams.
Biofiltration is one of the proven technologies for removing VOCs form high volume stream as it is environmentally friendly, cost effective and releases fewer byproducts [19]. The use of aerobic biofiltration technique has been reported for the biotreatment of CF with other mixtures of different VOCs [19, 20]. Yoon et al. (2002) have shown the degradation potential of nine VOCs including CF and found the highest removal was for toluene (99%) and the lowest removal was for CF (89.4%). Similarly, Balasubramanian et al. (2012) evaluated the biodegradation of CF along with a mixture of VOCs commonly found in pharmaceutical emissions, using a biotrickling filter [20]. Their study showed that increasing the rate of CF loading significantly reduced the degradation efficiency of the reactor for the mixture of VOCs. The increase in CF loading rate from 14.22 g/(m3.hr) to 55.83 g/(m3.hr) and further to 107.85 g/(m3.hr) affected the efficiency of both CF as well as the other VOCs [20]. However, there has been a limited research work reported on the use of anaerobic BTF for the removal of CF from the gas phase at low concentration. Multiple investigations have been performed on co-metabolism, specifically for the treatment of trichloroethylene (TCE) contaminated air using BTF. Chlorinated compounds can be degraded by using a more easily biodegradable non-chlorinated hydrocarbon serving as a co metabolite [20]. Biofilters using methane, butane, propane, propylene, phenol, toluene, or ammonia as a co metabolite have shown success in the treatment of trichloroethylene
53 | P a g e
(TCE) [21]. Mahon S. (2013) observed a 25% destruction of TCE by feeding both the propane and the TCE into the biofilter [21].
CF is a halogenated compound and recalcitrant to biodegrade. As halogenated organics, it’s halogen bond can significantly affect the biofiltration process [22]. Hence, the biodegradation of higher order chlorinated volatile compounds like CF occur mostly under co- metabolic conditions in the presence of primary substrates such as methane, propane, phenol and toluene [23, 24].
A nonionic surfactant can be introduced in the biofiltration system as means for enhancing solubility. Hence, surfactant could increase the apparent solubility of CF by micellar formation, which commences at the critical micelle concentration and the apparent solubility is proportional to surfactant concentration [25]. The use of surfactants in enhancing the bioavailability of hydrophobic compounds by facilitating their biotransformation under aerobic conditions has been studied by several researchers [26-28]. The utilization of surfactant increases the solubility which overcomes the rate limiting step [29]. Researchers have found that addition of surfactants stimulated polycyclic aromatic hydrocarbon (PAH) biodegradation [30-32].
However, there are limited studies where surfactant was used for the dehalogenation in biofiltration system. The use of surfactants in BTF enhanced n-hexane [33], toluene [34], and styrene [35] degradation in fungi biofilter. Yuan et al. (2010) also reported in their study that nonionic surfactant such as TX-100 greatly increased the dechlorination of chlorobenzenes [36].
In the present work, an effort was made to evaluate the performance of anaerobic BTF for biodegradation of CF. The experimental plan was designed to operate the BTF in the presence and absence of co metabolite and surfactant at different phases. The research also included the application of molecular tools for characterization of the microbial community diversity
54 | P a g e
throughout the BTF. The BTF was used for the continuous removal of gaseous CF at three different phases and to gain an insight into the effects of halorespiration for dehalogenation of
CF. The change in the biodiversity of the bacterial community during the three phases helps to identify the most abundant bacterial populations responsible for the dehalogenation of CF during the operation of the BTF.
3.2 Materials and methods
3.2.1Materials
Chloroform was used as a model THM compound; ethanol laden nitrogen gas stream was supplemented as co metabolite and Tomadol 25 – 7 was used as a surfactant for the biodegradation. CF with 99.8% purity obtained from Fisher Scientific (Pittsburgh, PA, USA),
Ethanol with 99.5% purity obtained from Sigma Aldrich (St. Louis, MO, USA) and Tomadol 25
– 7 was obtained from Tomah products (Tomah products, LA). CF is highly hydrophobic with a
−3 3 Henry’s law constant, KH of 3.5×10 atm.m /mol, and the KH value of a hydrophilic ethanol is known to be 5.1×10−6 atm.m3 / mol at 25 °C. The measuring sensors for pH, nitrate, dissolved oxygen (DO), and ammonia were acquired from Accumate Instruments. Genomic DNA extraction of bacterial strains was performed using the Mo Bio PowerSoil DNA (M Bio Lab,
Inc., Carlsbad, CA) Kit which was done by Molecular Research LP (MR DNA, Shallowater,
TX).
55 | P a g e
3.2.2 Experimental methods
3.2.2.1Biotrickling filter bed experiment
The experimental set-up for BTF consisted of a cylindrical glass column with multiple sections providing a total length of 130 cm and an internal diameter of 7.6 cm. A schematic set up of the experimental system is shown in Fig.3-1. The mid-portion of the column was packed with pelletized diatomaceous earth biological support media, to a depth of about 60 cm long
(Celite® 6 mm R-635 Bio-Catalyst Carrier; Celite Corp., Lompoc, CA). The BTF was operated in a co-current mode where CF and co metabolite laden gases, and liquid nutrient solution were introduced into the column from the top. The gas flow was controlled with two mass flow controllers (Sierra Instruments). The growth media of essential nutrients and vitamins was prepared with medium concentrations of 996 mg/L NH4Cl, 414 mg/L KH2PO4, 390 mg/L
MgCl2.6H2O, 280 mg/L CaCl2.2H2O, 2 mg/L FeCl2.4H2O, 4.79 mg/L CuSO4.5H2O, 6.53 mg/L
MnSO4.H2O, 5.24 mg/L ZnCl2, 4.58 mg/L CoCl2.6H2O, 0.32 mg/L B(OH)3, 4.79 mg/L
NiCl2.6H2O, 0.12 mg/L 4-aminobenzoic acid (99%), 0.048 mg/L biotin, 0.0024 mg/L cyanocobalamin, 0.05 mg/L, folic acid dihydrate (99%), 0.12 mg/L nicotinic acid (98%), 0.12 mg/L pantothenic acid Ca-salt hydrate (98%), 0.24 mg/L pyridoxine hydrochloride (98%), 0.12 mg/L riboflavin (98%), 0.12 mg/L thiamine hydrochloride (99%), and 0.12 mg/L thioctic acid
(98 %). The composition of the nutrient solution was similar to the one used by Atikovic et al.
[37] and Wu et al. [38].
The BTF was seeded with methanogenic microorganisms and maintained at 35oC.
Initially, these bacteria were obtained from nutrient enriched solution kept under a blanket of nitrogen gas that was acclimated in our lab to CF in a 4 liter amber batch reactor for two months.
The CF feed was step-wise increased from 5 to 50 ppmv within the two-month period. This
56 | P a g e
inoculum was mixed in the ratio of 1:1 with other methanogenic bacteria acquired from another bioreactor that was treating food waste prior to seeding the BTF. A mixture of CF and ethanol
(co metabolite) was continuously fed into the BTF in a flowing nitrogen carrier gas at a flow rate of 0.5 L/min with a corresponding empty bed residence time of five minutes. The nutrient solution buffered at pH 7 was fed intermittently to the BTF bed by a solenoid valve at a rate of
2.0 L/day. The presence of co metabolite enabled treatment strategies that stimulated biodegradation of CF, which alone was not enough to provide the carbon or energy benefit to the microorganism. To further improve, the removal efficiency of BTF’s for CF degradation, a non- ionic surfactant (Tomadol 25 – 7) was used.
3.2.2.2 Analytical methods
Gas and liquid samples were collected daily from the BTF system five days per week for the measurements of the ammonia -N the composition of feed and effluent gas / liquid streams.
The total number of samples for phase I, II and III were 44, 97 and 95 respectively.
Measurements of the effluent liquid pH, and organic matter, the gas flow pressure drop across the bed, and operating temperature were taken. Gas phase samples were taken on-line from different points along the BTF column using an electrically controlled low-bleed eight-port
Valco valve and analyzed by gas chromatograph. The samples were analyzed for CF, ethanol, and by-products such as methane and carbon dioxide. They were injected into GC – HP,
Column: HP, 608, 30 m X 530 μm film thickness, injection splitless through 1ml sample loop equipped with a flame ionization detector (FID). The GC oven was programmed isothermal at
60 °C (2min) ramped to 90 °C at a rate of 10 oC/min. The carrier gas (He) flow rate was set at
3.5 mL/min at constant flow rate. The FID was used with N2 make-up gas at a flow rate of 30
57 | P a g e
mL/min, a fuel gas flow (H2) of 40 mL/min. Retention time for CF was 3.8 min under the above conditions used. Levels of reaction products, such as carbon dioxide (CO2) samples were also taken automatically by GC HP- TCD from each sampling port in the BTF. The GC oven was programmed isothermal at 60°C (1min), ramped to 115°C at 25°C /min. The carrier gas (He) flow rate was set at 3.5 mL/min, the TCD was used with N2 make-up gas at a flow rate of 5 mL/min, a fuel gas flow (H2) of 30 mL/min and airflow 400 L/min.
Liquid samples were collected from the effluent stream of BTF once a week and analyzed for volatile suspended solids (VSS) and total organic carbon (TOC). The samples were filtered through a 0.45 µm membrane filter (Whatman Co.) and analyzed for influent and effluent concentrations of ammonia, dissolved total carbon, dissolved inorganic carbon, and volatile suspended solids. The concentration of ammonia was determined using an ammonia electrode sensor. Dissolved total carbon and dissolved inorganic carbon content of the liquid samples were determined with a Shimadzu total organic carbon analyzer model TOC - L (Shimadzu
Corp., Tokyo, Japan). The volatile suspended solids analysis was conducted by Standard
Method 2540G [39].
3.2.2.3 Microbial community molecular analysis
Biofilm samples were collected from the BTF within the media as shown in Fig. 3-1. The samples were taken from port 2 (first port from the top within the media) at the end of each phase before proceeding to the next phase. This method was previously done by other researchers [40, 41]. The samples consisted of about five media pellets covered with biomass and placed in O2 free sampling tubes. All the samples collected were stored in a -20 °C freezer until being sent to molecular research laboratory (Molecular Research LP Shallowater, TX) for
58 | P a g e
biological analysis. The DNA of microbial mass in the samples was extracted using Mo Bio
PowerSoil DNA (M Bio Lab, Inc., Carlsbad, CA) following manufacturer’s instruction that includes cell breakage steps followed by the addition of detergents and high salt buffers and enzymatic digestion with lysozyme and proteases. For ion torrent sequencing, the 16S rRNA gene V4 variable region PCR primers 515/806 were used in a single-step 30 cycle PCR using the
HotStarTaq Plus Master Mix Kit (Qiagen, USA) under the following conditions: 94°C for 3 minutes, followed by 28 cycles (5 cycle used on PCR products) of 94°C for 30 seconds, 53°C for
40 seconds and 72°C for 1 minute, after which a final elongation step at 72°C for 5 minutes was performed. Sequencing was carried out at Molecular Research LP (www.mrdnalab.com,
Shallowater, TX, USA) on an Ion Torrent PGM following the manufacturer’s guidelines.
Sequence data were processed using a proprietary analysis pipeline. Sequences were first depleted of barcodes and primers, and those under 150bp or with ambiguous base calls or with homopolymer runs exceeding 6bp were removed. Operational taxonomic units (OTUs), which were defined by clustering at 3% divergence (97% similarity) [42-46], were generated after denoising sequences and removing chimeras. The last OTUs were taxonomically classified using
BLASTn against a database derived from RDPII (http://rdp.cme.msu.edu) and NCBI
(www.ncbi.nlm.nih.gov) [47].
3.3. Results and discussion
3.3.1 Biotrickling filter performances
In this research, the effects of co metabolite and surfactant on the performance of BTF were evaluated. Even though CF is a recalcitrant compound to biological transformation under conventional aerobic conditions, it can be transformed in the presence of a co metabolite under
59 | P a g e
aerobic and anaerobic environments [48]. The co metabolite was mixed with CF stream to achieve higher removal efficiency by providing additional electron donor to the micro- organisms. Anaerobic dehalogenation of CF has been observed by different researchers by using methanogenic microbes with electron donating co metabolites in reductive CF biotransformation
[17, 49-51]. Ethanol was used as a co metabolite since it readily mixes with CF and water, and is non-toxic to microbial community. In Phase I, the BTF started up with CF influent concentration of 5 ppmv and ethanol concentration of 25 ppmv providing a corresponding CF loading rate of 0.27 g/(m3.hr). The operating conditions and different phases of operation are summarized in Table 3-1. It is worth noting that the removal efficiency of ethanol was always above 98% for the given loading rate conditions studied. Therefore, emphasis is placed on the performance of the BTF on CF removal. Daily dehalogenation rate of the BTF was used to track the performance and robustness of the BTF. Fig. 3-2 presents a statistical summary of the removal efficiency as a box plot at different loading rates. The lower boundary of the box denotes the lower quartile, a line within the box marks the median, and the boundary of the box furthest from zero indicates the upper quartile. Whiskers (error bars) above and below the box indicate the 90th and 10th percentiles. The BTF was run for 44 days under the conditions of
Phase I, and the average removal efficiency for this Phase was 49±8%, which provided an average EC of 0.13±0.02 g/(m3.hr) (Table 3-1). CF is used as an electron acceptor by the reducing bacteria and dehalogenation was enhanced by adding electron donor supply. In previously reported studies, co metabolite assisted degradation of c CF were all conducted in batch liquid phase reactions. Thus, there are limited studies on the use of co metabolite for a continuous gas phase studies. Significant CF anaerobical degradation was obtained by Bouwer et al (1981) in seeded cultures with initial concentrations at less than 100 ppb and observed over
60 | P a g e
90% removal efficiency in the presence of methanol. It was also observed that lower fractional degradation of CF was obtained as the initial concentration increased. Gupta et al. (1996) investigated the use of acetic acid as a co metabolite for CF biotransformation and achieved more than 99% of removal efficiency in a methanogenic environment in a chemostat [52].
To further improve the dehalogenation of CF, a nonionic surfactant was mixed with the nutrient feed solution. Hence, in Phase II, a surfactant was added along with the co metabolite.
A biodegradable nonionic surfactant, Tomadol 25 - 7 was mixed with the nutrient solution to obtain a concentration of 75 mg/L that corresponds to 0.5 of the critical micelle concentration. It is worthwhile to note that the behavior of this BTF was very similar to the reported performance of n-hexane trickle bed air biofilter using Tomadol 25 – 7 [29]. The microorganisms utilize surfactants as substrates for energy and nutrients during biodegradation process [53]. Phase II test was run for 97 days to give the microbes sufficient time to acclimate and attain a stable condition. The increased run time allowed the BTF more cell production and increase of biomass throughout the bed. As shown in Fig. 3-2, this condition resulted in an increase of removal efficiency to 64±8% with EC of 0.17±0.02 g/(m3.hr) (Table 3-1). In a previous study, the use of Tomadol 25 – 7 improved the performance of BTF by doubling the elimination capacity for n-hexane, allowing a robust operation, and decreasing the fluctuation of the effluent stream n-hexane concentrations [29]. The amount of methane produced in our research, during
Phases I and II, were 0.15 mg/day and 0.25 mg/day, respectively, which illustrated the more enhanced performance of the BTF in the presence of surfactant.
During phase III, the BTF was augmented with surfactant only without a co metabolite.
This phase has the same operating conditions as Phase II and was run for 95 days. The removal efficiency was higher than that of Phase I and the overall dehalogenation of CF was high in the
61 | P a g e
absence of ethanol. The removal efficiency was 60±7% with EC of 0.16±0.02 g/(m3.hr) (see
Table 3-1). The methane production for phase III was 0.17 mg/day. The drop in the performance compared to the previous Phase (II) could be due to the limitation of biomass growth as was depicted by measurements of VSS.
3.3.2. Dehalogenation pathways and kinetics for the different phases
The reductive dehalogenation of CF along the BTF was measured weekly by collecting gas samples from ports that are located at 7.6 cm, 23 cm, 38 cm, 53 cm and 60 cm down from the top of the packed bed (i.e., ports 2-5 Fig. 3-1). Biological reductive dehalogenations of chloromethanes are co-metabolic processes, or they may be coupled to energy conservation, where anaerobic bacteria use a specific halo organic compound as electron acceptor of a respiratory process. In the absence of byproduct formation, the hydrodehalogenation and hydrogenation of CF can be hypothesized as in Fig 3-3. The reaction may follow a sequential dehalogenation path, in which chlorine atoms are removed and replaced with hydrogen atoms, along with the formation of HCl and yielding CH4. In addition to reductive dehalogenation, mineralization of CF to CO2 and HCl have been observed under anoxic conditions [24].
Distinguishing between a sequential, or a direct pathway to methane (Fig. 3-3) is important for assessing the potential formation of toxic intermediates, The kinetic analysis was conducted using the data from sampling ports within the media as there is a possibility of biodegradation on the top portion of the BTF above the media or at the bottom disengagement chamber used for separation of liquid and gas effluents. The CF concentrations in these samples along with the influent stream concentration were used to develop the transformation kinetics as a pseudo first
62 | P a g e
order reaction rate based on a plug flow reactor model. The kinetics reaction rate constants were obtained from the slopes of the regression lines, from the following equation:
푙푛 ( 퐶 ) = 푘푡 ……………………………………………………………… (1) 퐶표
Where t is time, C and Co are the effluent and the influent concentration, respectively, and k reaction rate constant (seconds-1). Fig. 3-4 shows the reaction rate constant for the BTF for the three phases of reactions. The error bars represent the standard deviations of three replicas. The results in the figure show higher reaction rate of 0.002 s-1 was obtained for Phase II as compared to Phase I, whereas the reaction rate did not change significantly between Phase II and Phase III.
The addition of surfactant prompted higher reduction rate constant. This could be due to the co- metabolism and dissolution effects of the surfactants that resulted in excess biomass retention within the BTF bed. As discussed in the performance section, these values correlated with efficiencies of the BTF for the different phases. The higher reaction rate was obtained due to the enhancement and improvement gained from the co metabolite and surfactant added to the system, which increased microbial activity within the biofilter bed. The results correspond well with the increase in CF removal efficiency during Phase II and III.
3.3.3 Carbon mass balance
The cumulative CO2 equivalent of for this research is presented in Fig.3-5. To study the carbon cycle within the bed for both the liquid and gaseous phases, all the carbon sources and products were measured. The influent consisted of gaseous concentrations of CF and ethanol, plus aqueous inorganic and organic carbon. The effluent included aqueous inorganic and organic carbon, a carbon equivalence of volatile suspended solids, gaseous carbon dioxide, gaseous
63 | P a g e
methane and a carbon equivalence of CF and ethanol concentrations. The CO2 equivalence of all the carbon components was calculated in moles and a cumulative input and output CO2 equivalence of carbon was plotted on sequential time (Fig.3-5). The carbon recovery obtained in this research was 63% with a standard deviation of 10% and similar carbon recovery was reported for n-hexane as single VOC with surfactant [54]. To avoid toxicity to the microbes, CF was supplied to the BTF at a relatively lower flow rate, which resulted in the lower recovery.
The loss of influent and effluent carbon was produced as biomass within the BTF [54]. This hypothesis is justified by comparing the loss of carbon to the amount of biomass accumulated within the bed. The cellular composition for typical heterogeneous anaerobic microorganisms is represented as C4.9H9.4NO2.9 [55]. These compositions were used as the basis for relating the ammonia consumed in building up new biomass to estimate the amount of biomass retained within the BTF. A two-tailed t-test was performed to compare the results of the carbon consumed and the biomass produced. The test generated a p- value < 0.05 indicating that the difference between the carbon retained and the biomass produced was statistically significant, therefore, confirming that the loss of carbon within the BTF was utilized for biomass growth.
It is worthwhile to note that the main carbon contributors to the carbon balance of BTF are the gas phase concentrations of the influent and effluent CF and ethanol concentration, and effluent gaseous carbon dioxide and methane. The share of carbon in the liquid phase due to the amount calculated from volatile suspended solids, influent and effluent organic carbon in the aqueous phase could be considered negligible since the total aqueous carbon did not exceed 5 % of the total carbon in the system.
64 | P a g e
3.3.4 Microbial ecological analyses and correlation
As a microbial degradation process, biofiltration is based on the ability of microorganisms to degrade organic and inorganic compounds. Webster et al. (1997) studied the microbial ecosystems in compost and granular activated carbon biofilters and found out both biofilters dominated by gram-negative bacteria [56]. Although, gram negative bacteria has been reported as the dominant bacterial community of some biofilters, few studies have been done on the microbial diversity of BTF [55, 56]. Similarly, Amann et al. (1995) suggested that up to 99
% of the microorganisms that are active in biofilters are not culturable by traditional cultivation methods [57]. Hence, in order to understand and analyze the microbial community structure and the diversity of the BTF during the dechlorination processes, molecular methods have been proven to be powerful tools for analyzing the diversity and structure of the microbiome [58]. In this research, Ion Torrent system was used to examine the impact of the presence and absence of co metabolite and surfactant on the bacterial community structure in the BTF. To get a high diversity of microbes, inocula usually come from digested activated sludge or previously cultivated microflora [59]. Initially, in this research microbes were acclimated from CF based culture and methanogenic bacteria from food waste. Overall, the BTF bacterial communities consisted of mainly betaproteobacteria (45%) following deltaproteobacteria (19%), synergistia
(8%), alphaproteobacteria (7%) and gammabacteria, Fig. 3-6 shows the class level compositions of bacteria based on 97% identity of 16S rRNA gene sequences and it is based on 97% identity of 16S rRNA gene sequences in class level. Table 3-2 shows dominant groups in species level
(97% identity), their taxonomical class, potential metabolic role and related receptors. The community dominant compositions were A. oryzae and A. restrica, Geobacter spp. & G. sulfurreducens, Smithella spp, Azonexus spp, Aminobacterium spp. and Anaerbaculum mobile.
