au · 2013. 1. 9. · “cebona” (cost effective biofilter for odor nuisance abatement for...
TRANSCRIPT
Function and structure of biofilters removing sulfur gasses
PhD Thesis
by Claus Lunde PedersenDepartment of BioscienceAarhus UniversityAU
Cover page: Cryo scanning microscopy image of fungal aerial hyphae taken from a biofilter treating waste air from a pig farm.
Function and structure of biofilters removing sulfur gasses
Department of BioscienceAarhus University
PhD Thesis
Claus Lunde Pedersen
1
Preface: This PhD thesis represents my work as a PhD student under supervision of Lars Peter
Nielsen from September 2009 till December 2012. During my project I have been
investigating the biological aspects of biofiltration of odorous sulfur compounds from
farming industry, in collaboration with partners from Aarhus University, Aalborg
University, Danish Pig Production, Skov A/S and Weber Saint-Gobain A/S. My work
was funded by The Danish Council for Strategic Research, as part of the project:
“CEBONA” (Cost Effective Biofilter for Odor nuisance Abatement for intensive pig
production).
The aim of the CEBONA project was to gain knowledge and improve existing
biofilter technology, where my PhD project focused on the uptake and metabolism of
odorous sulfur compounds in biofilters from pig farms. Pig farms emit odors at low
concentrations at high ventilation rates and it seemed contradictive that some
biofilters still managed to reduce emissions of odorous hydrogen sulfide (H2S). These
preliminary observations initiated the PhD project, where biofilms from several
biofilters were investigated for their surface structures, which could explain the
unexpected removal of H2S. This was done by combining online mass spectrometry
with different types of microscopy, investigating the surface structures of the biofilms,
which lead to a preliminary hypothesis (Chapter 2). The hypothesis was strengthened
and elaborated by repeated studies at the same biofilter (Chapter 3). Final studies of
the biofilter were aimed to identify the H2S oxidizing microorganisms with the use of
molecular techniques (Chapter 4). During the project, I was also involved in the setup
and testing of another full-scale biofilter treating waste air from a pig farm. The
material used in the biofilter contained high amounts of iron oxides, which are known
to catalyze H2S oxidation. The aim was to show, whether or not the iron oxides in the
operating biofilter were still catalyzing H2S (Chapter 5).
This PhD has been a journey into studies related to physical and physiological aspects
of biofiltration, which has gone far beyond the initial expectations. It has been
exciting and rewarding to be part of an ongoing investigation to reveal new insights in
this topic of applied microbiology.
Claus Lunde Pedersen, December 2012
2
!
3
"#$%&!'(!)'*+&*+,!
List of supporting papers! "!
#$%%&'(! )!
*+,$%-! .!
/0,1!23!&44'+50&1026,! 7!
861'29$:1026! ;<!!"#$% &'!"#$%!&'%('&)*$+! ,,!
"#$%!#')'()*$+! ,-!
"#$%!.()*/*)0! ,1!
()#*)+,$-,)#.% &/!2&).3'!$4!($5&$6+#7!4%$5!.*%! ,8!
9*:)'%!#'7*;+! ,8!
9*:)'%!5.)'%*.:! ,<!
0123-.)454%*#$%678%9:,-;1% &<!96+;.:!76:4*#'!$=*#.)*$+! -,!
>*)$(?$+#%*.:!@-A!$=*#.)*$+! -1!
B%$+!(.).:0C'#!@-A!$=*#.)*$+! -8!
=#.2+94)#.4%-."%:1$4:12,)>14% 7?!
=+6+'&>!90,:$,,026!23!%&6$,:'0?1,! @<!0-.942$):,%@% A'!0-.942$):,%@@% A'!0-.942$):,%@@@% A&!0-.942$):,%@B% A7!
Manuscript I 39
Manuscript II 56
Manuscript III 67
Manuscript IV 81
Acknowledgements 96
Appendix: Conference activity and Curriculum Vitae
4
List of supporting papers
I. Pedersen, C. L.; Hansen, M. J.; Jensen, L. H. S.; Guldberg, L. B.; Feilberg, A.;
Nielsen, L. P. "Correlation between fungi and removal of hydrogen sulfide
in air treating biofilter" (in preparation)
II. Pedersen, C. L.; Liu, D.; Feilberg, A.; Nielsen, L. P. "Fungi metabolize
hydrogen sulfide in an air treating biofilter" (in preparation)
III. Pedersen, C. L.; Schramm, A.; Nielsen, L. P. "Bacterial and eukaryotic
community in a biofilter treating hydrogen sulfide" (analysis in progress)
IV. Pedersen, C. L.; Liu, D.; Feilberg, A.; Nielsen, L. P. "Non-biological
oxidation of hydrogen sulfide and methane thiol in an air treating
biofilter containing iron oxides" (in preparation)
Supporting papers will be referred to by their roman numerals e.g. (Manuscript II).