65 | P a g e
Especially, A. oryzae and A. restrica, Geobacter spp. and, Smithella spp. were the main species in all the three phases. The prevalence of these species has also been reported previously from various microbial utilization and studies related to anaerobic biodegradation. A. oryzae [60] was reported for perchlorate remediation and Smithella spp. utilize to degrade propionate first to acetate and then to butyrate [61]. Both have been studied in the biological removal of perchlorate and propionate anaerobically [60, 61]. Similarly, Anderson et al. (2003) reported that Geobacter species can be important agents for in situ bioremediation of uranium [62]. Bae et al. (2007) studied about the species of A. restrica and found out that it’s a nitrogen-fixing bacterium.
Within these dominant species, the bacteria with the highest relative abundances across all the samples were A. orazae, Geobacter spp, G. sulffurrducens and Smithella.
Fig. 3-7 shows that the relative abundance and the diversity of the anaerobic microbial community observed in the BTF for samples collected from port 2 (see Fig. 3-1). During Phase
I, the most dominant species were A. restrica and A. oryzae (46% and 21%) followed by
Geobacter spp. (16%) and Aminivibrio pyruvatiphilus (6%). However, during Phase II, no significant sequences of A. restrica were retrieved, the relative abundance of Geobacter spp. reduced to 13%. On the other hand, the amount of A. pyruvatiphilus decreased to less than 1%, while A. oryzae showed a significant increase to 76% and Azonexus spp. and Smithella spp. were significantly detected in 0.6% and 3%, respectively. During Phase III, A. oryzae, Smithella spp.
Clostridium saccharoperbutylacetonicum and Geobacter spp. were the dominant species with the relative abundance of 50%, 14%, 5% and 5%, respectively. With the addition of surfactant in the BTF system, the growth of A. oryzae was greatly enhanced. Furthermore, the addition of the surfactant has inhibited the growth of CF degrading species like A. restrica which were the dominant species during Phase I. This effect was clearly noticed when the feed stream CF was
66 | P a g e
supplemented with a co metabolite with a surfactant in the BTFs during Phase II, where the concentration of A. oryzae increased significantly from 21% to 76%. Furthermore, the presence of co metabolite during Phase II increased the growth of A. restrica than the other dominant species. In general, the relative abundance of A. oryzae increased with the degradation of CF which correlates to the corresponding removal efficiency and EC. Hence, in Phase II, where ethanol was combined with Tomadol, the average removal efficiency was 64% and at this higher efficiency, the abundance of A. oryzae also increased to 76%. The percentage of A. oryzae for
Phase I and III are 21% and 50%, respectively. It is, therefore, speculated that A. oryzae could be the primary bacteria for the dehalogenation of CF under conditions of this research.
3.4 Conclusion
This research evaluated the effectiveness of BTF for the degradation of CF under anaerobic condition. The BTFs performance for the CF was evaluated for three phases, namely, in the presence of co metabolite (Phase I), in the presence of co metabolite and surfactant (Phase
II) and the presence of surfactant only (Phase III). Enhanced removal efficiency was obtained when surfactant was added to the system together with ethanol (Phase II), which provided a removal efficiency of 64% with EC of 0.17 g/ (m3.hr) for a loading rate of 0.27g./(m3.hr).
Consequently, higher reaction rate constants were attained in this phase as compared to the other two phases. The results of the microbial analysis showed a diverse group of bacterial community in the biofilter. From the biological community structure investigation, the presence of co metabolite supports the growth of A. restrica species than others. A mixture of co metabolite and surfactant benefits the growth of A. oryzae and Geobacter spp. However, the absence of co metabolite greatly encourages Smithella spp more than A. oryzae and Geobacter spp. The
67 | P a g e
increase in the percentage of A. orazae also correlates with the removal efficiency of CF. The overall results obtained for the BTF showed clearly that using both co metabolite and surfactant had effectively enhanced the biodegradation of CF due to providing more favorable conditions for the growth of bacteria colonies.
68 | P a g e
Table 3-1. Operating condition for the BTF degrading chlorform under anaerobic condition. Influent CF concentration was 5 ppmv
Operating Conditions Phases I II III with with Co No
Co metabolite metabolite Co metabolite and with no Surfactant Surfactant Surfactant Operation time, days 44 97 95 Average Chloroform Removal Efficiency 49 ± 9 64±8 59±8 (%) Loading Rate (g/m3. h) 0.27 0.27 0.27 Elimination Capacity (g/m3. h) 0.13 ± 0.02 0.17 ±0.02 0.16 ±0.02
69 | P a g e
Table 3-2: Summary of the community diversity along with their metabolic characteristics
Bacteria Taxonomic Metabolic role Electronic classification receptor (Class) Azospira suillum, Betaproteobacteria dissimilatory reduction of humic substances A. oryzae, perchlorate and chlorate and nitrate and Azonexus metals fungiphilus, Azonexus spp. A. restrica Betaproteobacteria - metabolizes b- D,L. Lactic acid, hydroxybutyrate, Putrescine and c-hydroxybutyrate, D,L- B. bhydroxybutyric lactate and putrescine as acid growth substrates -capacity of nitrogen fixation and reduction acetylene to ethylene Geobacter spp. Deltaproteobacteria Oxidize organic iron oxide or compounds and metals, other available including iron, radioactive metals metals and petroleum compounds Smithella spp. Deltaproteobacteria anaerobic, syntrophic, fumarate propionate-oxidizing bacteria Anaerobaculum Synergistia anaerobic, peptide- Crotonate, mobile fermenting Amino acid Aminobacterium Synergistia an amino acid-degrading Amino acid anaerobe isolated from anaerobic sludge
70 | P a g e
Figure 3-1 Schematic diagram of the biotrickling filter (BTF)
71 | P a g e
Figure 3-2 Performance of the BTF in the three phases. Phase I (44 days) presence of co metabolite, phase II (97 days) presence of co metabolite and surfactant, phase III (95 days) presence of surfactant with no co metabolite
72 | P a g e
Cl C** CO CO2
Cl Hydrolytic Reduction k1 HCOOH
k 2 k 4 k5 - 2e HCl - - + HCl H 2e 2e HCl + + H H C H OH Reductive Hydrogenolysis 2 5 k3
Halorespiration
CO2 CH4 + HCl
Figure 3-3 Sequential and direct reaction pathways for reductive dehalogenation of CF
73 | P a g e
Figure 3-4 Reaction rate constants for chloroform in three phases. Three data sets have been collected per phase and the error bars present the standard deviations for these replicas. Phase I:
Chloroform with Co metabolite (Ethanol), Phase II: Chloroform with Co metabolite (Ethanol) and nonionic surfactant and Phase III: Chloroform with nonionic surfactant.
74 | P a g e
Figure 3-5 Carbon mass balance: Cumulative carbon input and output as CO2 equivalent in mole for the BTF
75 | P a g e
Figure 3-6 Class level compositions of bacteria based on 97% identity of 16S rRNA gene sequences
76 | P a g e
Figure 3-7 Bacterial community diversity for the three conditions for samples collected at the top port of the biofilter. Phase I: Chloroform with Co metabolite (Ethanol), Phase II: Chloroform with Co metabolite (Ethanol) and nonionic surfactant and Phase III: Chloroform with nonionic surfactant
77 | P a g e
3.5 References 1. Li, J. and E.R. Blatchley, Volatile disinfection byproduct formation resulting from chlorination of organic− nitrogen precursors in swimming pools. Environ. Sci. Technol., 2007. 41(19): p. 6732-6739. 2. Durmishi, B.H., et al., Disinfection of Drinking Water and Trihalomethanes: A Review. International J. Advan. Res. Chem. Sci. (IJARCS), 2015. 2(11): p. 45-56. 3. Richardson, S.D., et al., What's in the pool? A comprehensive identification of disinfection by-products and assessment of mutagenicity of chlorinated and brominated swimming pool water. Environ. Health Perspec., 2010. 118(11): p. 1523. 4. ATSDA, A.f.T.S.a.D.R. Toxic Substances Portal - Chloroform: (Accessed 09/10/2016). 2015 [cited 2016 September, 05]; Available from: http://www.atsdr.cdc.gov/toxprofiles/tp.asp?id=53&tid=16. 5. USEPA. Ground Water and Drinking Water: Table of Regulated Drinking Water Contaminants,(Accessed 09/10/2016) : https://www.epa.gov/ground-water-and-drinking- water/table-regulated-drinking-water-contaminants. 2016 [cited 2016 09/10/2016]; Available from: https://www.epa.gov/ground-water-and-drinking-water/table-regulated- drinking-water-contaminants. 6. Shemer, H. and N. Narkis, Trihalomethanes aqueous solutions sono-oxidation. Water Res., 2005. 39(12): p. 2704-2710. 7. Clark, R.M., J.Q. Adams, and B.W. Lykins Jr, DBP control in drinking water: cost and performance. J. Environ. Eng., 1994. 120(4): p. 759-782. 8. Houel, N., R.E. Selleck, and F.H. Pearson, Air stripping of chloroform from water. J. Environ. Eng. Divis., 1979. 105(4): p. 777-781. 9. McCulloch, A., Chloroform in the environment: occurrence, sources, sinks and effects. Chemosphere, 2003. 50(10): p. 1291-1308. 10. Yoon, I.-K. and C.-H. Park, Effects of gas flow rate, inlet concentration and temperature on biofiltration of volatile organic compounds in a peat-packed biofilter. J. Biosci. Bioeng., 2002. 93(2): p. 165-169. 11. Egli, C., et al., Anaerobic dechlorination of tetrachloromethane and 1, 2-dichloroethane to degradable products by pure cultures of Desulfobacterium sp. and Methanobacterium sp. FEMS Microbiol. Letters, 1987. 43(3): p. 257-261. 12. Yu, Z. and G.B. Smith, Chloroform dechlorination by a wastewater methanogenic consortium and cell extracts of Methanosarcina barkeri. Water Res., 1997. 31(8): p. 1879-1886. 13. Egli, C., et al., Transformation of tetrachloromethane to dichloromethane and carbon dioxide by Acetobacterium woodii. Appl. Environ. Microbiol., 1988. 54(11): p. 2819- 2824. 14. Egli, C., et al., Transformation of tetra-and trichloromethane to CO2 by anaerobic bacteria is a non-enzymic process. FEMS Microbi Letters, 1990. 68(1-2): p. 207-212. 15. Picardal, F.W., et al., Involvement of cytochromes in the anaerobic biotransformation of tetrachloromethane by Shewanella putrefaciens 200. Appl. Environ. Microbiol., 1993. 59(11): p. 3763-3770. 16. Novak, P., L. Daniels, and G. Parkin, Enhanced dechlorination of carbon tetrachloride and chloroform in the presence of elemental iron and Methanosarcina barkeri, Methanosarcina thermophila, or Methanosaeta concillii. Environ. Sci. Technol., 1998. 32(10): p. 1438-1443.
78 | P a g e
17. Mikesell, M.D. and S.A. Boyd, Dechlorination of chloroform by Methanosarcina strains. Appl. and Environ. Microbiol., 1990. 56(4): p. 1198-1201. 18. Becker, J.G. and D.L. Freedman, Use of cyanocobalamin to enhance anaerobic biodegradation of chloroform. Environ. Sci. Technol., 1994. 28(11): p. 1942-1949. 19. Yoon, I.-K., C.-N. Kim, and C.-H. Park, Optimum operating conditions for the removal of volatile organic compounds in a compost-packed biofilter. Korean J. Chem. Eng., 2002. 19(6): p. 954-959. 20. Balasubramanian, P., L. Philip, and S.M. Bhallamudi, Biotrickling filtration of complex pharmaceutical VOC emissions along with chloroform. Biores. Technol., 2012. 114: p. 149-159. 21. Mahon, S., Biofiltration for the Treatment of Recalcitrant Volatile Organic Compounds in Polluted Air Streams. Air & Water Pollut., 2013. 22. Leson, G. and A.M. Winer, Biofiltration: an innovative air pollution control technology for VOC emissions. J. Air & Waste Manage. Assoc., 1991. 41(8): p. 1045-1054. 23. Kim, Y., L. Semprini, and D.J. Arp, Aerobic cometabolism of chloroform and 1, 1, 1- trichloroethane by butane-grown microorganisms. Bioremediation J., 1997. 1(2): p. 135- 148. 24. Cappelletti, M., et al., Microbial degradation of chloroform. Appl. Microbiol. Biotechnol., 2012. 96(6): p. 1395-1409. 25. Edwards, D.A., R.G. Luthy, and Z. Liu, Solubilization of polycyclic aromatic hydrocarbons in micellar nonionic surfactant solutions. Environ. Sci. Technol., 1991. 25(1): p. 127-133. 26. Yeh, D.H., K.D. Pennell, and S.G. Pavlostathis, Effect of Tween surfactants on methanogenesis and microbial reductive dechlorination of hexachlorobenzene. Environ. Toxicol. Chem., 1999. 18(7): p. 1408-1416. 27. Laha, S. and R.G. Luthy, Effects of nonionic surfactants on the solubilization and mineralization of phenanthrene in soil–water systems. Biotechnol. Bioeng., 1992. 40(11): p. 1367-1380. 28. Aronstein, B.N. and M. Alexander, Surfactants at low concentrations stimulate biodegradation of sorbed hydrocarbons in samples of aquifer sands and soil slurries. Environ. Toxicol. Chem., 1992. 11(9): p. 1227-1233. 29. Hassan, A.A. and G.A. Sorial, n-Hexane biodegradation in trickle bed air biofilters. Water, Air, & Soil Pollut., 2008. 8(3-4): p. 287-296. 30. Tiehm, A., Degradation of polycyclic aromatic hydrocarbons in the presence of synthetic surfactants. Appl.and Environ. Microbiol., 1994. 60(1): p. 258-263. 31. Volkering, F., et al., Influence of nonionic surfactants on bioavailability and biodegradation of polycyclic aromatic hydrocarbons. Appl. Environ. Microbiol., 1995. 61(5): p. 1699-1705. 32. Boonchan, S., M.L. Britz, and G.A. Stanley, Surfactant‐ enhanced biodegradation of high molecular weight polycyclic aromatic hydrocarbons by Stenotrophomonas maltophilia. Biotechnol. Bioeng., 1998. 59(4): p. 482-494. 33. Hassan, A.A. and G. Sorial, Biological treatment of benzene in a controlled trickle bed air biofilter. Chemosphere, 2009. 75(10): p. 1315-1321. 34. Woertz, J., K. Kinney, and P. Szaniszlo, A fungal vapor-phase bioreactor for the removal of nitric oxide from waste gas streams. J. Air & Waste Manage. Assoc., 2001. 51(6): p. 895-902.
79 | P a g e
35. Volkering, A. Breure, and W. Rulkens, Microbiological aspects of surfactant use for biological soil remediation. Biodegradation, 1997. 8(6): p. 401-417. 36. Yuan, S., et al., Effect of nonionic and cationic surfactants on the dechlorination kinetics and products distribution of various polychlorinated benzenes by Cu/Fe particles. Sep. Purif. Technol., 2010. 74(1): p. 130-137. 37. Atikovic, E., M.T. Suidan, and S.W. Maloney, Anaerobic treatment of army ammunition production wastewater containing perchlorate and RDX. Chemosphere, 2008. 72(11): p. 1643-1648. 38. Wu, S., et al., Anaerobic biodegradation of soybean biodiesel and diesel blends under methanogenic conditions. Water research, 2015. 87: p. 395-402. 39. APHA, Standard methods for the examination of water and wastewater. American Public Health Assoc. and Water Environment Federation, Washington, DC, USA, 2005. 40. Zehraoui, A., et al., Impact of alternate use of methanol on n-hexane biofiltration and microbial community structure diversity. Biochem. Eng. J., 2014. 85: p. 110-118. 41. Zhai, J., et al., Microbial Community in a Biofilter for Removal of Low Load Nitrobenzene Waste Gas. PloS one, 2017. 12(1): p. e0170417. 42. Dowd, S.E., et al., Bacterial tag-encoded FLX amplicon pyrosequencing (bTEFAP) for microbiome studies: bacterial diversity in the ileum of newly weaned Salmonella-infected pigs. Foodborne Pathog. and Dis., 2008. 5(4): p. 459-472. 43. Edgar, R.C., Search and clustering orders of magnitude faster than BLAST. Bioinform., 2010. 26(19): p. 2460-2461. 44. Capone, K.A., et al., Diversity of the human skin microbiome early in life. J. Investigative Dermatolo., 2011. 131(10): p. 2026-2032. 45. Eren, A.M., et al., Exploring the diversity of Gardnerella vaginalis in the genitourinary tract microbiota of monogamous couples through subtle nucleotide variation. PloS one, 2011. 6(10): p. e26732. 46. Swanson, K.S., et al., Phylogenetic and gene-centric metagenomics of the canine intestinal microbiome reveals similarities with humans and mice. The ISME J., 2011. 5(4): p. 639-649. 47. DeSantis, T.Z., et al., Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl. and Environ. Microbiol., 2006. 72(7): p. 5069- 5072. 48. Zitomer, D.H. and R.E. Speece, Methanethiol in nonacclimated sewage sludge after addition of chloroform and other toxicants. Environ. Sci. Technol., 1995. 29(3): p. 762- 768. 49. Bagley, D.M. and J.M. Gossett, Chloroform degradation in methanogenic methanol enrichment cultures and by Methanosarcina barkeri 227. Appl. Environ. Microbiol., 1995. 61(9): p. 3195-3201. 50. Bouwer, E.J., B.E. Rittmann, and P.L. McCarty, Anaerobic degradation of halogenated 1-and 2-carbon organic compounds. Environ. Sci. Technol., 1981. 15(5): p. 596-599. 51. Krone, U.E., et al., Coenzyme F430 as a possible catalyst for the reductive dehalogenation of chlorinated C1 hydrocarbons in methanogenic bacteria. Biochem., 1989. 28(26): p. 10061-10065. 52. Gupta, M., et al., Biotransformation rates of chloroform under anaerobic conditions—I. Methanogenesis. Water Res., 1996. 30(6): p. 1377-1385.
80 | P a g e
53. Ying, G.-G., Fate, behavior and effects of surfactants and their degradation products in the environment. Environ. International, 2006. 32(3): p. 417-431. 54. Hassan, A.A. and G.A. Sorial, A comparative study for destruction of n-hexane in trickle bed air biofilters. Chem. Eng. J., 2010. 162(1): p. 227-233. 55. Bruce, E.R. and L.M. Perry, Environmental biotechnology: principles and applications. New York: McGrawHill. Vol. 400. 2001: McGraw-Hill 56. Webster, T.S., et al., Microbial ecosystems in compost and granular activated carbon biofilters. Biotechnol. Bioeng., 1997. 53(3): p. 296-303. 57. Amann, R.I., W. Ludwig, and K.-H. Schleifer, Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Microbiol. reviews, 1995. 59(1): p. 143-169. 58. Fakruddin, M. and K. Mannan, Methods for analyzing diversity of microbial communities in natural environments. Ceylon J. Sci. (Biol. Sci.), 2013. 42(1). 59. Wagner, M., et al., Microbial community composition and function in wastewater treatment plants. Antonie Van Leeuwenhoek, 2002. 81(1-4): p. 665-680. 60. Hutchison, J.M., et al., Perchlorate reduction using free and encapsulated Azospira oryzae enzymes. Environ. Sci. Technol., 2013. 47(17): p. 9934-9941. 61. Dolfing, J., Syntrophic propionate oxidation via butyrate: a novel window of opportunity under methanogenic conditions. Appl. and Environ. Microbiol., 2013. 79(14): p. 4515- 4516. 62. Anderson, R.T., et al., Stimulating the in situ activity of Geobacter species to remove uranium from the groundwater of a uranium-contaminated aquifer. App. Environ. Microbiol., 2003. 69(10): p. 5884-5891.
81 | P a g e
4. Comparative Study on the Performance of Anaerobic and Aerobic Biotrickling Filter for the Removal of Chloroform
4.1 Abstract
The use of Biotrickling filter (BTF) for gas phase treatment of volatile trihalomethanes
(THMs) stripped from water treatment plants could be an attractive treatment option. The aim of this study is to use lab scale anaerobic BTF to treat gaseous CF (recalcitrant to biological transformation) as a model THMs and compare the results with aerobic BTF. Additional investigations were conducted to determine the microbial diversity present with in the BTFs. CF is a hydrophobic volatile THM known to be difficult to bio degrade. To improve the degradation process, ethanol was used as a co metabolite at a different ratio to CF. The experimental plan was designed to operate one BTF under anaerobic condition and the other one under aerobic acidic condition. Higher elimination capacity of 0.23 ± 0.01 g/m3.hr was observed with a removal efficiency of 80.9 ± 4% for the aerobic BTF operating at pH 4 for the concentration ratio of 1:40 CF to ethanol. For similar ratio, the anaerobic BTF supported lower removal efficiency of 59 ± 10% with corresponding lower elimination capacity of 0.16 ± 0.01 g/m3.hr.
The carbon recovery acquired for the anaerobic and aerobic BTFs were 59% and 63%, respectively. The loading rate for CF on both BTFs was 0.27 g/m3.hr (per m3 of filter bed volume). The variations of the microbial community were attributed to the degradation of CF in each BTF. Azospira oryzae and Azospira restrica were the dominant bacteria and potential candidate for CF degradation for the anaerobic BTF. Whereas, Fusarium sp. and Fusarium solani were the dominant fungi and potential candidates for CF degradation in the aerobic BTF.
This chapter is based on a publication: Bineyam Mezgebe, Keerthisaranya Palanisamy, George A. Sorial , E. Sahle-Demessie, Ashraf Aly Hassan, and Jingrang Lu. 2017. Comparative Study on the Performance of Anaerobic and Aerobic Biotrickling Filter for the Removal of Chloroform." Environmental Engineering Science: EES-2017-0275-ver9-Mezgebe. 82 | P a g e
4.2 Introduction
Drinking water disinfection by chlorination is the most important step in water treatment to kill pathogens and reduce waterborne diseases. However, chlorine also reacts with the natural organic matter (NOM) that are present in most surface water, and produces many harmful disinfection byproducts (DBP). Most DBPs are known to be toxic and pose a risk to human health [1]. Many DBPs are also bio accumulative, and thus long-term exposure to low DBPs causes a chronic health risk. The common DBPs from chlorination of water include trihalomethanes (THMs), and haloacetic acids (HAAs) [2, 3]. The main THMs include CF, dichlorobromomethane (DCBM), dibromochloromethane (DBCM) and bromoform [4]. Various factors affecting the formation of DBP include the water pH and temperature, the concentration and contact time of chlorine and bromine, and the concentration of natural organic matters [5].
The methods currently used to reduce NOMs and to minimize the formation of DBPs include the use of activated carbon filters and conventional water treatment processes including clarification, coagulation, flocculation, sedimentation, and filtration [6].
However, these controlling methods can only remove about 30% of the precursors for
THMs [7]. In addition, removing these THMs by physical and/or chemical methods at low concentrations found in drinking water is expensive and may generate a secondary pollutant. The high Henry's law constant of many of the THMs allows alternative approaches for treatment such as gas stripping combined with biological treatment [8]. Thus, the formation of THMs in drinking water has highlighted the need for exploring alternative disinfectants for chlorine and new treatment technologies for removing THMs after they are formed.
In this study, CF was taken as a model DBPs since it is the most toxic and most abundant of the THMs. CF is a volatile THMs and could be removed from contaminated waters to the
83 | P a g e
gaseous phase by air stripping [4, 9, 10]. Biological treatment techniques for volatile organic compounds (VOC) removal have several advantages. Compared to the conventional methods, such as incineration, catalytic oxidation, and adsorption, biological treatments could be cost effective as safer and eco-friendly [11]. Most of the research on the biological treatment of CF has been limited to batch liquid phase processes at wastewater treatment plants or hazardous waste disposal sites. Under anaerobic conditions, CF could undergo a reductive biotransformation by pure cultures of methanogens [12, 13], acetogenic bacteria [14], sulfate- reducing bacteria [15] and iron-reducing bacteria [15, 16] producing partial dehalogenation and mineralization [13-16]. Thus, biological techniques have resulted in dechlorination of CF to dichloromethane, methane, and carbon dioxide [15, 17, 18].
CF is a tri-chlorinated methane compound and is recalcitrant to biological transformation.
It can only be transformed or biodegrade in the presence of a co metabolite under anaerobic or aerobic environments [19-21]. Furthermore, the halogenic nature of CF can affect the biodegradation process [22]. To overcome this obstacle, halogenated organic compounds often require the presence of an easily degradable substrate that can increase their biodegradability by co-metabolism [22]. Anaerobic dechlorination of CF has been observed by different researchers by using methanogenic microbes with electron donating co metabolites in reductive CF biotransformation [17, 23-25]. Additionally, CF removal ranging between 13% and 43% was obtained in a study of co-metabolism of CF and other THMs [26].
Although most studies show successful biodegradation of CF in the liquid phase, there is a limited amount of reported work on the use of biofiltration for the removal of CF from gaseous streams. Biofiltration is one of the proven technologies for removing VOCs from high volume stream as it is environmentally friendly, cost effective and releases fewer byproducts [27]. The
84 | P a g e
use of aerobic biofiltration technique has been reported for the biotreatment of CF with other mixtures of different VOCs [27, 28]. Yoon et al. have shown the degradation potential of nine
VOCs including CF and found the highest removal was for toluene (99%) and the lowest removal was for CF (89.4%) [27]. Similarly, Balasubramanian et al. evaluated the biodegradation of CF along with a mixture of VOCs commonly found in pharmaceutical emissions, using a biotrickling filter. Their study showed that increasing the rate of CF loading significantly reduced the degradation efficiency of the reactor for the mixture of VOCs [28].
Similarly, in our previous work, an aerobic BTF was used to treat gaseous CF in the presence of ethanol as co metabolite [29]. However, to the best of our knowledge, no reported work in literature is available for the use of anaerobic BTF in treating CF.
The main goal of current study is to examine gas phase CF removal by using anaerobic
BTF in the presence of ethanol as a co metabolite. In addition, comparison was conducted on the performances of this current anaerobic and previously studied aerobic BTF. The study also investigated the microbial ecology within both BTFs in order to get a deep insight of the factors affecting BTFs.
4.3 Materials and methods
4.3.1 Materials
Chloroform with 99.8% purity was obtained from Fisher Scientific (Pittsburgh, PA,
USA) and Ethanol with 99.5% purity obtained from Sigma Aldrich (St. Louis, MO, USA). CF is
−3 3 o highly hydrophobic with a Henry’s law constant, KH of 3.5×10 atm.m / mol at 25 C, and the
−6 3 KH value of the hydrophilic ethanol is 5.1×10 atm.m / mol at 25 °C [30, 31]. The measuring
85 | P a g e
sensors for pH, dissolved oxygen (DO), and ammonia were acquired from Accumate
Instruments. Genomic DNA extractions of bacterial and fungi strains were performed using the
Mo Bio PowerSoil DNA (M Bio Lab, Inc., Carlsbad, CA) Kit, which was done by Molecular
Research LP (MR DNA, Shallowater, TX).
4.3.2 Biotrickling filter (BTF)
In this current work, an anaerobic BTF is evaluated for degrading CF. The results were used to compare the performance to a previously studied aerobic BTF. The loading rate of CF for both BTFs was kept at 0.27 g/m3.hr (i.e., per m3 of filter volume) throughout the experiment.
Ethanol (hydrophilic VOC) was introduced as gaseous co metabolite at different loading rates for both BTFs. Table 4-1 shows all the operational parameters for the anaerobic BTF. Fig.4-1 also shows the schematic diagram of each BTF. Each BTF column consists of seven cylindrical glass sections with an internal diameter of 7.6 cm and a total length of 130 cm and is packed with pelletized diatomaceous earth biological support media to a depth of about 60 cm (Celite® 6 mm
R-635 Bio-Catalyst Carrier; Celite Corp., Lompoc, CA). Both BTFs operated in a co-current mode with both gas and liquid flow downwards to acclimatize and enhance the growth of biomass. In this anaerobic BTF system, nitrogen was used as a carrier gas with a flowrate of 0.5
L/ min which provides a corresponding empty bed residence time (EBRT) of 5.44 min. the initial
CF concentration was 5 ppmv.
Methanogenic microorganisms were used to inoculate the filter bed. Initially, these bacteria were obtained from a nutrient enriched solution kept under a blanket of nitrogen gas that was acclimated in our lab to CF in a 4 liter amber batch reactor for two months. The CF feed
86 | P a g e
was step-wise increased from 5 – 50 ppmv within the two-month period. This inoculum was mixed in the ratio of 1:1 with another methanogenic bacteria acquired from another bioreactor that was treating food waste prior to seeding the BTF. The origin of these methanogenic bacteria was from an anaerobic digester at a local wastewater treatment plant.
The buffered nutrient solution containing ammonia as electron donor was supplied at an average rate of 2.0 L/ day. The growth media for anaerobic BTF was prepared with medium concentrations of 996 mg/L NH4Cl, 414 mg/L KH2PO4, 390 mg/L MgCl2.6H2O, 280 mg/L
CaCl2.2H2O, 2 mg/L FeCl2.4H2O, 4.79 mg/L CuSO4.5H2O, 6.53 mg/L MnSO4.H2O, 5.24 mg/ L
ZnCl2, 4.58 mg/L CoCl2.6H2O, 0.32 mg/L B(OH)3, 4.79 mg/L NiCl2.6H2O, 0.12 mg/L 4- aminobenzoic acid (99%), 0.048 mg/L biotin, 0.0024 mg/L cyanocobalamin, 0.05 mg/L, folic acid dihydrate (99%), 0.12 mg/L nicotinic acid (98%), 0.12 mg/L pantothenic acid Ca-salt hydrate (98%), 0.24 mg/L pyridoxine hydrochloride (98%), 0.12 mg/L riboflavin (98%), 0.12 mg/L thiamine hydrochloride (99%), and 0.12 mg/L thioctic acid (98%). The composition of the nutrient solution was used according to the ones provided in literature [19, 32, 33]. One Molar
o NaHCO3 was used as a buffer to maintain the pH at 7. The temperature was kept at 35 C in a temperature-controlled room to maintain favorable methanogens growth. Whereas, in the aerobic system, air was used as a carrier gas with a flowrate of 0.5 L/ min at corresponding
EBRT of 5.44 min.
In this case, the buffered nutrient solution containing nitrate was supplied at an average rate of 2.0 L/ day. The nutrients were supplied at an acidic pH of 4 by the addition of sodium formate buffer to encourage the growth of fungi colonies. The buffered solution contains all necessary macronutrients, micronutrients, and buffers, is described by Sorial et al. [34]. The temperature of the aerobic BTF was maintained at 35 °C similar to the anaerobic BTF. Liquid
87 | P a g e
CF and ethanol were injected via separate syringe pumps in series and vaporized into the nitrogen or air stream.
4.3.3. Strategies of biomass control
The aerobic BTF operation was tested for different biomass control technologies namely stagnation and backwashing. Stagnation non-use period was observed during two consecutive days per week. During stagnation period, the BTF did not get any nutrients, VOCs, or air.
Whereas backwashing involves flushing the media bed with 18 L of buffered nutrient solution, inducing medium fluidization at approximately 50% bed expansion when the system is offline.
Following this, the recirculating nutrient solution will be stopped, the biofilter is drained, and then another 18 L of the nutrients will be supplied for a final rinse. More details on biomass control technologies can be found in Hassan and Sorial [35]. However, for the case of anaerobic system there was no need to use any kind of biomass controlling technique since there was no related biomass growth problem.
4.3.4 Sampling and analysis
Gas and liquid samples were collected daily from the BTF systems five days per week for the measurement of composition of feed and effluent gas/ liquid streams. Liquid samples were collected for the measurement of the influent and effluent liquid pH, ammonia, and organic matter. The gas flow pressure-drop across the bed, and operating temperature were taken on daily basis. Dissolved oxygen for the anaerobic BTF was taken every day in order to check for any leak by using Accumate DO probe. Gas phase samples for anaerobic BTF were taken on-line from different points along the BTF column using an electrically controlled low-bleed eight-port
88 | P a g e
Valco valve and analyzed by gas chromatograph. The samples were analyzed for CF, ethanol, and methane (CH4) as a by-product. They were injected into GC – HP, Column: HP, 608, 30 m x
530 μm film thickness, injection splitless through 5ml sample loop equipped with a flame ionization detector (FID). The GC oven was programmed isothermal at 60 °C (2min) ramped to
90 °C at a rate of 10 oC/ min. The carrier gas (He) flow rate was set at 3.5 mL/ min at constant flow rate. The FID was used with N2 make-up gas at a flow rate of 30 mL/ min, a fuel gas flow
(H2) of 40 mL/ min, and airflow of 400 mL/ min. Retention time for CF was 3.8 min under the above conditions used. For determining levels of reaction products, such as CO2, samples were also taken automatically by GC HP- TCD from each sampling port in the BTF. The GC oven was programmed isothermal at 60°C (1min), ramped to 115°C at 25°C /min. The carrier gas (He) flow rate was set at 3.5 mL/ min, the TCD was used with helium make-up gas at a flow rate of 5 ml /min.
Liquid samples were collected from the effluent stream of BTF once a week. The samples were filtered through a 0.45 µm membrane filter (Whatman Co.) and analyzed for influent and effluent concentrations of ammonia, nitrate, dissolved total carbon, dissolved inorganic carbon, and volatile suspended solids. The concentration of ammonia and nitrate were determined using an ammonia and nitrate electrode sensors. Dissolved total carbon (TC) and dissolved inorganic carbon (IC) content of the liquid samples were determined with a Shimadzu total organic carbon analyzer model TOC - L (Shimadzu Corp., Tokyo, Japan). The volatile suspended solids analysis was conducted according to Standard Method 2540G [36].
It should be noted that before samples are analyzed in the GC/FID, GC/TCD, electrode instruments, or TC/ IC, the instruments are checked for meeting an instrument stability calibration criterion. This criterion is determined by using six concentration levels for target
89 | P a g e
analytes. The response factor (RF) for each standard concentration level is then determined. The instrument stability for initial calibration is acceptable when the RF for each concentration level of the standard solutions is below 10% from the overall mean value for the six standard solutions.
4.3.5 Microbial community molecular analysis
Biofilm samples were collected from anaerobic and aerobic BTF within the media as shown in Fig.4-1. The samples were taken from port 2 (first port from the top within the media) at the end of each phase before proceeding to the next phase. In order to get microbial analysis result, samples from biofilter were collected at the end of each experimental phase [37, 38].
The samples consisted of about five media pellets covered with biomass suspended in liquid.
All the samples collected were stored in a -20 °C freezer prior sending them to molecular research laboratory (Molecular Research LP Shallowater, TX). In this microbial analysis study, bacteria and fungi were chosen for anaerobic and aerobic BTF, respectively. The main reason for bacteria used in the anaerobic BTF is that fungi couldn’t grow under anaerobic environment at neutral condition. Some researchers confirmed the strong correlation of bacterial community growth with pH, while decrease in pH favorably increased fungal growth [37, 39]. The DNA of microbial mass in the samples was extracted using Mo Bio PowerSoil DNA (M Bio Lab, Inc.,
Carlsbad, CA) following manufacturer’s instruction that includes cell breakage steps followed by the addition of detergents and high salt buffers and enzymatic digestion with lysozyme and proteases.
90 | P a g e
For ion torrent sequencing, the 16S rRNA gene V4 variable region PCR primers 515/806 were used in a single-step 30 cycle PCR using the HotStarTaq Plus Master Mix Kit (Qiagen,
USA) under the following conditions: 94 °C for 3 minutes, followed by 28 cycles (5 cycle used on PCR products) of 94 °C for 30 seconds, 53°C for 40 seconds and 72 °C for 1 minute, after which a final elongation step at 72 °C for 5 minutes was performed. Sequencing was carried out at Molecular Research LP (www.mrdnalab.com, Shallowater, TX, USA) on an Ion Torrent
Personal Genome machine (PGM) following the manufacturer’s guidelines. Sequence data were processed using a proprietary analysis pipeline. Sequences were first depleted of barcodes and primers, and those under 150bp or with ambiguous base calls or with homopolymer runs exceeding 6bp were removed. Operational taxonomic units (OTUs), which were defined by clustering at 3% divergence (97% similarity) [40-44], were generated after denoising sequences and removing chimeras. The last OTUs were taxonomically classified using BLASTn against a database derived from RDPII (http://rdp.cme.msu.edu) and NCBI (www.ncbi.nlm.nih.gov) [45].
4.4 Experimental results
4.4.1 Anaerobic biotrickling filter performance
In this study, the effects of co metabolite at different loading rates on the performance of anaerobic BTF, was evaluated. The co metabolite was allowed to mix with CF in the mixing chamber to achieve higher removal efficiency by providing additional electron donor to the micro-organisms. Ethanol was used as a co metabolite since it readily mixes with CF and water.
It is worth noting that the removal efficiency of ethanol was always above 98% for the given loading rate conditions studied for both BTFs. Therefore, the emphasis is placed on the performance of the BTF for CF degradation. The details of operation for anaerobic BTF is given
91 | P a g e
in Table 4-1 where at every phase of operation the corresponding influent concentration, loading rate and days of operation are provided. The table also summarizes the results of the BTF including average removal efficiency and its standard deviation and the elimination capacities of each phase of operation. Fig. 4-2 presents examples of a statistical summary of the removal efficiency as a box plot at different loading rates. The lower boundary of the box denotes the lower quartile, a line within the box marks the median, and the boundary of the box furthest from zero indicates the upper quartile. Whiskers (error bars) above and below the box indicate the 90th and 10th percentiles.
In Phase I, the BTF started up with CF influent concentration of 5 ppmv and ethanol concentration of 25 ppmv providing a corresponding CF loading rate of 0.27 g/m3.hr. The BTF was run for 44 days under the conditions of Phase I, and the average removal efficiency for this phase was 49 ± 9%, which provided an average elimination capacity (EC) of 0.13 ± 0.02 g/m3.hr
(Table 4-1). On day 45, the influent concentration of ethanol was further increased to 50 ppmv with a corresponding ethanol – CF ratio of 1:10. In phase II, the removal efficiency slightly increased to 52 ± 7% with an EC of 0.14 ± 0.01 g/m3.hr. After the system left to run for 33 days
(during phase II), the ethanol concentration was increased to 100 ppmv in phase III. At this level, the system ran for 41 days and the removal efficiency with a corresponding EC was 56 ± 7% and
0.15 ± 0.02 g/m3.hr respectively. On day 118, the ratio of CF to ethanol was further increased to
1:40. During phase (IV), the removal efficiency was at 59 ± 10% which provided a higher elimination capacity of 0.16 ± 0.01 g/m3.hr as compared to the other previous phases.
92 | P a g e
4.4.2 Aerobic biotrickling filter performance
The result for aerobic BTF was reported in our previous study (Palanisamy et al., 2016).
The details of operation for aerobic BTF is given in Table 4-1 where at every phase of operation the corresponding influent concentration, loading rate and days of operation are provided. The table also summarizes the results of the BTF including average removal efficiency with its standard deviation and the elimination capacity. During Phase I, the removal efficiency of CF was 69.9±9% with a corresponding EC of 0.21 ± 0.01 g/m3.hr. On phase II, the removal efficiency of CF was 71.6 ± 5% with EC of 0.22 ± 0.01 g/m3.hr. On Phase III, the removal efficiency of CF increased to 75.1 ± 9% providing EC of 0.22 ± 0.01 g/m3.hr. Finally, on phase
IV, the removal efficiency of CF increased to 80.9% with a standard deviation of 4%. The corresponding EC for this phase was 0.23 ± 0.01 g/m3.hr.
4.5 Discussion of the results
4.5.1. Performance comparison for anaerobic and aerobic BTFs
The use of co metabolite improved CF degradation for both BTFs. It has been observed that for both BTFs the performance increased with an increase in the co metabolite concentration. Few studies have been conducted for the use of co metabolite for CF degradation.
The study conducted by Gupta et al. investigated the use of acetic acid as a co metabolite in anaerobic CF biotransformation in the liquid phase which resulted in higher removal efficiency
[32]. Similarly, aerobic CF biodegradation has been observed during the oxidation of other co metabolites. CF co-oxidation with formate or methane, with butane oxidizing, and nitrifying bacterium has been reported [20]. In our current study, CF displayed significant biodegradation rates when using ethanol as a co-substrate at the ratio of 1:40 (phase IV). Similar conclusion was
93 | P a g e
reported on our previous study in a fungal-based system. In the work, fungi utilization greatly enhanced the performance of the aerobic BTF as compared to the anaerobic one. The highest removal efficiency reported under acidic aerobic condition was significantly reached to 80.9 ±
4% (see Table 4-1). Interestingly, highest elimination capacity was obtained during phase IV of the aerobic BTF (Table 4-1). It is postulated that the use of fungi in the aerobic system helped in enhancing the elimination capacity of CF.
This enhanced performance could be due to the resilience of fungi to acid and dry conditions as compared to bacteria, which is a helpful property when operating biofilters.
Moreover, it is hypothesized that the aerial mycelia of fungi, which are in direct contact with the gas, can take up hydrophobic compounds faster than flat aqueous bacterial biofilm surfaces.
Although, the aerobic condition showed enhanced performance for degradation of CF, the significance of the anaerobic degradation is the renewable energy source. The anaerobic process produces methane rich biogas suitable for energy production helping to replace fossil fuels. The ratio of methane to carbon dioxide ranged from 1.77 to 2.05 (table 4-1) for this system. These values also correlated with the corresponding removal efficiency values. As the removal efficiency increased, the ratio also increased.
4.5.2 Kinetics of CF removal in the BTFs
The removal performances as a function of depth within each BTF were measured weekly. For aerobic BTF, it was conducted one day following stagnation at the sampling ports located along the depth. At the same time, similar measurement was taken for the anaerobic
BTF. The samples were taken along the BTFs from ports that are located at 7.6 cm, 23 cm, 38
94 | P a g e
cm, 53 cm and 60 cm down from the top of the packed bed. The kinetic analysis was conducted using the data from sampling ports within the media as there is a possibility of biodegradation on the top portion of the BTF above the media or at the bottom disengagement chamber used for separation of liquid and gas effluents. The BTF is assumed to function as a plug flow reactor, and the removal kinetics was based on the pseudo first-order reaction as a function of the depth of each BTF. At least three sampling data sets from each port was taken for every phase. The sampling data for every phase were fitted to a linear model with the independent variable, time
(sec), and the dependent variable, loge(C/C0), where C is the effluent concentration and C0 is the influent concentration. The kinetics reaction rate constants were obtained from the slopes of the regression lines. Fig. 4-3 provides the results where the error bars represent the standard deviation from at least three data sets. Fig. 4-3 clearly shows the advantage of fungi utilization in the BTF which is indicated by a higher reaction rate constant as compared to the anaerobic
BTF at the same influent concentration.
CF reaction rate constant increased as the influent co metabolite loading increased. The reaction rate constant values for the four phases of the anaerobic BTF ranged from 0.001 to
0.0014 /s. On the other hand, the reaction rate constant for the aerobic BTF ranged from 0.0011 to 0.0018 /s. The highest reaction rate constant was observed in phase IV of each BTF. In the case of the anaerobic BTF, it correlates with the increase of ethanol loading rate. It is worth to note that increasing ethanol-loading rates favored the growth of microbial population, which resulted in an increase in the biocatalyst, and thus improving the rates of biodegradation. During similar ratio of CF to ethanol the reaction rate constant for anaerobic BTF was always less than that of aerobic BTF which correlates well with the removal efficiencies reported in Table 4-1.
95 | P a g e
4.5.3 Carbon mass balance
The cumulative CO2 equivalent of CF in the influent was compared to the same equivalent in the effluent for both BTFs. The influent cumulative CO2 consists of influent gaseous concentration and influent aqueous inorganic and organic carbon. The effluent CO2 equivalent includes the effluent aqueous inorganic and organic carbon, effluent VSS, gaseous
CO2 and CH4 (only for anaerobic BTF) and effluent CF and ethanol concentrations. Fig. 4-4 presents the cumulative influent and effluent for anaerobic BTF as an example. The CO2 equivalence of all the carbon components was calculated in moles and a cumulative input and output CO2 equivalence of carbon was plotted on sequential time (Fig.4-4). The difference between the influent and effluent carbon on average was 41% with a standard deviation of 8.8%.