!
5
-.//#01!Odors from farming industry have a negative impact on human living conditions for
people present in rural areas. The odors comprise of a complex matrix of organic
compounds produced by bacteria in the manure from farmed animals. Hydrogen
sulfide has been identified as the most potent of odors, when comparing people’s
sensitivity to an individual compound, with emission concentrations equal to the ones
from pig stables. The human nose is still considered to be the best way of detecting
odor compounds, however with the arrival of proton transfer reaction mass
spectrometry (PTR-MS), it is now possible to measure concentrations accurately close
to the human detection limits of sulfur compounds, including hydrogen sulfide. The
PTR-MS was installed at two pilot filters, which were observed to remove hydrogen
sulfide at rates that were too high to be explained by conventional biofiltration, where
water solubility and diffusion limit the bacterial biofilm´s capability to take up and
metabolize hydrogen sulfide. Fungal aerial hyphae were present at short intervals in
one of the filters and a strong correlation was found between removal of hydrogen
sulfide and presence of fungal aerial hyphae, however there had been no previous
reports on fungi metabolizing hydrogen sulfide. Nevertheless eukaryotic organisms
have been found to be using hydrogen sulfide as a signal molecule, and based on other
studies we have proposed a pathway, in which hydrogen sulfide oxidation may be
energetically favorable for fungal aerial hyphae in biofilters. This pathway explains
the high removal rate of hydrogen sulfide by fungi. Samples of filter material were
taken, when fungi were present and absent to identify the organisms that oxide
hydrogen sulfide. Phylogenetic analysis of 16S and 18S ribosomal DNA clone
libraries showed no strict sulfide oxidizing bacteria, and that most of the fungi present
in the biofilter were related to food items in the stable, such as fodder, or to the skin of
farm animals. One fungal strain was found to be degrading cellulose, which the filter
material was composed of, suggesting that fungal aerial hyphae may be supported by
the filter itself, given the right nutrients.
The second filter, with high removal of hydrogen sulfide had expanded clay
aggregates as filter material. This porous material contained high amounts of iron
oxides, which are known to catalyze hydrogen sulfide oxidation to elemental sulfur.
The material without bacteria showed a 50% removal of H2S compared to material,
which came from the biofilter. Pasteurizing material from the biofilter confirmed that
6
half of the removed hydrogen sulfide is due to the iron oxides present in the filter
material. No elemental sulfur was observed in the biofilter, indicating that bacteria
oxidize the iron-catalyzed products.
In conclusion iron oxides and fungi proved to be two novel ways of removing
hydrogen sulfide from air and may in future contribute to the design of more efficient
biofilters, provided that presence of fungal aerial hyphae can be maintained in the
biofilter. Iron oxides seem more easily applicable: they could be either incorporated
into the biofilter material, as with the clay pellets, or applied on surfaces of the filter
material.
7
2&,./3!Lugtgener fra landbruget har en negativ indvirkning på helbredet for befolkningen i
landdistrikterne. Lugtene består af en kompleks matrix af organiske forbindelser, der
er produktet af bakteriel aktivitet i gyllen fra husdyravl. Svovlbrinte er blevet
identificeret som den mest potente lugt, når emissionskoncentrationer fra svinestalde
sammenlignes med folks følsomhed over for de enkelte lugtstoffer. Den menneskelige
næse anses stadig for at være den bedste måde at detektere lugtstoffer, men med
introduktionen af proton transfer reaction mass spectrometry (PTR-MS), er det nu
muligt at måle lugt koncentrationer tæt på den menneskelige detektionsgrænse for
svovlforbindelser, såsom svovlbrinte. PTR-MS målte svovlbrinte koncentrationer i to
pilot filtre , men fjernelsesgraden for svovlbrinte var for høj til at kunne forklares ved
konventionel forståelse af biofiltrering, hvor vandopløselighed og diffusion begrænser
bakteriel omsætning af svovlbrinte. Svampehyfer blev observeret i et af filtrene med
korte mellemrum og en korrelation blev påvist mellem fjernelse af svovlbrinte og
tilstedeværelsen af svampehyfer. Der har dog ikke været tidligere indikationer af
svampe der kan omsætte svovlbrinte. Eukaryote organismer har vist sig at benytte
svovlbrinte som et signalmolekyle og baseret på andre studier foreslår vi en
reaktionsvej, hvor svovlbrinte omsætning kan være energetisk favorabel for svampene
i biofiltret. Svampene kan derved forklare den høje omsætning af svovlbrinte. Prøver
af filtermateriale blev taget med og uden svampes tilstedeværelse i filteret for at
identificere de organismer der omsætter svovlbrinte. Fylogenetisk analyser af 16S og
18S ribosomale DNA klon-biblioteker viste ingen svovlbrinte oksiderende bakterier
og viste, at de fleste af de svampe, der findes i biofilteret var relateret til fødevarer i
stalden, såsom foder, eller de boede på huden af husdyr. En af de fundne svampe
nedbryder cellulose, som filtermaterialet var sammensat af, hvilket antyder, at
svampehyfer kan eksistere i filteret, hvis de rigtige næringsstoffer er til stede. Det
andet filter med høj fjernelse af svovlbrinte anvendte brændte ler partikler (LECA)
som filtermateriale. Dette porøse materiale indeholdt store mængder af jernoxider,
som er kendt for at katalysere svovlbrinte oksidation til elementært svovl.