A difference of 27% with standard deviations of 3.1% was obtained for aerobic BTF. The carbon recovery for the anaerobic BTF was 59% and the recovery for the aerobic BTF for the four phases was 63% [29]. The loss of influent and effluent carbon was produced as biomass within the BTF. This hypothesis is justified by comparing the loss of carbon to the amount of biomass accumulated within the bed. The cellular composition for typical heterogeneous anaerobic microorganisms is represented as C4.9H9.4NO2.9 and the aerobic filamentous fungi is also presented by C9H15O5N [46].
These compositions were used as the basis for relating the ammonia and nitrate consumed in building up new biomass to estimate the amount of biomass retained within each
BTF. A t-test was performed to compare the results of the carbon consumed and the biomass produced. The anaerobic test results ranged from 7.32 x10-8 to 4.52x10-6 with p- value < 0.05 indicating that the difference between the carbon retained and the biomass produced was
96 | P a g e
statistically significant, therefore, confirming that the loss of carbon within the BTF was utilized for biomass growth.
It is worthwhile to note that the main carbon contributors to the carbon balance for both
BTFs are the gas phase concentrations of the influent and effluent CF and ethanol concentration, and effluent gaseous carbon dioxide. Methane is another effluent gas for the anaerobic BTF. The share of carbon in the liquid phase due to the amount calculated from volatile suspended solids, influent and effluent organic carbon in the aqueous phase could be considered negligible since the total aqueous carbon did not exceed 5% of the total carbon in the system.
4.5.4 Microbial ecological analyses and correlation
The bacterial and fungi structures of anaerobic and aerobic BTFs were studied by using
Ion Torrent PGM system. Samples for the microbial analysis were collected from each BTF after re-acclimation to the different phase when 99 % of the original performance was attained.
To get a high diversity of microbes, inoculums usually come from digested activated sludge or previously cultivated micro flora [47]. For the anaerobic biofilter, initially, microbes were acclimated for CF based culture by using methanogenic bacteria from food waste. Fig. 4-5 shows the relative abundance and the diversity of the anaerobic microbial community observed for phases I to III of the anaerobic BTF. Due to the erratic performance of the anaerobic BTF after day 143, no microbial samples were taken on the last phase (phase IV). The microbial analysis is based on 97% identity of 16S rRNA gene sequences in class level. The figure (Fig. 4-5) provides the results of analysis for the samples collected from port 2 (Fig. 4-1) of each phase. During
Phase I, the most dominant species were A. restrica and A. oryzae (46% and 21%) followed by
97 | P a g e
Geobacter spp. (16%) and Aminivibrio pyruvatiphilus (6%). However, during Phase II, the amounts of A. restrica and A. oryzae reduced to 18% and 37%, respectively.
The retrieved amount of Geobacter spp. also reduced to 2 %. The amount of A. pyruvatiphilus also decreased to less than 1%, while Azonexus fungiphilus (15 %) showed a significant relative abundance than in phase I. The amount of clostridium spp. was also higher in phase II 7% compared to 2% in phase I. In phase III, A. restrica , A. oryzae,. Azonexus fungiphilus. and Anaerobaculum mobile were the dominant species with the relative abundance of 47%, 29%, 6% and 4%, respectively. With the addition of ethanol in the anaerobic BTF system, the growth of A. restrica and A. oryzae were greatly enhanced. Furthermore, the addition of more ethanol on phase II has affected the growth of CF degrading species like A. restrica, A. oryzae and Geobacter spp. which were the dominant species during Phase I. This effect was clearly noticed when the CF feed stream was supplemented with more co metabolite in the BTFs during Phase II, where the concentration of Azonexus fungiphilus and
Anaerobaculum mobile increased significantly from 1% each to 6 and 15% respectively.
Moreover, during phase III with higher co metabolite concentration (100 ppmv), it can be noticed that the growth of A. restrica and A. oryzae increased more than the other dominant species.
In general, the relative abundance of A. oryzae increased with the degradation of CF, which correlates to the corresponding removal efficiency and EC. It is, therefore, speculated that
A. oryzae could be the primary bacteria for the degradation of CF under anaerobic conditions. A. oryzae and A. restrica were the main species in all the three phases. The prevalence of these species has also been reported previously from various microbial utilization and studies related
98 | P a g e
to anaerobic biodegradation. A. oryzae [48]. Similarly, Bae et al. studied the species of A. restrica and found out that it’s a nitrogen-fixing bacteria [49].
In the case of aerobic BTF, Fusarium sp and F. solani were the major species detected for the four phases. Fig. 4-6 provides the fungi community diversity observed over the four phases of aerobic BTF for samples collected from the top port of the biofilter. The figure suggests that significant phase dependent changes in the detected fungi communities of the BTF.
Phase I fed with CF and 5 ppmv of ethanol, the most dominant species were Fusarium sp.,
Aspergillus sp., and Ascotricha sp with relative abundancy of 64%, 15% and 11% respectively.
The availability of F. Solani was 4%. However, on phase II, when the BTF was fed with more ethanol (50 ppmv), the dominant species were Fusarium sp. with 95% followed by F. Solani and
F. nectria haematococca with 2% each. In this phase, the amount of Aspergillus sp., and
Ascotricha sp. reduced to less than 0.3% which supported more growth to Fusarium sp. Other very important observation is that the amount of Fusarium sp. increased more than 30% from the previous phase (phase I). This could be due to the increase in ethanol concentration, which favors more carbon source for the microbes. During phase III, again Fusarium sp. was dominant by 86% and followed by F. solani. at 10%.
As reported in our previous work [29], in this phase the system left to run for more than
100 days and could be the main reason for the increase and dominancy of Fusarium sp. and F.
Solani species over other fungi species within the aerobic BTF. It is also very important to note that, when ethanol concentration increased to 100 ppmv (ratio of 1:20), the percentage of F. solani also increased more than 8% from the previous phase. In addition, a new kind of fungi species called Cylindrocarpon sp.(1%) was detected on this phase. During phase IV, the aerobic
BTF was mainly dominated by Fusarium sp. (59%) and F. Solani (36%). It is interesting to note
99 | P a g e
that F. Solani increased significantly in this phase as compared to the previous phase. It could be attributed to the increase of ethanol concentration to 200 ppmv. Similarly, Cylindrocarpon sp. increased to 4% during this phase. Finally, it can be concluded that the abundance of fungi population might explain the high removal efficiency of CF in the acidic aerobic BTF.
Especially, Fusarium sp. and F. solani were the most dominant and abundant fungi species in this aerobic BTF. Other studies reported that F. solani used to biodegrade n-hexane [50, 51].
Sagar and Singh conducted a study on the biodegradation of lindane pesticide by Fusarium sp., and demonstrated that F. solani biodegraded lindane up to 59.4% [52].
4.6 Conclusion
In this study, we examined the removal of gas phase CF under two environmental conditions (anaerobic and aerobic), and in the presence of ethanol as co metabolite.
Investigations of the biological community structure within the BTFs were also conducted. The use of aerobic fungi BTF under acidic condition successfully enhanced the biodegradation process of CF. The BTF provided more stable performance by having smaller standard deviation in the removal efficiency as compared to the anaerobic BTF. Hence, acidic aerobic BTF had achieved significant improvement in the removal of CF. Operation at acidic pH enhanced greatly the performance providing a removal efficiency around the 80.9% level. Using fungi culture led to higher loading rates that could not be achieved by anaerobic microbial culture. The result obtained from microbial analysis showed that the most dominant fungi, which promote higher removal efficiency, were Fusarium sp. and F. solani. A. oryzae and A. restrica were the responsible bacteria community species responsible for anaerobic BTF. The current study proves the effectiveness of the use of BTF in post aeration processes installed at different points in the
100 | P a g e
water distribution system for the removal of DBPs. The added stability in performance could put more trust in the cost effectiveness of biological treatment of hydrophobic compounds.
101 | P a g e
Table 4-1 Operating conditions for anaerobic and aerobic BTFs degrading chloroform at a loading rate of 0.27 g/m3h.
Phases Operating condition BTF Type I II III IV Anaerobic Influent Ethanol concentration, ppmv 25 50 100 200 Aerobic Anaerobic 44 33 41 35 Operation time, days Aerobic 30 29 122 33 Anaerobic 49±9 52±7 56±7 59±10 Average Chloroform Removal Efficiency (%) Aerobic 69.9±9 71.6±5 75.1±9 80.9±4 Anaerobic 0.13±0.02 0.14±0.01 0.15±0.02 0.16±0.01 Elimination Capacity (g/m3. h) Aerobic 0.21±0.01 0.22±0.01 0.22±0.01 0.23±0.01 Ratio: Methane / Carbon dioxide Anaerobic 1.77 1.95 2.01 2.05
102 | P a g e
Figure 4-1 Schematic diagram of biotrickling filtration system (BTFs)
103 | P a g e
Figure 4-2 Performance of the anaerobic BTF in the four phases with increasing chloroform-to- ethanol, Phase I: (1: 5), Phase II: (1: 10), Phase III: (1: 20), Phase IV: 1: 40
104 | P a g e
Figure 4-3 Reaction rate constants for chloroform for both anaerobic and aerobic BTFs in four corresponding phases ratio of chloroform-to-ethanol (co metabolite). Phase I: (1: 5), Phase II: (1:10), Phase III: (1: 20), Phase IV: (1: 40).
105 | P a g e
Figure 4-4 Carbon mass balance of BTF over 160 operation days where cumulative carbon input
and output are given as CO2 equivalent in mole for the anaerobic process.
106 | P a g e
Figure 4-5 Bacterial community diversity for the three phases of anaerobic BTF for samples collected at the top port of the biofilter for selected ratio of chloroform-to-ethanol Phase I: (1: 5), Phase II: (1: 10) and Phase III: (1: 20).
107 | P a g e
Figure 4-6 Fungi community diversity for the four phases of aerobic BTF for samples collected at the top port of the biofilter for selected chloroform to ethanol. Phase I: (1: 5), Phase II: (1:10), Phase III: (1: 20), Phase IV: (1: 40).
108 | P a g e
4.7 References 1. Gopal, K., et al., Chlorination byproducts, their toxicodynamics and removal from drinking water. Journal of hazardous materials, 2007. 140(1): p. 1-6. 2. Krasner, S.W., et al., The occurrence of disinfection by-products in US drinking water. Journal-American Water Works Association, 1989. 81(8): p. 41-53. 3. Dalvi, A.G.I., R. Al-Rasheed, and M. Javeed, Haloacetic acids (HAAs) formation in desalination processes from disinfectants. Desalination, 2000. 129(3): p. 261-271. 4. Lichtfouse, E., Environmental chemistry: green chemistry and pollutants in ecosystems. 2005, Berlin, Germany: Springer Science & Business Media. 5. Pourmoghaddas, H. and A.A. Stevens, Relationship between trihalomethanes and haloacetic acids with total organic halogen during chlorination. Water Research, 1995. 29(9): p. 2059-2062. 6. Xie, Y., Disinfection By-product Removal Using Point-of-use Carbon Filters. Proc. AWWA WQTC, Quebec, 2005. 7. Gh, A. and G. Gh, Adsorption of humic acid from aqueous solutions onto modified pumice with hexadecyl trimethyl ammonium bromide. Journal of Babol University of Medical Sciences, 2011. 14(1): p. 14-22. 8. Staudinger, J. and P.V. Roberts, A critical compilation of Henry's law constant temperature dependence relations for organic compounds in dilute aqueous solutions. Chemosphere, 2001. 44(4): p. 561-576. 9. LaKind, J.S., S.D. Richardson, and B.C. Blount, The good, the bad, and the volatile: can we have both healthy pools and healthy people? 2010, ACS Publications. 10. McGregor, F., P. Piscaer, and E. Aieta, Economics of treating waste gases from an air stripping tower using photochemically generated ozone. 1988. 11. Delhoménie, M.C., L. Bibeau, and M. Heitz, A Study of the Biofiltration of High‐ Loads of Toluene in Air: Carbon and Water Balances, Temperature Changes and Nitrogen Effect. The Canadian Journal of Chemical Engineering, 2005. 83(2): p. 153-160. 12. Egli, C., et al., Anaerobic dechlorination of tetrachloromethane and 1, 2-dichloroethane to degradable products by pure cultures of Desulfobacterium sp. and Methanobacterium sp. FEMS Microbiology Letters, 1987. 43(3): p. 257-261. 13. Yu, Z. and G.B. Smith, Chloroform dechlorination by a wastewater methanogenic consortium and cell extracts of Methanosarcina barkeri. Water Research, 1997. 31(8): p. 1879-1886. 14. Egli, C., et al., Transformation of tetrachloromethane to dichloromethane and carbon dioxide by Acetobacterium woodii. Applied and environmental microbiology, 1988. 54(11): p. 2819-2824. 15. Egli, C., et al., Transformation of tetra-and trichloromethane to CO2 by anaerobic bacteria is a non-enzymic process. FEMS Microbiology Letters, 1990. 68(1-2): p. 207- 212. 16. Picardal, F.W., et al., Involvement of cytochromes in the anaerobic biotransformation of tetrachloromethane by Shewanella putrefaciens 200. Applied and environmental microbiology, 1993. 59(11): p. 3763-3770. 17. Mikesell, M.D. and S.A. Boyd, Dechlorination of chloroform by Methanosarcina strains. Applied and environmental microbiology, 1990. 56(4): p. 1198-1201.
109 | P a g e
18. Becker, J.G. and D.L. Freedman, Use of cyanocobalamin to enhance anaerobic biodegradation of chloroform. Environmental science & technology, 1994. 28(11): p. 1942-1949. 19. Zitomer, D.H. and R.E. Speece, Methanethiol in nonacclimated sewage sludge after addition of chloroform and other toxicants. Environmental science & technology, 1995. 29(3): p. 762-768. 20. Field, J.A. and R. Sierra-Alvarez, Biodegradability of chlorinated solvents and related chlorinated aliphatic compounds. Reviews in environmental Science and Bio/technology, 2004. 3(3): p. 185-254. 21. Cappelletti, M., et al., Microbial degradation of chloroform. Applied microbiology and biotechnology, 2012. 96(6): p. 1395-1409. 22. Leson, G. and A.M. Winer, Biofiltration: an innovative air pollution control technology for VOC emissions. Journal of the Air & Waste Management Association, 1991. 41(8): p. 1045-1054. 23. Bagley, D.M. and J.M. Gossett, Chloroform degradation in methanogenic methanol enrichment cultures and by Methanosarcina barkeri 227. Applied and environmental microbiology, 1995. 61(9): p. 3195-3201. 24. Bouwer, E.J., B.E. Rittmann, and P.L. McCarty, Anaerobic degradation of halogenated 1-and 2-carbon organic compounds. Environmental science & technology, 1981. 15(5): p. 596-599. 25. Krone, U.E., et al., Coenzyme F430 as a possible catalyst for the reductive dehalogenation of chlorinated C1 hydrocarbons in methanogenic bacteria. Biochemistry, 1989. 28(26): p. 10061-10065. 26. Wahman, D.G., L.E. Katz, and G.E. SPEITEL, Trihalomethane cometabolism by a mixed-culture nitrifying biofilter. Journal (American Water Works Association), 2006. 98(12): p. 48-60. 27. Yoon, I.-K., C.-N. Kim, and C.-H. Park, Optimum operating conditions for the removal of volatile organic compounds in a compost-packed biofilter. Korean Journal of Chemical Engineering, 2002. 19(6): p. 954-959. 28. Balasubramanian, P., L. Philip, and S.M. Bhallamudi, Biotrickling filtration of complex pharmaceutical VOC emissions along with chloroform. Bioresource technology, 2012. 114: p. 149-159. 29. Palanisamy, K., et al., Biofiltration of Chloroform in a Trickle Bed Air Biofilter Under Acidic Conditions. Water, Air, & Soil Pollution, 2016. 227(12): p. 478. 30. Butler, J., C. Ramchandani, and D. Thomson, 58. The solubility of non-electrolytes. Part I. The free energy of hydration of some aliphatic alcohols. Journal of the Chemical Society (Resumed), 1935: p. 280-285. 31. Chen, F., et al., Henry’s law constants of chlorinated solvents at elevated temperatures. Chemosphere, 2012. 86(2): p. 156-165. 32. Gupta, M., et al., Biotransformation rates of chloroform under anaerobic conditions—I. Methanogenesis. Water Research, 1996. 30(6): p. 1377-1385. 33. Wu, S., Anaerobic Biodegradation Patterns for Biodiesel. 2014, University of Cincinnati. 34. Sorial, G.A., et al., Evaluation of trickle bed biofilter media for toluene removal. Journal of the Air & Waste Management Association, 1995. 45(10): p. 801-810. 35. Hassan, A.A. and G. Sorial, Biological treatment of benzene in a controlled trickle bed air biofilter. Chemosphere, 2009. 75(10): p. 1315-1321.
110 | P a g e
36. Association, A.P.H., Standard methods for the examination of water and wastewater. American Public Health Association (APHA): Washington, DC, USA. 2005, Washington, DC, USA: American Public Health Association 37. Zehraoui, A., et al., Impact of alternate use of methanol on n-hexane biofiltration and microbial community structure diversity. Biochemical Engineering Journal, 2014. 85: p. 110-118. 38. Zhai, J., et al., Microbial Community in a Biofilter for Removal of Low Load Nitrobenzene Waste Gas. PloS one, 2017. 12(1): p. e0170417. 39. Bárcenas-Moreno, G., J. Rousk, and E. Bååth, Fungal and bacterial recolonisation of acid and alkaline forest soils following artificial heat treatments. Soil Biology and Biochemistry, 2011. 43(5): p. 1023-1033. 40. Dowd, S.E., et al., Bacterial tag–encoded FLX amplicon pyrosequencing (bTEFAP) for microbiome studies: Bacterial diversity in the ileum of newly weaned salmonella-infected pigs. Foodborne pathogens and disease, 2008. 5(4): p. 459-472. 41. Edgar, R.C., Search and clustering orders of magnitude faster than BLAST. Bioinformatics, 2010. 26(19): p. 2460-2461. 42. Capone, K.A., et al., Diversity of the human skin microbiome early in life. Journal of Investigative Dermatology, 2011. 131(10): p. 2026-2032. 43. Eren, A.M., et al., Exploring the diversity of Gardnerella vaginalis in the genitourinary tract microbiota of monogamous couples through subtle nucleotide variation. PloS one, 2011. 6(10): p. e26732. 44. Swanson, K.S., et al., Phylogenetic and gene-centric metagenomics of the canine intestinal microbiome reveals similarities with humans and mice. The ISME journal, 2011. 5(4): p. 639-649. 45. DeSantis, T.Z., et al., Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Applied and environmental microbiology, 2006. 72(7): p. 5069-5072. 46. Rittmann, B.E. and P.L. McCarty, Environmental biotechnology: principles and applications. 2001, New York: McGraw-Hill. 47. Wagner, M., et al., Microbial community composition and function in wastewater treatment plants. Antonie Van Leeuwenhoek, 2002. 81(1): p. 665-680. 48. Hutchison, J.M., et al., Perchlorate reduction using free and encapsulated Azospira oryzae enzymes. Environmental science & technology, 2013. 47(17): p. 9934-9941. 49. Bae, H.-S., et al., Description of Azospira restricta sp. nov., a nitrogen-fixing bacterium isolated from groundwater. International journal of systematic and evolutionary microbiology, 2007. 57(7): p. 1521-1526. 50. Hernández‐ Meléndez, O., et al., Fungal removal of gaseous hexane in biofilters packed with poly (ethylene carbonate) pine sawdust or peat composites. Biotechnology and bioengineering, 2008. 100(5): p. 864-871. 51. Arriaga, S. and S. Revah, Removal of n-hexane by Fusarium solani with a gas-phase biofilter. Journal of Industrial Microbiology and Biotechnology, 2005. 32(11-12): p. 548- 553. 52. Sagar, V. and D. Singh, Biodegradation of lindane pesticide by non white-rots soil fungus Fusarium sp. World Journal of Microbiology and Biotechnology, 2011. 27(8): p. 1747- 1754.
111 | P a g e
5. Impact of Co metabolite Concentration on the Removal of Trihalomethanes by Biotrickling Filter.
5.1 Abstract
Biotrickling filter (BTF) could be an attractive treatment option for removal of gas phase trihalomethanes (THMs). In this study, mixtures of chloroform (CF) and dichlorobromomethane
(DCBM) are used as model THMs. The BTF was run with nutrients buffered at pH 4 for encouraging the growth of fungi. The loading rates for CF and DCBM were 0.14 and
0.18 g m-3h-1, respectively. Ethanol was used as co-metabolite at different loading rates ranging from 0.57 to 4.59 g m-3h-1. Effect of co metabolite on the BTF performance, removal profile along BTF depth, COD/nitrogen consumption and CO2 production were studied under continuous loading operation conditions. The result showed that higher removal efficiencies of
CF (85±6%) and DCBM (87±6%) were obtained for the higher ethanol loading rate. The maximum reaction rate constant for CF and DCBM obtained from this study were 0.00203 s-1 and 0.0022 s-1 respectively. Similarly, highest ratio of 15.46 chemical oxygen demand (COD) removal to nitrogen utilization was observed at the higher co metabolite loading rate. The study also investigated the microbial community diversity within the biofilter for the different phases. The most responsible fungi for the degradation of CF and DCBM was Fousarium Sp. with dominancy over 98% for the four studied phases.
This chapter is to be submitted to Chemosphere: Bineyam Mezgebe, George A. Sorial, E. Sahle-Demessie, Ashraf Aly Hassan, and Jingrang Lu. (under review). “Impact of Co metabolite Concentration on The Removal Trihalomethanes by Biotrickling Filter." 112 | P a g e
5.2 Introduction
Chlorinated disinfectants are routinely used in water treatment plants to inactivate pathogenic microbes, but they may also react with dissolved natural organic matters to produce a range of disinfection by-products (DBPs), including trihalomethanes (THMs). Trihalomethanes are chlorinated byproducts which include CF, DCBM, dibromochloromethane (DBCM) and bromoform (BF). The major concern of the THMs is possible cancer risks to human and other chronic health effects, which includes cardiac irregularities, loss of a fetus, low birth weight and also pre-term delivery [1]. Exposure to THMs could be through inhalation and dermal contact during regular indoor and outdoor activities. The maximum allowable contaminant level in drinking water for total THMs is 0.08 ppm [2]. In addition, CF and other THMs could also originate from sources other than by-products of water disinfection. In the United States, these direct releases to the environment have been reported annually and include air emissions, surfacewater discharges, underground injections, and releases to land. According to the United
States Protection Agency’s (USEPA’s) Toxic Release Inventory (TRI), approximately 1.1 million pounds of CF and 89 pounds of DCBM were released in 2003 by these routes across the nation [3]. At present, water treatment plants use aeration in the storage tank for reducing THMs
[4]. However, after aerating the utilities emit the gas to the atmosphere with no controlling system. Hence, the passage of the 1990 Amendments to the Clean Air Act have led to the development of more stringent regulations, standards, guidelines, and codes of volatile organic compounds (VOCs) emissions. It also paved the way to the development of processes aimed at reducing air pollution.
In this study, a mixture of CF and DCBM are taken as a model THMs. They are both hydrophobic VOCs, with Henry’s law constant at 25˚C, 0.0025 and 0.0016 atm.m3/mol),
113 | P a g e
respectively. At present, activated carbon, carbon filters and conventional water treatment processes, including clarification, coagulation, flocculation, sedimentation, and filtration are commonly used to remove THMs precursors [2]. All these methods can only remove 30% of the precursors of trihalomethanes [5]. In addition removing CF and DCBM by physical and chemical methods is expensive and may generate a secondary pollutant [6]. Their high Henry's law constant [7] allows alternative approaches for treatment such as gas stripping combined with biological treatment. For instance, Lou et al. [8] reported that the use of biological activated carbon could achieve removal efficiency of CF and DCBM greater than 45% [8]. Similarly,
Buchanan et al. concluded that biologically activated carbon treatment was most effective for the removal of CF and DCBM [9]. Hence, biological treatment techniques for VOC removal can be made cost effective for low concentration and high volume streams as compared to conventional techniques such as incineration, catalytic oxidation, and adsorption but they are safe and eco-friendly [10]. CF and DCBM are volatile THMs and could be transferred from contaminated waters to the gaseous phase by air stripping [11, 12].
Biofiltration is a biological process that uses microorganisms to convert water soluble
VOCs into harmless byproducts. Although most studies show successful biodegradation of CF and DCBM in the liquid phase, there has been limited reported work on the use of biofiltration for the removal of CF and DCBM from gaseous streams. In addition, THMs’ halogen bond could significantly affect the biofiltration process [13]. To overcome this, halogenated organic compounds often require a co metabolite that can increase their biodegradability [13]. A Study has been conducted by Wahman et al. 2011 on co metabolism of gaseous THMs with a nitrifying biofilm in a biofiltration. The result obtained from the study indicated that CF removal was ranging between 13 % and 43 % [14]. Similarly, the same group confirmed that ammonia-
114 | P a g e
oxidizing bacteria can transform the four regulated THMs which include CF, DCBM, DBCM and BF. In their study they demonstrated that biofilters with co metabolite can remove THMs with removals ranging from 7-24% [15]. Other biofiltration studies have been reported for the treatment of gaseous CF with other mixtures of different VOCs [16, 17]. Yoon et al. [16] have shown the degradation potential of nine VOCs (benzene, toluene, m-xylene, o-xylene, styrene,
CF, trichloroethylene, isoprene, and dimethyl sulfide) and found the highest removal was for toluene and the lowest removal was for CF. Balasubramanian et al. [17] also evaluated the biodegradation of CF along with a mixture of VOCs (methanol, ethanol, acetone and toluene) commonly found in pharmaceutical emissions, using a BTF. Their study showed that increasing the rate of CF loading significantly reduced the degradation efficiency of the reactor for the mixture of VOCs.
Additionally, the hydrophobicity nature of CF and DCBM makes them recalcitrant to biodegradation and slow down the mass transfer of this contaminant into the liquid phase within the BTF bed affecting the reaction kinetics. Van Groenestijn et al. discussed that replacing the working consortium in a biofilter from Bacteria to fungi has more advantageous to overcome this obstacle [18]. Fungi are more resistant to acidification and drying out, which is a major advantage of the natural media biofilters but does not necessary count as an advantage in the
BTF. Fungi can stand over a wide pH range and are tolerant to pH fluctuations unlike bacteria which requires neutral pH for sustenance [19]. The aerial mycelia of fungi form a larger surface area in the gas phase than bacterial biofilms, which may facilitate the uptake of hydrophobic volatile compounds overtaking the rate limiting step by accelerating the mass transfer of hydrophobic compounds from the air to the biofilm [18]. Few reported research utilized fungi in the operation of a traditional biofilter and proven to be a better option for insoluble compounds
115 | P a g e
like alkyl benzenes [19] and n-hexane [20]. On the other hand, acidic biofiltration of CF in the presence of fungi, to the best of our knowledge, is rarely reported in literature except our previous work in treating CF with co metabolite (ethanol) [21].
In this study, a BTF was run continuously to degrade gaseous mixtures of CF and
DCBM. The loading rate of the mixtures was kept at 0.14 gm-3h-1for CF and 0.18 gm-3h-1 for
DCBM throughout the experimental study. The experimental plan was designed to operate the
BTF under acidic condition. Ethanol was used as co metabolite at different loading rates. It often emitted together with THMs from pharmaceutical industries and wastewater sources [22, 23].
The study also extends the application of BTF for the removal of THMs emitted from other sources. The primary objective of this study is, therefore, to evaluate the biodegradation of mixtures of gaseous THMs under acidic conditions. The effects of mixtures of THMs to co metabolite ratios, step-increase in the influent concentration of co metabolite on the performance of the BTF were evaluated. The study also investigated the microbial ecology within the BTF to get a deep insight of the factors affecting the performances. Other operational parameters were also investigated such as reaction kinetics, carbon mass balance and chemical oxygen demand
(COD) removal vs nitrogen utilization.
5.3 Materials and methods
5.3.1Materials
CF and DCBM with 99.8% purity were obtained from Fisher Scientific (Pittsburgh, PA,
USA) & Ethanol with 99.5% purity obtained from Sigma Aldrich (St. Louis, MO, USA). CF and DCBM are highly hydrophobic compounds with corresponding Henry’s law constant
116 | P a g e
−3 −3 3 -1 (at 25 °C), KH of 3.67×10 and 2.12×10 atm.m mol , respectively. The KH value of ethanol is known to be 5.1×10−6 atm.m3 mol-1 at 25 °C which clearly indicates its hydrophilicity as compared to CF and DCBM. The measuring sensors for pH and nitrate were acquired from
Accumate Instruments. Genomic DNA extractions of bacterial and fungi strains were performed using the Mo Bio PowerSoil DNA (M Bio Lab, Inc., Carlsbad, CA) Kit, which was done by
Molecular Research LP (MR DNA, Shallowater, TX).
5.3.2 Biotrickling filter (BTF)
A schematic diagram of the BFT system is shown in Fig. 5-1. The system’s column consisted of seven cylindrical glass sections with an internal diameter of 7.6 cm and a total length of 130 cm and is packed with pelletized diatomaceous earth biological support media to a depth of about 60 cm (Celite® 6 mm R-635 Bio-Catalyst Carrier; Celite Corp., Lompoc, CA).
The BTF ran at an average temperature of 22±1 0C and operated in a co-current mode of both gas and liquid flows downward. The BTF was fed with different mixing ratio of mixtures of THMs to ethanol (co metabolite) ranging from 1:5 to 1:40.
The BTF operation was tested for stagnation type of biomass control strategy. In this stagnant period, the system runs with no flow of VOC, nutrient, and air. The duration and frequency for this strategy were 2 days per week for the length of the experimental phase. This biomass control techniques were previously applied in the biofiltration of hydrophilic and hydrophobic VOCs and was proven to be very effective in biomass growth control. More details on biomass control technologies can be found in Hassan and Sorial [24].
In this BTF system, air was used as a carrier gas with a flowrate of 0.5 L min-1 providing a corresponding empty bed residence time (EBRT) of 5 min. The buffered nutrient solution was
117 | P a g e
supplied at an average rate of 2.0 L day-1. The composition of the nutrient solution was used according to the one provided by Zehraoui et al. [25]. The nutrients were supplied at an acidic pH of 4 by the addition of sodium formate buffer to encourage the growth of fungi colonies.
Liquid mixtures of THMs and ethanol were injected via two separate syringe pumps in series and vaporized into the air stream. The mixtures of THMs (CF and DCBM) were at 1:1 ratio.
5.3.3 Sampling and analysis
Gas and Liquid samples were collected daily from the BTF system five days per week for the measurements and composition of feed and effluent gas/ liquid streams. Liquid samples were collected for the measurements of the influent and effluent liquid pH, nitrate, and organic matter.
The gas flow pressure drop across the bed and operating temperature were taken on daily basis.
Gas phase samples were taken manually using airtight syringes. The samples were analyzed for
CF, DCBM, ethanol, and carbon dioxide as a by-product. They were injected into GC – HP,
Column: HP, 608, 30 m X 530 μm film thickness, injection splitless through 5ml sample loop equipped with a flame ionization detector (FID). The GC oven was programmed isothermal at
60 °C (2min) ramped to 90 °C at a rate of 10 oC min-1. The carrier gas (He) flow rate was set at
-1 3.5 mL min at constant flow rate. The FID was used with N2 make-up gas at a flow rate of 30
-1 -1 -1 mL min , a fuel gas flow (H2) of 40 mL min , and airflow of 400 mL min . Retention time for
CF and DCBM were 3.8 min and 4.5 respectively under the above conditions used. Levels of reaction products, such as carbon dioxide (CO2) samples were also taken automatically by GC
HP- TCD from each sampling port in the BTF. The GC oven was programmed isothermal at
60°C (1min), ramped to 115°C at 25°C min-1. The carrier gas (He) flow rate was set at
-1 -1 3.5 mL min , the TCD was used with helium make-up gas at a flow rate of 5 mL min .
118 | P a g e
Liquid samples were collected from the effluent stream of BTF once a week and analyzed for nitrate, volatile suspended solids (VSS) and dissolved total carbon (TC) and dissolved inorganic carbon (IC). The samples were filtered through a 0.45 µm membrane filter (Whatman
Co.). The concentration of nitrate was determined using a nitrate – N Accumate electrode sensor.
TC and IC content of the liquid samples were determined with a Shimadzu total organic carbon analyzer model TOC - L (Shimadzu Corp., Tokyo, Japan). The VSS analysis was conducted by
Standard Method 2540G [27].
5.3.4 Microbial community molecular analysis
Biofilm samples were collected from two sampling ports within the BTF media (see
Fig.1). To get best microbial analysis result, samples from biofilter were collected at the end of each experimental phase [25, 28]. The microbial samples were taken from port 2 (top) and port
5 (bottom) of the BTF and placed in sampling tubes. The samples consisted of about five media pellets covered with biomass were taken to determine the microbial distribution comparison for each phase. All the samples collected were stored in a -20 °C freezer until being sent to molecular research laboratory (Molecular Research LP Shallowater, TX) for biological analysis.
The DNA of microbial mass in the samples was extracted using Mo Bio PowerSoil DNA (M Bio
Lab, Inc., Carlsbad, CA) following manufacturer’s instruction that includes cell breakage steps followed by the addition of detergents and high salt buffers and enzymatic digestion with lysozyme and proteases. For ion torrent sequencing, the 16S rRNA gene V4 variable region
PCR primers 515/806 were used in a single-step 30 cycle PCR using the HotStarTaq Plus Master
Mix Kit (Qiagen, USA) under the following conditions: 94°C for 3 minutes, followed by 28 cycles (5 cycle used on PCR products) of 94°C for 30 seconds, 53°C for 40 seconds and 72°C for
119 | P a g e
1 minute, after which a final elongation step at 72°C for 5 minutes was performed. Sequencing was carried out at Molecular Research LP (www.mrdnalab.com, Shallowater, TX, USA) on an
Ion Torrent Personal Genome machine (PGM) following the manufacturer’s guidelines.
Sequence data were processed using a proprietary analysis pipeline. Sequences were first depleted of barcodes and primers, and those under 150bp or with ambiguous base calls or with homopolymer runs exceeding 6bp were removed. Operational taxonomic units (OTUs), which were defined by clustering at 3% divergence (97% similarity) [29-33], were generated after denoising sequences and removing chimeras. The last OTUs were taxonomically classified using
BLASTn against a database derived from RDPII http://rdp.cme.msu.edu) and NCBI
(www.ncbi.nlm.nih.gov) [34].
5.4. Results and discussion
5.4.1 Biotrickling filter performances
This study evaluated the biodegradations of gaseous mixtures of THMs (CF+DCBM) by using BTF in the presence of co-metabolite at different loading rates. The co-metabolite was mixed with CF and DCBM stream to provide additional electron donor to the micro-organisms.
Ethanol was used as a co-metabolite since it readily mixes with CF, DCBM and water. It is worth noting that the removal efficiency of ethanol was always above 98% for the given loading rate conditions studied for this BTF. Therefore, emphasis is placed on the performance of the
BTF on CF and DCBM. In Phase I, the BTF started up with equal mixtures of THMs (CF and
DCBM) providing total influent THMs concentration of 5 ppmv and ethanol concentration of 25 ppmv. The corresponding CF and DCBM loading rates were 0.14 and 0.18 g/m3.hr, respectively.
The operating conditions and different phases of operation are summarized in Table 5-1. Fig. 5-2
120 | P a g e
also presents a statistical summary of the removal efficiency as a box plot at different loading rates for the BTF. The lower boundary of the box denotes the lower quartile, a line within the box marks the median, and the boundary of the box furthest from zero indicates the upper quartile. Whiskers (error bars) above and below the box indicate the 90th and 10th percentiles. In phase I, the BTF was run for 36 days and the average removal efficiencies for CF and DCBM were 73% and 77%, respectively which provided average elimination capacities (ECs) of
0.11±0.02 and 0.14±0.03 g m-3h-1 for CF and DCBM, respectively (Table 5-1). During phase II, the THMs to ethanol ratio was raised to 1:10 where the removal efficiencies for CF and DCBM increased to 77±6 % and 78±6 % respectively. Their corresponding ECs were 0.12±0.02 and
0.15±0.02 g m-3h-1 for CF and DCBM, respectively. The system was left to run for 41 days.
Mixtures of THMs (CF and DCBM) to ethanol ratio in Phase III was increased to 1:20 providing ethanol loading rate of 2.30 g m-3h-1 while maintaining the same loading rate for both
CF and DCBM. The removal efficiency of CF and DCBM showed a slight increase to 79±8% and 80±9%, respectively. For this phase, the respective ECs for this phase were 0.12±0.01 and
0.16±0.02 g m-3h-1 for CF and DCBM. At the end of this phase, a thick layer of biomass was visible on the top surface of the media, and within the BTF. With the increase of VOCs’ loading rates, more biomass was accumulating on the upper surface of the BTF. This finding could be an indication that cell synthesis and ethanol oxidation were simultaneous processes in the upper section of the BTF. Kim and Sorial and Song and Kinney observed in their study for toluene biofiltration that cell synthesis occur in the upper section of the biofiltration and VOC oxidation within the whole bed [35, 36]. The stagnation abetted in controlling this excessive biomass.
Simillar biomass controlling same technique has been used by Zehraoui et al. for the biofiltration of n-hexane [37] and by Palanisamy et al. for biofiltration of CF [21].
121 | P a g e
On day 110 (phase IV) the loading rate of ethanol was increased to 4.59 g m-3h-1. This phase was run for 38 days. The removal efficiencies for both CF and DCBM increased to 85±6% and 87±6%, respectively. Similarly, in this phase stagnation was used to control the observed high pressure drops across the system. The pressure drop in this phase could be due to the high biomass accumulation within the bed which created BTF short circuit. The mixtures of THMs- ethanol ratio of 1:40 brought a higher elimination capacity of 0.13±0.01 and 0.17±0.01 g m-3h-1 for CF and DCBM compared to the other previous phases (table 5-1). The result also showed that the increase in the loading rates of the co-metabolite enhanced the performances of the BTF.
The same observation has been made in our previous studies. Although, our previous study focused on single solute (only CF), the result conformed that as the concentration of ethanol increased, the removal efficiency of CF also increased accordingly [38]. In our current study, the highest elimination capacity of CF and DCBM were obtained during phase IV (Table 5-1). The reported removal efficiency value of CF by Palanisamy et al. was 80.9% [21]. Similarly,
Wahman et al. 2011 reported that the use of biofilters seeded with three different mixed-culture sources can remove CF and DCBM up to 18% and 75% respectively [14]. However, this current study demonstrated that higher removal efficiencies can be obtained for both CF and DCBM by using fungi BTF under acidic condition. Fig. 5-3 provides a plot of CF and DCBM elimination capacities for each phase. It can be seen from the figure that ECs for CF and DCBM increased considerably with increase of ethanol (co-metabolite). This confirms our previous observation for the removal of CF alone, where further increase in ethanol loading rate caused a rise of CF elimination capacity.
122 | P a g e
5.4.2. Kinetics of the CF and DCBM for the different phases
Gas phase samples were drawn from each port of the BTF one day every week (following stagnation period) to evaluate the reaction rate kinetics of CF and DCBM corresponding to the total VOC loading rates. The samples were taken along the BTFs from ports that are located at
7.6 cm, 23 cm, 38 cm, 53 cm and 60 cm down from the top of the packed bed (fig.1). The kinetic analysis was conducted using the data from sampling ports within the media as there is a possibility of biodegradation on the top portion of the BTF above the media or at the bottom disengagement chamber used for separation of liquid and gas effluents. The BTF is assumed to function as a plug flow reactor, and the removal kinetics was based on the pseudo first-order reaction as a function of the depth of the BTF. Natural logarithm of the ratio of residual CF and
DCBM concentrations at each port to the inlet CF concentration (ln(C/C0)) is plotted against the independent variable, time (seconds). The data were fit to a linear model, and the slopes of the regression represented the reaction rate constants, k (s−1). Fig. 5-4 shows the reaction rate constant for the four phases of the BTF. The results in the figure show higher reaction rate of
0.00203 s-1 for CF and 0.0022 s-1 for DCBM were obtained for Phase IV as compared to the other phases of the BTF. A similar conclusion was given by Palanisamy et al. (2016) that highest reaction rate constant for single solute (CF) was obtained at higher ethanol loading rate
(0.0018 s-1). In this study, lowest reaction rate constants for the CF and DCBM (0.0016 s-1 and
0.0017 s-1) were obtained in phase I. It is worth noting that increasing ethanol-loading rates favored the growth of microbial population resulting in an increase in the biocatalyst, and thus improving the rates of biodegradation. It is also worth noting that the reaction rate constants for
CF obtained in this study is higher than that was reported in our previous research [21].
123 | P a g e
5.4.3 Carbon mass balance
The cumulative CO2 equivalent of CF, DCBM, ethanol, and the nutrients entering and leaving the BTF is presented in Fig. 5-5. To study the carbon cycle within the bed for both the liquid and gaseous phases, all the carbon sources for the feed and products were measured. The influent consisted of gaseous concentrations of CF, DCBM, and ethanol, plus aqueous inorganic and organic carbon. The effluent included aqueous inorganic and organic carbon, a carbon equivalence of volatile suspended solids, gaseous carbon dioxide, and a carbon equivalence of
CF and DCBM and ethanol concentrations. The CO2 equivalence of all the carbon components was calculated in moles and a cumulative input and output CO2 equivalence of carbon was plotted on sequential time (Fig.5-5). The carbon recovery for the BTF was 57 % and the recovery reported from our previous work for single solute for the similar four phases was 63 % [21]. It is stipulated that the loss of influent and effluent carbon was produced as biomass within the
BTF [24]. This hypothesis is justified by comparing the loss of carbon to the amount of biomass accumulated within the bed. The cellular composition for typical filamentous fungi is presented by C9H15O5N [39]. This composition was used as the basis for relating the nitrate consumed in building up new biomass to estimate the amount of biomass retained within the BTF. A t-test was performed to compare the results of the carbon consumed and the biomass produced. The obtained test results was with p- value < 0.05 indicating that the difference between the carbon retained and the biomass produced was statistically significant, therefore, confirming that the loss of carbon within the BTF was utilized for biomass growth.
124 | P a g e
5.4.4 Nitrogen utilization and COD reduction
Microorganisms’ uptake readily available inorganic nitrogen sources such as ammonia
+ − (NH4 ) and nitrate (NO3 ) which is very essential for their growth and development [40]. In this study, nitrate was the only source of nitrogen supplied with the nutrients. Daily analyses of
- influent and effluent concentrations of NO3 –N was performed. The net nitrogen utilization was computed from the NO3-–N in the BTF nutrients and the effluent liquid [37]. The chemical oxidation demand (COD) for ethanol, CF and DCBM degradation are illustrated using the following reactions.
퐶2퐻6 푂 + 3푂2 → 2퐶푂2 + 3퐻2푂 …………………………………………..…….……………(1)
2퐶퐻퐶푙3 + 2퐻2푂 + 푂2 → 2퐶푂2 + 6퐻퐶푙 …………………………………..…………………(2)
− 2퐶퐻퐶푙2퐵푟 + 2퐻2푂 + 푂2 → 2퐶푂2 + 4퐻퐶푙 + 퐵푟 ………………………………..…………(3)
Eqn 1 was used to determine the mass of COD consumed to the mass of ethanol supplied. The ratio of gCOD/gVOC oxidized was 2.09. This value was used to determine the COD consumed from the influent and effluent ethanol in the BTF. Similarly, Eqn 2 represents the oxidation of
CF to its end products, forming a gCOD/gVOC ratio of 0.13. This ratio was used to determine the COD consumed from the influent and effluent CF. Eqn 3 represents the oxidation of DCBM to its end products, forming a gCOD/gVOC ratio of 0.10. This ratio was used to determine the
COD consumed from the influent and effluent DCBM.