Filtermateriale uden bakterier viste en 50% fjernelse i forhold til materialet, som var
blevet podet med bakterier i det biologiske filter, hvilket blev bekræftet ved
pasteurisering af materiale fra det eksisterende filter. Ingen elementært svovl blev
8
observeret i det biologiske filter, hvilket indikerer at bakterier oxiderede det
elementære svovl til sulfat.
Anvendelsen af jernoxider og svampe er to hidtil ukendte måder at fjerne
hydrogensulfid i luften fra landbruget og kan i fremtiden bidrage til udformningen af
mere effektive biofiltre. Dog forudsat at tilstedeværelsen af svampehyfer kan
opretholdes i de biologiske filtre. Jernoxider synes mere relevant, og kan anvendes
enten på overflader af filtermateriale eller inkorporeres i det biologiske filter materiale
som det var tilfældet i ler partiklerne.
9
45,+!'(!#$$0&65#+5'*,! Ammonia (NH3)
Dimethyl disulfide (DMDS)
Dimethyl sulfide (DMS)
Gas chromatography mass spectrometry (GC/MS)
Hydrogen sulfide (H2S)
Light expanded clay aggregates (LECA)
Membrane inlet mass spectrometry (MIMS)
Methane thiol (MT)
Odor active value (OAV)
Odor threshold value (OTV)
Proton transfer reaction mass spectrometry (PTR-MS)
Protonated water molecule H3O+
Solid phase microextraction (SPME)
10
7*+0'8.9+5'*!
:8'0!!
Odors from intensive pig farming are common problems in rural areas. Besides the
nuisance associated with odors, they have also been shown to have a negative impact
on the public health of people living in close proximity of the pig farms, where
several studies have suggested odors from intensive pig farms to be the cause for
several ailments (Schiffman et al. 1995; Merchant and Ross 2002; Wing et al. 2008).
Over the past 20 years, pig production has been increasing in Denmark the production
has been intensified to fewer units (Council 2012). This development has led to a
demand for odor control from intensive pig farms, since nuisance associated with
odors are expected to be increasing too. The odors emitted from pig farms are mainly
coming from fermentation of slurry (Riis and Lyngbye 2005). The slurry is rich in
organic compounds, reflecting the nutrients in ingested food (Le et al. 2007). Inside
the pig, gut microbiota inoculates the food, where bacteria ease the uptake of nutrients
for the pig by cleaving larger molecules into smaller molecules by fermentation under
anoxic conditions. Some of the fermentation products are highly volatile and released
as odorous flatulence. The fermentation continues after the stool leaves the pig, since
the slurry remains anoxic with the exception of a few micrometers on the slurry
surface, where oxygen penetrates (Markfoged et al. 2011). Despite that fermentation
is substrate and product limited, the process still continues since some of the
fermentation products have low water solubility and proceed to the air phase,
lowering the product concentration in the slurry. Since the fermentation products are
reduced organic compounds formed under anoxic conditions, they can react with
oxidized compounds as soon as they are emitted to the air phase, or represent an
energy source for bacteria by utilizing them as electron donors. The emitted
compounds from the slurry reflect the fermentation substrate e.g. an increase in
protein content in the pig diet yields higher ammonia emissions, due to higher
nitrogen availability (Le et al. 2007). These emitted odor compounds are typically
categorized by their chemical composition, e.g. ketones, esters and aromatic
compounds, where several hundred compounds has been identified from pig
production facilities (Schiffman et al. 2001). The fermentation of sulfur containing
11
amino acids lead to the formation of reduced sulfur compounds, such as hydrogen
sulfide (H2S), methane thiol (MT), dimethyl sulfide (DMS) and dimethyl disulfide
(DMDS). Along with ammonia (NH3), these are some of the compounds found in the
exhaust air from a pig farm (Feilberg et al. 2010). It is not only the concentration that
determines odors from pig farms, but also the perception, where most people are able
to distinguish fermentation products at very low concentrations.
:8'0!;&09&;+5'*!