The net chemical oxygen demand was calculated as the difference between COD of the feed and the COD of the effluent gas and liquid streams [37]. Fig. 5-6 shows dimensionless
CODremoval/Nutilization (CODR/NU) ratios plotted against the total loading rates of ethanol and mixtures of THMs in box plots. Phase I produced a CODR/NU ratio of 3 for a total VOCs loading
125 | P a g e
rate of 0.89 g m-3h-1 and increased to 5.12 for a total loading rate of 1.47 g m-3h-1 in phase II.
-3 -1 During phase III, for the total loading rate of 2.62 g m h the corresponding CODR/NU ratio increased to 7.58. The CODR/NU further increased to 15.46 at phase IV operated with the highest VOC loading rate of 4.91 g m-3h-1. The ratios showed apparent dependency on the loading rate for all the phases. The maximum reported CODR/NU value for single CF was 14.5
[21]. It is worth mentioning that the use of ethanol, a high yield VOC, contributed greatly to the high nitrogen consumption.
5.4.5 Microbial ecological analyses and correlation
Microbial diversity structure of the BTF was studied for each phase by using Ion Torrent
PGM system. Samples for the microbial analysis were collected from the BTF after re- acclimation to a phase was attained which is defined by a steady performance. Microorganisms are the core component of biofiltration. To get a high diversity of microbes, inoculums usually come from digested activated sludge or previously cultivated micro flora [41]. For this BTF, initially, microbes were acclimated from previous biofilter used for the removal of CF alone studied by Palanisamy et al.[21]. Fig. 5-7 (A& B) provides the fungi community diversity observed in this biofilter for samples collected from the top (port 2) and bottom (port 5) ports of each phase (Fig. 5-1).
On phase I, the BTF was running under acidic conditions and fed with 5 ppmv of mixtures of CF and DCBM and 25 ppmv of ethanol, the most dominant species at port 2 was
Fusarium sp., (98.4%) followed by F. Solani at 1.3 %. Whereas in port 5, Fusarium sp., was dominant by 96.4% followed by F. Solani and Mnemiopsis leidyi with 1.28 %, and 1.18 %, respectively. During phase II, when the BTF was fed with more ethanol (50 ppmv), at port 2 and
126 | P a g e
5, the amount of Fusarium sp showed a slight increase to 98. 5 %. In phase III, when the co metabolite was increased to 100 ppmv, Fusarium sp. dominated at 98.7 % as compared to 1.15
% for F. solani (port 2). In port 5, Fusarium sp., reduced to 93.82 % and another new species
Mnemiopsis leidyi showed at 4 %. In phase IV, with higher ethanol concentration (200 ppmv), the BTF was dominated by Fusarium sp.(98.6%). The relative abundancy of F. oxysporum and
F. Solani (port 2) for this phase were 0.5 % and 0.4%, respectively. At port 5 of this phase, similarly F. oxysporum was the dominant followed by F. Solani at 1.25%. Hence, it can be concluded that Fusarium sp. was the most dominant fungi for the degradation of CF and DCBM.
Similarly, our previous study for the removal of single solute (CF), the dominant fungi was
Fusarium sp.as well [21].
5.6 Conclusion
The developed BTF for the removal of gaseous THMs from flowing stream has a potential large-scale application. This study investigated the effect fungi based BTF for the removal of mixtures of THMs (CF and DCBM) under acidic condition. Ethanol was used as a co metabolite in enhancing the degradation of mixtures. The ratios of mixtures of THMs to ethanol were 1:5, 1:10, 1:20 and 1:40. THMs influent loading rates for CF and DCBM kept constant at
0.14 g m-3h-1 and 0.18 g m-3h-1, respectively. While for ethanol the influent loading rates varied from 0.57 to 4.59 0.14 g m-3h-1. The maximum elimination capacity for CF was 0.13 g m-
3h-1 and for DCBM was 0.17 g m-3h-1 for the given influent loading rates. This maximum elimination capacity resulted when the ratio of THMs to ethanol was 1:40. This is due to high bioavailability of mixtures in presence of ethanol within the BTF. However, for lower ethanol loading rate, lower elimination capacity of mixtures observed (ratio of 1:5). It is worth noting
127 | P a g e
that treating THMs in presence of ethanol is accompanied by high nitrogen consumption of 0.17 g d-1 for the BTF. This entails that, for the current operation conditions, nitrogen concentration in the influent nutrient need to be maintained in the average of 0.525 g d-1. Maintaining a sustainable residual of nitrogen in the effluent waters will ameliorate the ratio CODR/NU consumed. The microbial ecology analysis result showed that Fusarium sp was the most dominant fungi responsible for the degradation of CF and DCBM. The results of this study prove that BTF is an effective technique to treat gas phase mixtures of THMs through co metabolism under acidic condition. Hence, BTF is an effective technique to treat gas phase THMs which emit from water treatment and other industries.
128 | P a g e
Table 5- 1 Operating conditions of the BTF degrading 1:1 mixtures of CF and DCBM under acidic condition. Initial concentration of the mixtures was 5 ppmv
Phases Operating condition I II III IV CF DCBM CF DCBM CF DCBM CF DCBM Influent Ethanol 25 50 100 200 concentration, ppmv Operation time, days 36 41 33 38 Removal Efficiency, % 73±11 77±10 77±6 78±6 79±8 80±7 85±6 87±6 Loading Rate (g m-3h-1) 0.14 0.18 0.14 0.18 0.14 0.18 0.14 0.18 Elimination Capacity 0.11±0.02 0.14±0.03 0.12±0.02 0.15±0.02 0.12±0.01 0.16±0.02 0.13±0.01 0.17±0.01 (g m-3h-1)
129 | P a g e
Figure 5-1 Schematic diagram of the biotrickling filters (BTF)
130 | P a g e
Figure 5-2 Performance of the BTF in the four phases. Phase I: 1: 5 ratio of mixtures (CF and DCBM) to ethanol (co metabolite), phase II: 1: 10 ratio mixtures to ethanol, phase III: 1: 20 ratio of mixtures to ethanol and phase IV: 1: 40 ratio of mixtures to ethanol.
131 | P a g e
Figure 5-3 Elimination capacities for mixtures in four phases. Phase I: 1: 5 ratio of mixtures of THMs to ethanol (co metabolite), phase II: 1: 10 ratio of mixture of THMs to ethanol, phase III: 1: 20 ratio of mixture of THMs to ethanol and phase IV: 1: 40 ratio of mixture of THMs to ethanol.
132 | P a g e
Figure 5-4 Reaction rate constants for mixtures of THMs in four phases. Phase I: 1: 5 ratio of mixtures of THMs to ethanol (co metabolite), phase II: 1: 10 ratio of mixture of THMs to ethanol, phase III: 1: 20 ratio of mixture of THMs to ethanol and phase IV: 1: 40 ratio of mixture of THMs to ethanol.
133 | P a g e
Figure 5-5 Carbon mass balance: Cumulative carbon input and output as CO2 equivalent in mole for the BTF
134 | P a g e
Figure 5-6 Ratio of chemical oxygen demand (COD) to nitrogen utilization vs total loading rates of ethanol and mixtures of THMs. Phase I: 1: 5 ratio of mixture of THMs to ethanol (co metabolite), Phase II: 1: 10 ratio of mixture of THMs to ethanol, Phase III: I:20 mixture of THMs to ethanol and Phase IV: 1: 40 ratio of mixture of THMs to ethanol.
135 | P a g e
Figure 5-7 A) Fungi community diversity for all the four phases of the BTF for samples collected at the top port (port 2) of the biofilter. B) at the bottom port (port 5) of the biofilter Phase I: 1: 5 ratio of mixture of THMs to ethanol (co metabolite), Phase II: 1: 10 ratio of mixture of THMs to ethanol, Phase III: I:20 mixture of THMs to ethanol and Phase IV: 1: 40 ratio of mixture of THMs to ethanol
136 | P a g e
5.7 References 1. Krasner, S.W., The formation and control of emerging disinfection by-products of health concern. Philosophical Transactions of the Royal Society of London A: Math., Phys. Eng. Sci., 2009. 367(1904): p. 4077-4095. 2. Xie, Y., Disinfection By-product Removal Using Point-of-use Carbon Filters. Proc. AWWA WQTC, Quebec, 2005. 3. Watts, P., G. Long, and M. Meek, Concise international chemical assessment document 58: chloroform. World Health Organization (IPCS), Geneva, 2004. 4. McDonnell, B.E., Controlling disinfection by-products within a distribution system by implementing bubble aeration within storage tanks. 2012, University of Cincinnati. 5. Ghanizadeh, G. and G. Asgari, Adsorption of humic acid from aqueous solutions onto modified pumice with hexadecyl trimethyl ammonium bromide. J. Babol Univer. of Medical Sci., 2011. 14(1): p. 14-22. 6. Zilli, M., et al., Phenol removal from waste gases with a biological filter by Pseudomonas putida. Biotechnol. Bioeng., 1993. 41(7): p. 693-699. 7. Staudinger, J. and P.V. Roberts, A critical compilation of Henry's law constant temperature dependence relations for organic compounds in dilute aqueous solutions. Chemosphere, 2001. 44(4): p. 561-576. 8. Lou, J.-C., et al., Removal of trihalomethanes and haloacetic acids from treated drinking water by biological activated carbon filter. Water, Air, Soil Pollut., 2014. 225(2): p. 1-9. 9. Buchanan, W., F. Roddick, and N. Porter, Removal of VUV pre-treated natural organic matter by biologically activated carbon columns. Water Res., 2008. 42(13): p. 3335- 3342. 10. Delhoménie, M.C., L. Bibeau, and M. Heitz, A Study of the Biofiltration of High‐ Loads of Toluene in Air: Carbon and Water Balances, Temperature Changes and Nitrogen Effect. The Canadian J. Chem. Eng., 2005. 83(2): p. 153-160. 11. McGregor, F., P. Piscaer, and E. Aieta, Economics of treating waste gases from an air stripping tower using photochemically generated ozone. J. International Ozone Assoc., 1988. 12. LaKind, J.S., S.D. Richardson, and B.C. Blount, The good, the bad, and the volatile: can we have both healthy pools and healthy people? Environ. Sci. Technol., 2010. 13. Leson, G. and A.M. Winer, Biofiltration: an innovative air pollution control technology for VOC emissions. J. Air & Waste Manage. Assoc., 1991. 41(8): p. 1045-1054. 14. Wahman, D.G., L.E. Katz, and G.E. Speitel, Performance and biofilm activity of nitrifying biofilters removing trihalomethanes. Water Res., 2011. 45(4): p. 1669-1680. 15. Wahman, D.G., et al., Ammonia-oxidizing bacteria in biofilters removing trihalomethanes are related to Nitrosomonas oligotropha. Appl. Environ. Microbiol., 2011. 16. Yoon, I.-K., C.-N. Kim, and C.-H. Park, Optimum operating conditions for the removal of volatile organic compounds in a compost-packed biofilter. Korean J. Chem. Eng., 2002. 19(6): p. 954-959. 17. Balasubramanian, P., L. Philip, and S.M. Bhallamudi, Biotrickling filtration of complex pharmaceutical VOC emissions along with chloroform. Biores. Technol., 2012. 114: p. 149-159. 18. Van Groenestijn, J., W. Van Heiningen, and N. Kraakman, Biofilters based on the action of fungi. Water Sci. and Technol., 2001. 44(9): p. 227-232.
137 | P a g e
19. Kennes, C. and M.C. Veiga, Fungal biocatalysts in the biofiltration of VOC-polluted air. J. Biotechnol., 2004. 113(1): p. 305-319. 20. Hassan, A. and G.A. Sorial, A comparative study for destruction of n-hexane in trickle bed air biofilters. Chem. Eng. J., 2010. 162(1): p. 227-233. 21. Palanisamy, K., et al., Biofiltration of Chloroform in a Trickle Bed Air Biofilter Under Acidic Conditions. Water, Air, Soil Pollut., 2016. 227(12): p. 478. 22. Balasubramanian, P., L. Philip, and S.M. Bhallamudi, Biodegradation of chlorinated and non-chlorinated VOCs from pharmaceutical industries. Appl. Biochem. Biotechnol., 2011. 163(4): p. 497-518. 23. Cecen, F. and O. Aktas, Activated carbon for water and wastewater treatment: Integration of adsorption and biological treatment. 2011: John Wiley & Sons. 24. Hassan, A.A. and G. Sorial, Biological treatment of benzene in a controlled trickle bed air biofilter. Chemosphere, 2009. 75(10): p. 1315-1321. 25. Zehraoui, A., et al., Impact of alternate use of methanol on n-hexane biofiltration and microbial community structure diversity. Biochem. Eng. J., 2014. 85: p. 110-118. 26. Sorial, G.A., et al., Evaluation of trickle bed biofilter media for toluene removal. J. Air Waste Manage. Assoc., 1995. 45(10): p. 801-810. 27. APHA, A.W.W.A.A.P.W.A.W.E.F., Standard methods for the examination of water and wastewater. American Public Health Assoc. (APHA): Washington, DC, USA. 2017. 28. Zhai, J., et al., Microbial Community in a Biofilter for Removal of Low Load Nitrobenzene Waste Gas. PloS one, 2017. 12(1): p. e0170417. 29. Dowd, S.E., et al., Bacterial tag-encoded FLX amplicon pyrosequencing (bTEFAP) for microbiome studies: bacterial diversity in the ileum of newly weaned Salmonella-infected pigs. Foodborne Pathog. and Dis., 2008. 5(4): p. 459-472. 30. Edgar, R.C., Search and clustering orders of magnitude faster than BLAST. Bioinform., 2010. 26(19): p. 2460-2461. 31. Capone, K.A., et al., Diversity of the human skin microbiome early in life. J. Investigative Dermatolo., 2011. 131(10): p. 2026-2032. 32. Eren, A.M., et al., Exploring the diversity of Gardnerella vaginalis in the genitourinary tract microbiota of monogamous couples through subtle nucleotide variation. PloS one, 2011. 6(10): p. e26732. 33. Swanson, K.S., et al., Phylogenetic and gene-centric metagenomics of the canine intestinal microbiome reveals similarities with humans and mice. The ISME J., 2011. 5(4): p. 639-649. 34. DeSantis, T.Z., et al., Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl. Environ. Microbiol., 2006. 72(7): p. 5069-5072. 35. Kim, D. and G.A. Sorial, Role of biological activity and biomass distribution in air biofilter performance. Chemosphere, 2007. 66(9): p. 1758-1764. 36. Song, J. and K.A. Kinney, Effect of vapor‐ phase bioreactor operation on biomass accumulation, distribution, and activity: Linking biofilm properties to bioreactor performance. Biotechnol. and Bioeng., 2000. 68(5): p. 508-516. 37. Zehraoui, A., A.A. Hassan, and G.A. Sorial, Effect of methanol on the biofiltration of n- hexane. J. Hazard. Mater., 2012. 219: p. 176-182. 38. Palanisamy, K., et al., Biofiltration of Chloroform in a Trickle Bed Air Biofilter Under Acidic Conditions. Water, Air, & Soil Pollution, 2016. 227(12): p. 478.
138 | P a g e
39. Bruce, E.R. and L.M. Perry, Environmental biotechnology: principles and applications. New York: McGrawHill. Vol. 400. 2001, New York: McGraw-Hill 40. Moe, W.M., et al., Removal of the sesquiterpene β-caryophyllene from air via biofiltration: performance assessment and microbial community structure. Biodegradation, 2013. 24(5): p. 685-698. 41. Wagner, M., et al., Microbial community composition and function in wastewater treatment plants. Antonie Van Leeuwenhoek, 2002. 81(1-4): p. 665-680.
139 | P a g e
6. Effectiveness of Bio surfactant without Co metabolite in the Removal of Trihalomethanes by Biotrickling Filter
6.1 Abstract
Laboratory scale-studies on the biodegradation of mixtures of THMs (Chloroform (CF)
and Dichlorobromomethane (DCBM)) in two independent biotrickling filters (BTFs) were
carried out. One of the BTFs was using ethanol as co metabolite and the other one was using
surfactin (bio surfactant) without co metabolite. The result showed that, adding co metabolite at
higher rate resulted in removal efficiencies of 85% for CF and 87% for DCBM. Whereas for the
same loading rate, the use of surfactin without co metabolite showed similar removal efficiency
of 85% and 80% for CF and CBM, respectively. Further studies were conducted to investigate
and understand the microbial diversity within both BTFs. The result indicated that for BTF with
co metabolite Fusarium sp. was the most dominant fungi over 98% followed by F. Solani with
less than 2% . Whereas, F. oxysporum and Fusarium sp. were the dominant fungi for the BTF
with surfactin. Before introducing the surfactin into the BTF, batch experiment was conducted to
evaluate the effectiveness of synthetic surfactant as compared to a bio surfactant (surfactin). In
this regard, vials with surfactin showed better performance than vials with Tomadol 25 – 7
(synthetic surfactant).
This chapter is to be submitted to Chemical Engineering Journal: Bineyam Mezgebe, George A. Sorial, E. Sahle- Demessie, David Wendell. (under review). “Effectiveness of Bio surfactant without Co metabolite in the Removal of Trihalomethanes by Biotrickling Filter."
140 | P a g e
6.2 Introduction
Trihalomethanes (THMs) are formed during the chlorination of drinking water supply and emitted from other sources. They are formed when the excess chlorine present in the system reacts with the natural organic matter (NOM) and produces disinfectant byproducts with harmful long-term effects [1, 2]. The production of this harmful disinfection by products (DBPs) postures a chronic health risk. The common THMs include CF, DCBM, dibromochloromethane (DBCM) and bromoform (BF) [3]. Under the Safe Drinking Water Act (SDWA), the United States
Environmental Protection Agency (USEPA) establishes a Maximum Contaminant Level (MCL) of 0.08 mg/L for total THMs [4]. In addition, CF and other THMs could also originate from sources other than by-product. In the United States, these direct releases to the environment have been reported annually and include air emissions, surface water discharges, underground injections, and releases to land. At present water treatment plants that use aeration for reducing
THMs, emit the gas to the atmosphere. Hence, the passage of the 1990 Amendments to the Clean
Air Act have led to the development of more stringent regulations, standards, guidelines, and codes of volatile organic compounds (VOCs) emissions. According to the United States
Protection Agency’s (USEPA’s) Toxic Release Inventory (TRI), approximately 1.1 million pounds of CF and 89 pounds of DCBM were released in 2003 by these routes across the Nation
[5]. It paved the way to the development of processes aimed at reducing air pollution. The formation THMs in drinking water also has highlighted the need for exploring alternate disinfectants and new treatment technologies.
In this study, CF and DCBM are taken as a model DBPs. They are hydrophobic volatile organic compounds (VOC) from air streams, with Henry’s law constant at 25˚C, 0.0025 and
0.0016 atm.m3/mol, respectively. Both are recalcitrant to biodegradation. Several physical and
141 | P a g e
chemical removal methods such as adsorption, advanced oxidation [6], and air stripping [7] are used to treat THMs. Removing these THMs by physical and chemical methods is expensive and may generate a secondary pollutant [8, 9]. Their high Henry's law constant [10] allows alternative approaches for treatment such as gas stripping [3, 11, 12] combined with biological treatment. Biological treatment techniques for VOC removal have several advantages; it’s not only cost effective as compared to conventional techniques such as incineration, catalytic oxidation, and adsorption but it’s safe and eco-friendly [13] . Biofiltration is a biological process that uses microorganisms to convert water soluble VOCs into harmless byproducts. Although most studies show successful biodegradation of CF and DCBM in the liquid phase, there has been limited reported work on the use of biofiltration for the removal of CF and DCBM from gaseous streams.
The use of biofiltration technique has been reported for the biotreatment of CF with other mixtures of different VOCs [8, 14]. Yoon et al. [14] have shown the degradation potential of nine VOCs including CF and found the highest removal was for toluene and the lowest removal was for CF. Balasubramanian et al. [8] also evaluated the biodegradation of CF along with a mixture of VOCs commonly found in pharmaceutical emissions, using a biotrickling filter. Their study showed that increasing the rate of CF loading significantly reduced the degradation efficiency of the reactor for the mixture of VOCs. Additionally, in a study of co metabolism of
THMs (including CF and DCBM) conducted with a nitrifying biofilm in a biofiltration by
Wahman et al. 2007 obtained a CF removal ranging between 13% and 43%[15]. Similarly, the same group confirmed that ammonia-oxidizing bacteria are capable of aerobically transforming the four regulated THMs. As a result they reported that biofilters can removed THMs with removals ranging from 7-24% [16].
142 | P a g e
An increase in the bioavailability of VOC may enhance the rate of biodegradation if mass transfer is limiting. Synthetic ssurfactants or biological origin surfactants have been used to improve bioavailability of compounds. They can be introduced in the biofiltration system as a means for enhancing solubility. The use of synthetic surfactants in enhancing the bioavailability of hydrophobic compounds by facilitating their biotransformation under aerobic conditions has been studied by several researchers [17-19]. The use of synthetic surfactants in BTF enhanced n- hexane [20], toluene [21], and styrene [22] degradation in fungi biofilter. Yuan et al. (2010) also reported in their study that nonionic surfactant of TX-100 greatly increased the dechlorination of chlorobenzenes [23]. However, to the best of our knowledge, the use of bio surfactant like surfactin in biofilter system is rarely reported in literature. Surfactin is a lipopeptide-type bio surfactant which obtained from the gram-positive microorganism, Bacillus subtilis[6]. It is a supernatant of B. subtilis. B. subtilis is rod-shaped and gram-positive bacteria, which can tolerate extreme environmental conditions [6].
The hydrophobic nature of CF and DCBM could slow down their mass transfer into the liquid phase within the filter bed affecting their reaction kinetics. When compounds like CF and
DCBM are less soluble, it becomes more reluctant to be treated in a biofilter [16]. Van
Groenestein et al. [24] discussed that replacing the working consortium in a biofilter from bacteria to fungi has more advantageous. Fungi are more resistant to acidification and drying out, which is a major advantage of the natural media biofilters but does not necessary count as an advantage in the BTF. Fungi are stand over a wide pH range and are tolerant to pH fluctuations unlike bacteria which requires neutral pH for sustenance [25] . The aerial mycelia of fungi form a larger surface area in the gas phase than bacterial biofilms, which may facilitate the uptake of hydrophobic volatile compounds overtaking the rate limiting step by accelerating the mass
143 | P a g e
transfer of hydrophobic compounds from the air to the biofilm[24, 26]. In few researches fungi were utilized in the operation of a traditional biofilter and proven to be a better option for insoluble compounds like alkyl benzenes [25] and n-hexane [26]. On the other hand, acidic biofiltration of CF and DCBM in the presence of fungi, to the best of our knowledge, is rarely reported in literature except our previous work in treating CF with co metabolite [27].
In this study, two independent BTFs, namely BTF-A and B were used for the removal of mixtures of THMs. BTF-A run in the presence of co metabolite at different loading rates and the loading rate for the mixtures were kept at 0.14 g/(m3.h) for CF and 0.18 g/(m3.h) for DCBM throughout this experiment. And BTF-B run in the presence of surfactin without co metabolite with the loading rates of CF ranging from 0.14 to o.41 g/(m3.h) and DCBM ranging from 0.18 to
0.55 g/(m3.h) respectively. To enhance the fungi growth, both BTFs were operating under acidic environment. To get a deep insight of the factors affecting performances, the study investigated the microbial ecology within both BTFs.
6.3 Materials and methods
6.3.1Materials
The volatile THMs tested in the study were CF and DCBM with 99.8% purity obtained from Fisher Scientific (Pittsburgh, PA, USA) & Ethanol with 99.5% purity obtained from Sigma
Aldrich (St. Louis, MO, USA). CF and DCBM are highly hydrophobic compounds with
−3 −3 3 corresponding Henry’s law constant, KH of 3.67×10 and 2.12×10 atm.m /mol, respectively.
−6 3 The KH value of the hydrophilic ethanol is known to be 5.1×10 atm.m / mol at 25 °C. Two independent BTFs were employed in this study. BTF-A received continuous feed of mixtures of
144 | P a g e
THMs with ethanol as a co metabolite. BTF-B was fed continuously with the same flow of mixtures of THMs with surfactin seeded to the BTF. The surfactin for batch study was obtained from Sigma Aldrich (St. Louis, MO, USA). Genomic DNA extractions of bacterial and fungi strains were performed using the Mo Bio PowerSoil DNA (M Bio Lab, Inc., Carlsbad, CA) Kit which was done by Molecular Research LP (MR DNA, Shallowater, TX).
6.3.2 Biotrickling filter (BTF)
Fig. 6-1 shows the schematic diagram of both BTF-A and B. The columns of each BTF consists of seven cylindrical glass sections with an internal diameter of 7.6 cm and a total length of 130 cm and is packed with pelletized diatomaceous earth biological support media to a depth of about 60 cm (Celite® 6 mm R-635 Bio-Catalyst Carrier; Celite Corp., Lompoc, CA). Both
BTFs ran at an average temperature of 22±1 0C and operated in a co-current mode of both gas and liquid flows downward. BTF-A was fed with different mixing ratio of THMs (CF and
DCBM) to ethanol (co metabolite) ranging from 1:5 to 1:40. BTF-A was seeded with an aerobic microbial culture pre-acclimated to CF and ethanol, which had been obtained from a previous operation of the biofilter [28]. Whereas, BTF-B was fed with different loading rates ranging from 5 to 15 ppmv of equal mixtures of CF and DCBM with no co metabolite. Prior to operation of BTF-B an enriched solution of microorganisms was prepared. This solution contained the liquid effluent of BTF-A which already contained THMs degrading microorganisms and 10 mg/L of surfactin. Initially, surfactin was extracted from B. subtilis strain
#21332 (ATCC). B. subtilis were grown at 300C for 48-72 hours in LB broth. One-liter cultures were rotated at 250 rpm in flat-bottom flasks until collection, cells were spun down at 12K g's for 20 min to separate surfactin containing supernatant from bacteria. The used growth media
145 | P a g e
containing surfactin was decanted and applied to the BTF without further purification after B. subtilis cells were removed. The enriched surfactin solution was acclimated at pH 4 by using formate buffer with HCl. An air stream containing THMs (CF and DCBM) was bubbled through the solution for 45 days until the growth of new biomass was visible. The pH was continuously monitored, and it did not increase above 4 during this period. BTF-B was then seeded with this solution by pouring the solution on the media and leaving the solution intact within the media for
120 minutes and then drained.
The BTFs operations were tested for stagnation type of biomass control strategy. In the stagnation period both systems receive no flow of VOC, nutrient, and air. The duration and frequency for this strategy were 2 days per week throughout the duration of the study. This biomass control technique was previously applied in the biofiltration of hydrophilic and hydrophobic VOCs and was proven to be very effective in controlling excess biomass growth
[29, 30].
In both BTF systems, air was used as a carrier gas with a flowrate of 0.5 L/min with a corresponding empty bed residence time (EBRT) of 5 min. A buffered nutrient solution containing nitrate as a nitrogen source was supplied at an average rate of 2.0 L/day. The composition of the nutrient solution was used according to Hassan et al [29]. The nutrients were supplied at an acidic pH of 4 by the addition of sodium formate buffer to encourage the growth of fungi colonies. For BTF-A, liquid mixtures of THMs and ethanol were injected via two separate syringe pumps in series and vaporized into the air stream. However, for BTF-B, only liquid CF and DCBM with 1:1 ratio were injected via syringe pump and surfactin was seeded to the BTF.
146 | P a g e
6.3.3 Sampling and analysis
Gas and liquid samples were collected daily from the BTF system five days per week for the measurements and composition of feed and effluent gas/ liquid streams. Liquid samples collected for the measurements of the influent and effluent liquid pH, Nitrate, organic matter.
The gas flow pressure drop across the bed and operating temperature were also taken. Gas phase samples were taken manually using airtight syringes. The samples were analyzed for CF,
DCBM, ethanol (only for BTF-A), and carbon dioxide as a by-product. They were injected into
GC – HP, Column: HP, 608, 30 m X 530 μm film thickness, injection splitless through 1ml sample loop equipped with a flame ionization detector (FID). The GC oven was programmed isothermal at 60 °C (2min) ramped to 90 °C at a rate of 10 oC/min. The carrier gas (He) flow rate was set at 3.5 mL/min at constant flow rate. The FID was used with N2 make-up gas at a flow rate of 30 mL/min, a fuel gas flow (H2) of 40 mL/min and airflow of 400 mL/min.
Retention time for CF and DCBM were 3.8 and 4.5 min under the above conditions used.
Levels of reaction products, such as carbon dioxide (CO2) samples were also taken automatically by GC HP- TCD from each sampling port in the BTF. The GC oven was programmed isothermal at 60°C (1min), ramped to 115°C at 25°C /min. The carrier gas (He) flow rate was set at 3.5 mL/min, the TCD was used with N2 make-up gas at a flow rate of 5 mL/min.
Liquid samples were collected from the effluent stream of each BTF once a week and analyzed for volatile suspended solids (VSS) and total organic carbon (TOC). The samples were filtered through a 0.45 µm membrane filter (Whatman Co.) and analyzed for influent and effluent concentrations of nitrate, dissolved total carbon, dissolved inorganic carbon, and volatile suspended solids. The concentration of nitrate was determined using a nitrate – N Accumate electrode sensor. Dissolved total carbon and dissolved inorganic carbon content of the liquid
147 | P a g e
samples were determined with a Shimadzu total organic carbon analyzer model TOC - L
(Shimadzu Corp., Tokyo, Japan). The volatile suspended solids analysis was conducted by
Standard Method 2540G [31].
6.3.4 Microbial community molecular analysis
Biofilm samples were collected from port 2 (top sampling port) within each BTF media as shown in Fig.6-1. The samples were taken when the biofilters were running at the stated different phases. To get best microbial analysis result, samples from biofilter were collected at the end of each experimental phase [30, 32]. The microbial samples were then placed in sampling tubes. The samples consisted of about five media pellets covered with biomass suspended in liquid were taken to determine the microbial distribution comparison for each phase. All the samples collected were stored in a -20 °C freezer prior sending to molecular research laboratory (Molecular Research LP Shallowater, TX) for biological analysis. The DNA of microbial mass in the samples was extracted using Mo Bio PowerSoil DNA (M Bio Lab, Inc.,
Carlsbad, CA) following manufacturer’s instruction that includes cell breakage steps followed by the addition of detergents and high salt buffers and enzymatic digestion with lysozyme and proteases. For ion torrent sequencing, the 16S rRNA gene V4 variable region PCR primers
515/806 were used in a single-step 30 cycle PCR using the HotStarTaq Plus Master Mix Kit
(Qiagen, USA) under the following conditions: 94°C for 3 minutes, followed by 28 cycles (5 cycle used on PCR products) of 94°C for 30 seconds, 53°C for 40 seconds and 72°C for 1 minute, after which a final elongation step at 72°C for 5 minutes was performed. Sequencing was carried out at Molecular Research LP (www.mrdnalab.com, Shallowater, TX, USA) on an Ion
Torrent Personal Genome machine (PGM) following the manufacturer’s guidelines. Sequence
148 | P a g e
data were processed using a proprietary analysis pipeline. Sequences were first depleted of barcodes and primers, and those under 150bp or with ambiguous base calls or with homopolymer runs exceeding 6bp were removed. Operational taxonomic units (OTUs), which were defined by clustering at 3% divergence (97% similarity) [33-37], were generated after denoising sequences and removing chimeras. The last OTUs were taxonomically classified using BLASTn against a database derived from RDPII http://rdp.cme.msu.edu) and NCBI (www.ncbi.nlm.nih.gov) [38].
6.4. Results
6.4.1Batch comparison study of synthetic and bio surfactant for the solubility of CF in fungi based acidic nutrient solution
The goal of this batch study was to compare the widely used synthetic surfactant to the bio surfactant for CF’s degradation. The procedure utilized for this study was similar to the one conducted by Hassan et al. [39]. The experiment was performed by placing 200 mL of acidic nutrient with fungi in a 250 mL amber vials. Two kinds of surfactants namely, tomadol 25-7
(synthetic surfactant) and surfactin (bio surfactant) were used for this study. Each surfactant concentration was at 10 mg/L prepared in a replica of three vials for estimating the precision.
Additionally, nutrient solution with no surfactant added mixed with CF was used as a control.
The CF concentration in all vials was 25 µg/L. The vials were capped with a 24 mm replacement mini-inert valve (Supelco, PA). The vials were then continuously mixed in a tumbler. The vials were maintained at a constant room temperature of 22±1 oC. A 1 mL gas sample was drawn from the head space of each vial using gas-tight syringes and analyzed for CF and degradation by-products concentrations on day 1, 3, 7, and 10. Fig. 6-2 displays CF concentration in the head space obtained from the control, and the two surfactants containing vials (tomadol 25 -7 and
149 | P a g e
surfactin). It is seen from Fig. 6-2 that the vials with surfactin showed better performance than that with tomadol 25 -7. From the figure, the controls showed no change of concentration with time. Additionally, it is seen from Fig. 6-3, vials with surfactin produced 30% more carbon dioxide than the ones with Tomadol 25 – 7. Hence, the addition of surfactin demonstrated the extent of the bioavailability of CF for biodegradation by the microbes.
6.4.2 BTF Performances
6.4.2.1 Performance of BTF-A
In this BTF, the co metabolite was mixed with CF and DCBM stream to enhance their biodegradation by providing additional electron donor to the micro-organisms. Ethanol was used as a co metabolite since it readily mixes with CF, DCBM and water. It is worth noting that the removal efficiency of ethanol was always above 98% for the given loading rate conditions studied for this BTF. Therefore, emphasis is placed on the performance of the BTF on CF and
DCBM. In phase I, the BTF started up with influent concentration of mixtures of THMs 5 ppmv and ethanol concentration of 25 ppmv providing a ratio of 1:5 (THMs:ethanol). The corresponding CF and DCBM loading rates were 0.14 and 0.18 g/(m3.h), respectively. The details of operation for BTF-A for every phase of operation with corresponding influent concentration, loading rate, days of operation, removal efficiency, and average elimination capacity are provided in Table 6-1. The BTF was run for 36 days during this Phase and the average removal efficiencies for CF and DCBM were 73% and 77%, respectively which provided an average elimination capacities (ECs) of 0.11 ±0.02 g/(m3.h) and 0.14 ± 0.03 g/(m3.h), respectively (Table 6-1). In phase II, the THMs - ethanol ratio was raised to 1:10 and
150 | P a g e
the removal efficiencies for CF and DCBM increased to 77% and 78% with corresponding ECs
0f 0.12 ±0.02 and 0.15 ±0.02 g/(m3.h), respectively. Phase II was run for 41 days. During phase III and IV ethanol:THMs ratio were 1:20 and 1:40 providing loading rates of ethanol 2.30 and 4.59 g/(m3.h), respectively. The corresponding removal efficiencies in these two phases for
CF were 79% and 85%, respectively, and that for DCBM were 85% and 87%. The presence of co metabolite supported the THMs to stay within the biofilm to achieve higher removal efficiency by providing the micro-organisms with additional electron donor. Overall, the result showed that the increase in the loading rates of the co metabolite enhanced the performances of the BTF. The same observation was visualized in our previous study [27]. Although, our previous study focused on single solute (CF), the result conformed that as the concentration of ethanol increased, the removal efficiency of CF also increased accordingly. The reported removal efficiency value of CF using co metabolic biofilter by Palanisamy et al. was 80.9% [39].
Similarly, Wahman et al. (2011) reported that the use of co metabolic biofilters seeded with three different mixed-culture sources can remove CF and DCBM up to 18% and 75%, respectively [40]. However, this current study demonstrated that higher removal efficiencies can be obtained for both CF and DCBM by using fungi BTF under acidic condition.
6.4.2.2 Performance of BTF-B
Table 6-2 describes the details of operation of BTF-B. BTF-B was run at different influent concentrations of mixtures of CF and DCBM without co metabolite. The two components were at equal ratios 1:1 with total concentration ranging from 5 to 15 ppmv with corresponding loading rates of 0.14 to 0.55 g/m3.hr. The table also summarized the operating conditions and different phases of operation. Fig. 6-4 represents a statistical summary of the
151 | P a g e
removal efficiency as a box plot at different loading rates of the BTF. The lower boundary of the box denotes the lower quartile, a line within the box marks the median, and the boundary of the box furthest from zero indicates the upper quartile. Whiskers (error bars) above and below the box indicate the 90th and 10th percentiles. In phase I, after the acclimation period, the BTF was run for 60 days and the average removal efficiencies obtained for CF and DCBM were 85% and
80% respectively which afforded average elimination capacities (ECs) of 0.13 ±0.01 g/(m3.h) and 0.16 ±0.02 g/(m3.h) respectively (Table 6-2).
On day 61(phase II), the mixtures concentration was raised to 10 ppmv (loading rates of
0.27 and 0.37 g/(m3.h)) where both removal efficiencies for CF and DCBM decreased to 75% and 74%. Their corresponding ECs were 0.20 ±0.02 and 0.27 ±0.02 g/(m3.h). In this phase, the
BTF left to run for 55 days. During phase III, after increasing the concentrations to 15 ppmv
(loading rates of 0.41 and 0.55 g/(m3.h)), the system was left to run for 46 days. The obtained removal efficiencies were 69% and 68% for CF and DCBM, respectively. As shown on (Table 6-
2), the result obtained in this phase provide less removal efficiency than the rest of the previous phases. This could be that THMs have toxicity effect on the microbes at higher concentration.
The result showed that as the concentration THMs increased, the removal efficiency for CF and
DCBM decreased.
6.5 Discussion of the results
6.5.1 Performance comparison for BTF–A and B
For similar concentration of THMs, BTF–B showed supremacy in elimination capacity and in performance stability as compared to BTF–A. The highest removal efficiency obtained
152 | P a g e
for BTF-A was at Phase IV when the ethanol loading reached to 4.59 g/ m3.hr. At this point, the removal efficiencies for CF and DCBM were 85% and 87%, respectively. However, BTF-B provided removal efficiency of 87% for CF and 80% DCBM at THMs concentration of 5 ppmv.
This behavior indicates clearly that the significance of the presence of surfactin in degrading the
THMs. As it can be seen from Table 6-2, the presence of surfactin enhanced the bioavailability of the THMs by facilitating their biotransformation. Furthermore, it can be concluded that BTF-
B could tolerate high loading rates of THMs and provide reasonable elimination capacities without adding a co metabolite. It should be noted that ethanol is often emitted together with
THMs from industries and wastewater sources [8, 41]. Balasubramanian et al. [8] reported that ethanol could be degraded fast enough compared other organic solvents. Therefore, this study demonstrated that BTF could be an effective technology for degrading THMs when emitted with ethanol or can be used with surfactin without a co metabolite if emissions do not contain ethanol.
In general, bio surfactants like surfactin have the advantage over synthetic surfactants in that they are generally more acceptable and are less toxic.
6.5.2 Kinetics CF and DCBM for the different phases
Gas phase samples were drawn from each port of both BTF-A and B once every week to evaluate the reaction rate kinetics of CF and DCBM corresponding to the total VOC loading rates. The samples were taken along the BTFs from ports that are located at 7.6 cm, 23 cm, 38 cm, 53 cm and 60 cm down from the top of the packed bed. The kinetic analysis was conducted using the data from sampling ports within the media as there is a possibility of biodegradation on the top portion of the BTF above the media or at the bottom disengagement chamber used for separation of liquid and gas effluents. The CF and DCBM concentrations in these samples along
153 | P a g e
with the influent stream concentration were used to develop the transformation kinetics as a pseudo first order reaction rate based on a plug flow reactor model. These data were used to develop the pseudo first order reaction rate constant as a function of time. All sampling data at every concentration level (minimum of 3 dataset) was fit with a linear model with the independent variable, time (seconds), and the dependent variable, loge(C/C0), where C is the effluent concentration and C0 is the influent concentration. Fig. 6-5 (a &b) shows the reaction rate constant for both BTF-A and B for similar THMs loading rate. The figure represents reaction rate constant for all the four phases of BTF-A and phase I of BTF–B. Fig. 6-5(a) represents CF reaction rate constant and Fig. 6-5(b) represents DCBM reaction rate constant for both BTFs.
The results shown in Fig 6-5 (a &b) indicate that higher reaction rate of 0.0020 s-1 for CF and 0.0022 s-1 for DCBM were obtained for Phase IV of BTF-A as compared to the other phases of the BTF. Similar results were obtained for BTF-B of phase I. The reported highest reaction rate constant for single CF was 0.0018 s-1 [27]. These results indicate that the presence of
DCBM did not inhibit the degradation of CF.
6.5.3 Carbon mass balance
The cumulative CO2 equivalent of CF, DCBM and the nutrients entering and leaving the
BTF-B is presented in Fig. 6-6. To study the carbon cycle within the bed for both the liquid and gaseous phases, all the carbon sources and products were measured. The influent consisted of gaseous concentrations of mixtures of THMs plus aqueous inorganic and organic carbon. The effluent included aqueous inorganic and organic carbon, a carbon equivalence of volatile
154 | P a g e
suspended solids, gaseous carbon dioxide, and a carbon equivalence of mixtures of THMs concentrations. The CO2 equivalence of all the carbon components was calculated in moles and a cumulative input and output CO2 equivalence of carbon was plotted on sequential time (Fig.6-6).
The carbon recovery for BTF-B was 55 % which is similar to BTF-A reported in the previous chapter. The recovery reported from our previous work for single solute for the similar four phases was 63 % [27]. It is postulated that the loss of influent and effluent carbon was produced as biomass within the BTF [29]. This hypothesis is justified by comparing the loss of carbon to the amount of biomass accumulated within the bed. The cellular composition for typical filamentous fungi is also presented by C9H15O5N [42]. These compositions were used as the basis for relating the nitrate consumed in building up new biomass to estimate the amount of biomass retained within the BTF. A t-test was performed to compare the results of the carbon consumed and the biomass produced. The obtained test results were with p- value < 0.05 indicating that the difference between the carbon retained and the biomass produced was statistically significant, therefore, confirming that the loss of carbon within the BTF was utilized for biomass growth.
6.5.4 Nitrogen utilization and COD reduction
Microorganisms’ uptake readily available inorganic nitrogen sources such as ammonia
+ − (NH4 ) and nitrate (NO3 ) which is very essential for their growth and development [43]. In this study, nitrates were the only source of nitrogen supplied with the nutrients. Daily analyses of
- influent and effluent concentrations of NO3 –N were performed. The net nitrogen utilization was
- computed from the NO3 –N in the BTF nutrients and the effluent liquid [44]. The chemical
155 | P a g e
oxidation demand (COD) for CF and DCBM degradation is illustrated using the following reactions.
2퐶퐻퐶푙3 + 2퐻2푂 + 푂2 → 2퐶푂2 + 6퐻퐶푙 ……………….…………………..…………………(1)
− 2퐶퐻퐶푙2퐵푟 + 2퐻2푂 + 푂2 → 2퐶푂2 + 4퐻퐶푙 + 퐵푟 ...……………………………..…………(2)
Eqn 1 was used to determine the mass of COD consumed to the mass of CF supplied. The ratio of gCOD/gVOC oxidized was 0.13. This value was used to determine the COD consumed from the influent and effluent CF in the BTF. Similarly, Eqn 2 represents the oxidation of DCBM to its end products, forming a gCOD/gVOC ratio of 0.10. This ratio was used to determine the
COD consumed from the influent and effluent DCBM.
The net COD was calculated as the difference between COD of the feed and the COD of the effluent gas and liquid streams [44]. Fig. 6-7 shows dimensionless CODremoval/Nutilization
(CODR/NU) ratios plotted against the total loading rates of mixtures of THMs in box plots. Phase
3 I produced a CODR/NU ratio of 14.75 for a total VOCs loading rate of 0.32 g/m .hr and reduced to 11.31 for a total loading rate of 0.64 g/m3.hr in phase II which corresponds to the performance of the BTF. During phase III, for the total loading rate of 0.96 g/m3.hr the corresponding
CODR/NU ratio reduced further to 11.04. The reported CODR/NU value for single CF was 14.5
[27]. The CODR/NU ratios for BTF-A ranges from 5.12 in phase I to 15.46 in phase IV. It is worth mentioning that the use of surfactin in the BTF system can result in providing the same ratio by providing a high yield VOC which contributed greatly to the high nitrogen consumption.
156 | P a g e
6.5.5 Microbial ecological analyses and correlation
Ion Torrent PGM system was used to study the microbial diversity structure of both
BTFs. Samples for the microbial analysis were collected from each BTF after re-acclimation to each phase when 99% of the original performance was attained. Microorganisms are the core component of biofiltration. Fig. 6-8 (a) and (b) provides the fungi community diversity observed in BTF-A and BTF–B for samples collected from each BTF from the top (port 2) port respectively.
BTF-A: On phase I, the BTF was running under acidic conditions and fed with 5 ppmv of THMs and 25 ppmv of ethanol, the most dominant species at port 2 was Fusarium sp., with relative abundancy of 98.4% (Fig. 6-8(a)). During phase II, when the BTF was fed with more ethanol
(50 ppmv), the amount of Fusarium sp showed a slight increase to 98.5%. On phase III, when the co metabolite increased, Fusarium sp. dominates by 98.7%. In phase IV, with higher ethanol concentration, the BTF was again dominated by Fusarium sp. 98.6%. In this phase, a slight percentage of F.oxysporum and F. Solani has been observed with small abundancy of
0.5% and 0.4%, respectively
BTF-B: In phase I, the BTF was run with 5 ppmv of equal mixtures of CF and DCBM (2.5 ppmv each) without co metabolite. The most dominant species were F. oxysporum and Fusarium sp., with relative abundancy of 78.06%, 21.07%, respectively (Fig. 6-8 (b)). During phase II, when the BTF was fed with increased mixtures at 10 ppmv, the amount of F. oxysporum increased to
84.24% and the amount for Fusarium sp reduced to 14.44% . In this phase, Fusarium sp reduced by 6.63% from the previous phase. In phase III, when the influent concentration of the THMs mixture was increased to 15 ppmv, F. oxysporum. dominated again by 66.87% followed by
Fusarium sp. of 30.49%. In this phase, the amount of Fusarium sp increased by almost 16%
157 | P a g e
compared to phase II. The amount of F. oxysporum was also at lowest abundancy than Fusarium sp. In addition, it was observed in Fig. 6-8 (B) that this phase has more Fusarium sp species than the other two previous phases. It is speculated that when the overall removal efficiency reduced, the dominant fungi (F. oxysporum) of BTF-B showed a significant reduction which benefited the growth of species Fousarium sp. In comparison with BTF-A, it is interesting to observe that the most dominant fungi in BTF-B bed was F. oxysporum while in BTF-A was Fusarium sp,.
Hence, it can be concluded that F. oxysporum was the dominant fungi for the degradation of CF and DCBM without co metabolite and Fousarium sp. was the benefited fungi with co metabolite.
6.6 Conclusion
Initially, batch studies for comparing the effectiveness of bio surfactant (surfactin) to synthetic surfactant (Tomadol 25 -7) indicated the superiority of surfactin over Tomadol 25 -7.
Biofiltration study was then conducted to investigate the effect of co metabolite and surfactin for the removal of CF and DCBM. The use of surfactin successfully enhanced the biodegradation process of CF and DCBM in the BTF as compared to the BTF with co metabolite. Furthermore, surfactin provided better performance. This biofiltration could provide more trust in the removal of THMs since treatment facilities prefer not installing more expensive options to ensure compliance with regulatory rules. The use of surfactin in the BTF system does not require the addition of co metabolite and thus reduces the cost of treatment. Hence, the developed BTF for the removal of THMs from flowing stream has a potential large-scale application. Microbial analysis indicated that Fusarium sp and F. oxysporum were the most dominant and abundant fungi responsible for the degradation of CF and DCBM.
158 | P a g e
Acknowledgments
The work conducted was partly supported by the contract number EP11C000147 obtained from the EPA Path Forward Innovation Project from the EPA-University of Cincinnati Grants
Program.
Disclaimer
The views expressed in this article are those of the authors and do not reflect the official policy or position of the Unites State Environmental Protection Agency. Mention of trade names, products, or services does not convey official EPA approval, endorsement, or recommendation.
This manuscript has been subjected to the Agency’s review and has been approved for publication.
159 | P a g e
Table 6-1. Operating conditions for BTF-A, it’s degrading 1:1 mixtures of CF and DCBM under aerobic and acidic conditions. Initial concentration of the mixtures was 5 ppmv.
Mixtures of THMs + Mixtures of THMs + Co metabolite Surfactin Operating condition Phases
I II III IV I
Influent co metabolite (Ethanol) concentration, 25 50 100 200 - ppmv
Operation time, days 36 41 33 38 60
THMs CF DCBM CF DCBM CF DCBM CF DCBM CF DCBM
Removal Efficiency, % 73±11 77±10 77±6 78±6 79±8 80±7 85±6 87±6 85±9 80±7
Loading Rate (g/m3. h) 0.14 0.18 0.14 0.18 0.14 0.18 0.14 0.18 0.14 0.18
Elimination Capacity 0.11±0.02 0.14±0.03 0.12±0.02 0.15±0.0 0.12±0.01 0.16±0.02 0.13±0.01 0.17±0.01 0.13±0.01 0.16±0.02 (g/m3. h)
160 | P a g e
Table 6- 2. Operating conditions for BTF-B, the BTF was degrading 1:1 mixtures of CF and DCBM under aerobic and acidic conditions in the presence of surfactin.
Phases Experimental Conditions I II III THMs Concentration, ppmv 5 10 15 Operation time, days 60 55 46 THMs CF DCBM CF DCBM CF DCBM Removal Efficiency, (%) 85±7 80±7 75±6 74±7 69±6 68±7 Loading Rate (g/m3. h) 0.14 0.18 0.27 0.37 0.41 0.55 Elimination Capacity (g/m3. h) 0.13±0.01 0.16±0.02 0.20±0.02 0.27±0.02 0.30±0.02 0.37±0.02
161 | P a g e
Figure 6-1 Schematic diagram of the biotrickling filters (BTFs)
162 | P a g e
Figure 6-2 CF concentration in headspace at different time in the presence of Tomadol 25 – 7 (synthetic surfactant) and surfactin (bio surfactant)
163 | P a g e
Figure 6-3 CO2 production from the head space analysis
164 | P a g e
Figure 6-4 Performance of the BTF in the four phases. Phase I: 5 ppmv of mixtures of THMs (CF + DCBM), phase II: 10 ppmv of mixtures of THMs , phase III: 15 ppmv of mixtures of THMs.
165 | P a g e
Figure 6-5 Comparison of reaction rate constants for mixtures of THMs (CF and DCBM) for BTF-A and B at similar loading rate. a)for CF, & b) for DCBM
166 | P a g e
Figure 6-6 Carbon mass balance: Cumulative carbon input and output as CO2 equivalent in mole for BTF-B.
167 | P a g e
Figure 6-7 Ratio of chemical oxygen demand (COD) to nitrogen utilization vs total loading rates of mixtures of THMs. Phase I: 5 ppmv of mixture of THMs, Phase II: 10 ppmv of mixture of THMs, and Phase III: 15 ppm
168 | P a g e
Figure 6-8 a) Fungi community diversity for all the four phases of the BTF-A for samples collected at the top port (port 2) of the BTF. b) Fungi community diversity for all the four phases of the BTF-B for samples collected at the top port (port 2) of the BTF
169 | P a g e
6.7 References 1. Kim, J., et al., Chlorination by-products in surface water treatment process. Desalination, 2003. 151(1): p. 1-9. 2. Agus, E., N. Voutchkov, and D.L. Sedlak, Disinfection by-products and their potential impact on the quality of water produced by desalination systems: a literature review. Desalination, 2009. 237(1-3): p. 214-237. 3. Lichtfouse, E., Environmental chemistry: green chemistry and pollutants in ecosystems. 2005, Berlin, Germany: Springer Science & Business Media. 4. USEPA. Ground Water and Drinking Water: Table of Regulated Drinking Water Contaminants,(Accessed 09/10/2016) : https://www.epa.gov/ground-water-and-drinking- water/table-regulated-drinking-water-contaminants. 2016 [cited 2016 09/10/2016]; Available from: https://www.epa.gov/ground-water-and-drinking-water/table-regulated- drinking-water-contaminants. 5. Watts, P., G. Long, and M. Meek, Concise international chemical assessment document 58: chloroform. World Health Organization (IPCS), Geneva, 2004. 6. Mulligan, C.N., Environmental applications for biosurfactants. Environ. pollut., 2005. 133(2): p. 183-198. 7. Nakanishi, T. and H. Oku, Metabolism and accumulation of pentachloronitrobenzene by phytopathogenic fungi in relation to selective toxicity. Phytopathol., 1969. 59(11): p. 1761-1762. 8. Balasubramanian, P., L. Philip, and S.M. Bhallamudi, Biotrickling filtration of complex pharmaceutical VOC emissions along with chloroform. Biores. Technol., 2012. 114: p. 149-159. 9. Zilli, M., et al., Phenol removal from waste gases with a biological filter by Pseudomonas putida. Biotechnol. Bioeng., 1993. 41(7): p. 693-699. 10. Staudinger, J. and P.V. Roberts, A critical compilation of Henry's law constant temperature dependence relations for organic compounds in dilute aqueous solutions. Chemosphere, 2001. 44(4): p. 561-576. 11. McGregor, F., P. Piscaer, and E. Aieta, Economics of treating waste gases from an air stripping tower using photochemically generated ozone. J. International Ozone Assoc., 1988. 12. LaKind, J.S., S.D. Richardson, and B.C. Blount, The Good, the Bad, and the Volatile: Can We Have Both Healthy Pools and Healthy People? Environ. Sci. Technol., 2010. 44(9): p. 3205-3210. 13. Delhoménie, M.C., L. Bibeau, and M. Heitz, A Study of the Biofiltration of High‐ Loads of Toluene in Air: Carbon and Water Balances, Temperature Changes and Nitrogen Effect. The Canadian J. Chem. Eng., 2005. 83(2): p. 153-160. 14. Yoon, I.-K., C.-N. Kim, and C.-H. Park, Optimum operating conditions for the removal of volatile organic compounds in a compost-packed biofilter. Korean J. Chem. Eng., 2002. 19(6): p. 954-959. 15. Wahman, D.G., L.E. Katz, and G.E. Speitel, Trihalomethane cometabolism by a mixed- culture nitrifying biofilter. J. Amer. Water Works Assoc., 2006. 98(12): p. 48-60. 16. Wahman, D.G., et al., Ammonia-oxidizing bacteria in biofilters removing trihalomethanes are related to Nitrosomonas oligotropha. Appl. Environ. Microbiol., 2011.
170 | P a g e
17. Yeh, D.H., K.D. Pennell, and S.G. Pavlostathis, Effect of Tween surfactants on methanogenesis and microbial reductive dechlorination of hexachlorobenzene. Environ. Toxicol. Chem., 1999. 18(7): p. 1408-1416. 18. Laha, S. and R.G. Luthy, Effects of nonionic surfactants on the solubilization and mineralization of phenanthrene in soil–water systems. Biotechnol. Bioeng., 1992. 40(11): p. 1367-1380. 19. Aronstein, B.N. and M. Alexander, Surfactants at low concentrations stimulate biodegradation of sorbed hydrocarbons in samples of aquifer sands and soil slurries. Environ. Toxicol. Chem., 1992. 11(9): p. 1227-1233. 20. Hassan, A.A. and G. Sorial, Biological treatment of benzene in a controlled trickle bed air biofilter. Chemosphere, 2009. 75(10): p. 1315-1321. 21. Woertz, J., K. Kinney, and P. Szaniszlo, A fungal vapor-phase bioreactor for the removal of nitric oxide from waste gas streams. J. Air & Waste Manage. Assoc., 2001. 51(6): p. 895-902. 22. Volkering, F., A. Breure, and W. Rulkens, Microbiological aspects of surfactant use for biological soil remediation. Biodegradation, 1997. 8(6): p. 401-417. 23. Yuan, S., et al., Effect of nonionic and cationic surfactants on the dechlorination kinetics and products distribution of various polychlorinated benzenes by Cu/Fe particles. Sep. Purif. Technol., 2010. 74(1): p. 130-137. 24. Van Groenestijn, J., W. Van Heiningen, and N. Kraakman, Biofilters based on the action of fungi. Water Sci. Technol., 2001. 44(9): p. 227-232. 25. Kennes, C. and M.C. Veiga, Fungal biocatalysts in the biofiltration of VOC-polluted air. J. Biotechnol., 2004. 113(1): p. 305-319. 26. Hassan, A.A. and G.A. Sorial, A comparative study for destruction of n-hexane in trickle bed air biofilters. Chem. Eng. J., 2010. 162(1): p. 227-233. 27. Palanisamy, K., et al., Biofiltration of Chloroform in a Trickle Bed Air Biofilter Under Acidic Conditions. Water, Air, & Soil Pollut., 2016. 227(12): p. 478. 28. Palanisamy, K., et al., Biofiltration of Chloroform in a Trickle Bed Air Biofilter Under Acidic Conditions. Water, Air, & Soil Pollution, 2016. 227(12): p. 478. 29. Hassan, A.A. and G.A. Sorial, Biofiltration of n‐ hexane in the presence of benzene vapors. J. Chem. Technol. and Biotechnol., 2010. 85(3): p. 371-377. 30. Zehraoui, A., et al., Impact of alternate use of methanol on n-hexane biofiltration and microbial community structure diversity. Biochem. Eng. J., 2014. 85: p. 110-118. 31. APHA, Federation, Water Environmental American Public Health Association, Standard methods for the examination of water and wastewater. American Public Health Assoc. (APHA): Washington, DC, USA, 2005. 32. Zhai, J., et al., Microbial Community in a Biofilter for Removal of Low Load Nitrobenzene Waste Gas. PloS one, 2017. 12(1): p. e0170417. 33. Dowd, S.E., et al., Bacterial tag-encoded FLX amplicon pyrosequencing (bTEFAP) for microbiome studies: bacterial diversity in the ileum of newly weaned Salmonella-infected pigs. Foodborne Pathog. and Dis., 2008. 5(4): p. 459-472. 34. Edgar, R.C., Search and clustering orders of magnitude faster than BLAST. Bioinform., 2010. 26(19): p. 2460-2461. 35. Capone, K.A., et al., Diversity of the human skin microbiome early in life. J. Investigative Dermatolo., 2011. 131(10): p. 2026-2032.
171 | P a g e
36. Eren, A.M., et al., Exploring the diversity of Gardnerella vaginalis in the genitourinary tract microbiota of monogamous couples through subtle nucleotide variation. PloS one, 2011. 6(10): p. e26732. 37. Swanson, K.S., et al., Phylogenetic and gene-centric metagenomics of the canine intestinal microbiome reveals similarities with humans and mice. The ISME J., 2011. 5(4): p. 639-649. 38. DeSantis, T.Z., et al., Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl. Environ. Microbiol., 2006. 72(7): p. 5069-5072. 39. Palanisamy, K., et al., Biofiltration of Chloroform in a Trickle Bed Air Biofilter Under Acidic Conditions. Water, Air, Soil Pollut., 2016. 227(12): p. 478. 40. Wahman, D.G., L.E. Katz, and G.E. Speitel, Performance and biofilm activity of nitrifying biofilters removing trihalomethanes. Water Res., 2011. 45(4): p. 1669-1680. 41. Cecen, F. and O. Aktas, Activated carbon for water and wastewater treatment: Integration of adsorption and biological treatment. 2011: John Wiley & Sons. 42. Bruce, E.R. and L.M. Perry, Environmental biotechnology: principles and applications. New York: McGrawHill. Vol. 400. 2001, New York: McGraw-Hill 43. Moe, W.M., et al., Removal of the sesquiterpene β-caryophyllene from air via biofiltration: performance assessment and microbial community structure. Biodegradation, 2013. 24(5): p. 685-698. 44. Zehraoui, A., A.A. Hassan, and G.A. Sorial, Effect of methanol on the biofiltration of n- hexane. Journal of hazardous materials, 2012. 219: p. 176-182.
172 | P a g e
7. Conclusions and recommendations
7.1 Summary
The application of biofiltration for controlling VOCs emissions has become a great interest. The biofiltration technology is considered the best available control technology and cost effective. However, this technology faces several challenges when used for controlling chlorinated compounds. Typically, hydrophobic compounds like THMs are not readily available for the microorganisms, which could be challenging for utilizing biological treatment. The solubility of VOCs in water, which depends on Henry’s law constant, is the most important characteristics that affect the performance of biofilters. As a solution to the handling limitation in biofilter performance, this study investigated several techniques to successfully biodegrade
THMs. Adding co metabolite (an easily degradable compound) in the feed gas has the potential to enhance the biodegradation of these chlorinated compounds. Furthermore, the bio-availability of hydrophobic compounds can be enhanced by the introduction of nontoxic surfactants
(synthetic or bio – surfactant) that will serve dual processes, increasing solubility and limiting excess biomass growth.
7.2 Conclusions
7.2.1 Effects of aeration for THMs control
The result from the bubble aeration indicated that aeration significantly decreased the THMs and proved to best controlling technique. The gas flow exhibited greatest effect on THMs removal from water. As the airflow increased from 0.5 to 2 L/min , the removal of THMs increased from 81% to 99% for DCBM, 59% to 97% for DBCM and 42% to 88% for BF. The
173 | P a g e
tank temperature gradient study revealed that as the temperature increased the formation of
THMs also increased. The quantitative risk analysis after aeration of chloroform revealed that the values for potential non- cacogenic/ cacogenic are within the USEPA’s acceptable range. Thus, aeration significantly reduces chloroform concentration to safe drinking water levels.
7.2.2 Biological study a. Performance of the Anaerobic BTF
The anaerobic BTF’s performance for single solute (CF) was evaluated for three phases, namely, in the presence of co-metabolite (Phase I), in the presence of co-metabolite and surfactant (Phase II) and the presence of surfactant only (Phase III). Enhanced removal efficiency was obtained when surfactant was added to the system together with ethanol (Phase
II), which provided a removal efficiency of 64% with EC of 0.17 g/ (m3.hr) for a loading rate of
0.27g./(m3.hr). Consequently, higher reaction rate constants were attained in this phase as compared to the other two phases. A mixture of co metabolite and surfactant benefits the growth of A. oryzae and Geobacter spp. The increase in the percentage of A. orazae also correlates with the removal efficiency of CF. The overall results obtained for the anaerobic BTF showed clearly that using both co metabolite and surfactant had effectively enhanced the biodegradation of CF by providing more favorable conditions for the growth of bacteria colonies.
174 | P a g e
b. Anaerobic vs aerobic
Comparison was conducted on the performances anaerobic and aerobic BTFs for single solute (CF). As a result, the use of aerobic fungi BTF under acidic condition successfully enhanced the biodegradation process of CF. Operation at acidic pH enhanced greatly the performance providing a removal efficiency around the 80.9% level. Using fungi culture led to higher loading rates that could not be achieved by anaerobic microbial culture. The result obtained from microbial analysis showed that the most dominant fungi, which promote higher removal efficiency, were Fusarium sp. and F. solani. A. oryzae and A. restrica were the responsible bacteria community species responsible for anaerobic BTF. The comparative study proves the effectiveness of the use of aerobic BTF under acidic condition which supports fungi growth.
c. Mixtures of THMs with co metabolite
The mixtures study investigated the effect of fungi based BTF for the removal of CF and
DCBM mixtures under acidic condition. CF and DCBM were studied at equal volume ratios with each having a concentration of 2.5 ppmv. Ethanol was used as a co metabolite in enhancing the degradation of mixtures. The ratios of THMs mixture to ethanol were 1:5, 1:10, 1:20 and 1:40.
The maximum elimination capacity for CF was 0.13 g / m3.hr and for DCBM was 0.17 g / m3.hr for the given influent loading rates of 0.14 and 0.18 g / m3.hr. This maximum elimination capacity resulted when the ratio of THMs to ethanol was 1:40. The microbial ecology analysis result showed that fusarium sp was the most dominant fungi responsible for the degradation of
CF and DCBM.
175 | P a g e
d. Mixtures of THMs with surfactin
A comparative study was investigated for evaluating the effect of co metabolite and surfactin
(bio surfactant) on BTF performance for removal of mixtures of THMs (CF and DCBM). The use of surfactin successfully enhanced the biodegradation process of CF and DCBM in the BTF and provided more stable performance by having smaller standard deviation in the removal efficiency as compared to the BTF with co metabolite.
The added stability in performance could put more trust in the biological treatment of THMs as treatment facilities not prefer installing more expensive options to ensure compliance with regulatory rules. The use of surfactin in the BTF system can prevent the addition of co metabolite and avoid its cost. Hence, the developed BTF for the removal of THMs from flowing stream has a potential for large-scale application. The result from the microbial study showed that Fusarium sp and F. oxysporum were the most dominant fungi responsible for the degradation of CF and DCBM.
Finally, the results of this study proved that BTF is an effective technique to treat gas phase
CF and DCBM through co metabolism and surfactin under acidic condition. It is an effective technique to treat gas phase THMs that emit from aeration of storage tanks in the distribution system for drinking water and other industries.
7.3 Recommendations of future work
(1) Extended studies with multicomponent mixture of hydrophilic and hydrophobic VOCs are still needed as biofilters deployed in air emissions facilities are mostly exposed to
176 | P a g e multicomponent contaminants. More application of the BTF in degrading other chlorinated compounds is needed.
(2) Extensive investigation of the microbial community in the biofilters needed which will uncover how spatial and temporal variability of the biofilter ecology is VOCs dependent.
(3) The use of bio surfactant like surfactin need to be explored more for different chlorinated compounds other than THMs.
(4) Biofilters operating in industry are generally exposed to other compounds or multicomponent contaminants with varying transient loadings. As a solution to this problem one needs to evaluate the integrated two bed adsorption system and BTF under the transient operating conditions for the removal of hydrophobic compounds.
(5) Many studies have been conducted on the effectiveness of Zero Valent Iron (ZVI) for degradation of chlorinated hydrocarbons. Hence, a study is needed to investigate the effectiveness of ZVI along with the BTF in the removal of volatile DBPs.
(6) The empty bed residence time (EBRT) for this study was 5 minutes. Further studies need to be conducted to evaluate the impact of EBRT on the performance of the BTF.
177 | P a g e