The perception of odor from bacterial fermentation is useful and in some cases
critical, since fermentation products indicate the presence of microbial activity and
alert the person, if a potential food source has been spoiled by microbial degradation.
Whether an odorant is considered pleasant or nauseating depends on the concentration
(Lim et al. 2001). As an example some of the compounds emitted from pig farms are
also found in fermented food products such as beer or dairy products, where they are
perceived pleasantly, since they appear in much lower concentrations. Diluting a
particular odor can be used to determine the lowest detectable concentration, termed
odor threshold value (OTV) (van Gemert 2003). Comparing OTV between different
odor compounds may describe which compounds we are most sensitive to. The sense
of smell as with other senses is perceived on a logarithmic scale, meaning that when
the odor concentration increase tenfold, then humans perceive it as a doubling in
intensity. This makes it difficult to detect the exact concentration of a particular
compound by olfactometry. Furthermore the perception of a specific odor diminishes
over time as long as the person is exposed. This phenomenon is characterized as "odor
fatigue" and the sensitivity to smell is restored when the person is no longer exposed
to the particular odor. It follows, that the first impression of an odor is related to a
specific odor concentration, while changes of odor concentrations are perceived over
time due to odor fatigue. The sense of smell may determine the lowest possible
concentration, which can be detected, but it is not suitable to quantify higher
concentrations of odors. Therefore other techniques are needed in order to determine
odor compounds concentration in air.
12
:8'0!8&+&9+5'*!
Given the complexity of odor compounds emitted from pig farms makes them
difficult to identify and quantify. Several techniques have been introduced to
determine the concentration of odor compounds in air, where the most common
technique for identifying and quantifying odors is gas chromatography followed by
mass spectrometry (GC/MS). Such equipment is restricted to the laboratory and
sampling techniques are required for the collection of odors. By collecting air in
canisters and transporting it to the laboratory, odors may be detected, however given
the physical properties of odor, the compounds may react or adsorb with the material
used for sampling the odors (Hansen et al. 2011). The adsorbing properties are
exploited in another sampling technique, where odors are collected on adsorbing
material, such as Tenax TA (2,6-diphenylene oxide polymer) in small metal tubes
(Trabue et al. 2008) or with solid phase microextraction (SPME) (Cai et al. 2006). A
known volume of air is sampled on the adsorbent tube and transported to the laboratory,
where the tubes are heated and the odors desorb following detection by GC/MS. Several
studies have investigated the use of different materials for collecting air samples, or for
odor adsorption for method optimization, since both sampling methods have problems
with recovery of odors (Trabue et al. 2008). Humidity in the sampled air, reaction with
sampling material or leakage out through the material, are some of the factors that have a
negative influence on the recovery of odors in the laboratory. However, it has been shown,
that the use of Tedlar bags in combination with the analysis of air for sulfur compounds
within 24 hours minimize the loss of compounds (Hansen et al. 2011). Recently two
methods for odor measurements have been introduced to online measure odors from pig
farms by membrane inlet mass spectrometry (MIMS) (Feilberg et al. 2010) and proton
transfer reaction mass spectrometry (PTR-MS) (Feilberg et al. 2010). Both instruments
are based on the use of mass spectrometry to detect compounds mass to charge ratio.
MIMS lets the incoming air pass over a permeable membrane, where compounds in the
air diffuse through the membrane and compounds are then ionized and detected by the
mass spectrometer (Johnson et al. 2000). The membrane may select for specific
compounds due to their solubility to the specific membrane material used. PTR-MS
contains a reaction chamber or a drift tube, where compounds (R) passes through and
reacts with protonated water (H3O+), that act as ion source (Warneke et al. 1996):
R+H3O+ ! RH + +H2O
13
The ionized compounds are then detected by mass spectrometry as in MIMS.
Ionization takes place, if that the compounds proton affinity is higher than waters,
which is the case for most organic compounds. However for compounds with proton
affinity close to that of water, the reaction above may be reversed. This is the case for
H2S, which is found in the exhaust air from pig farms. Feilberg et al. found a
correlation between humidity and H2S concentration, which made it possible to
correct for humidity dependence and measure H2S concentrations down to 5 ppb.
Humidity may also affect detection limits for MIMS. Both PTR-MS and MIMS have to
compare measurements to a compound reference library, e.g. by GC/MS, in order to
assign detected mass spectra with their compound. Both instruments need recalibration
for measurements in the field, due to specific circumstances, such as humidity and
temperature (Hansen 2011). For field measurements and in combination with air sampling
in Tedlar bags, PTR-MS is the preferred technique for quantifying odorous sulfur
compounds from pig farms (Feilberg et al. 2011).
14
:8'0!#9+565+1!
By comparing the measured concentrations of odor compounds emitted with their
specific OTV, it is now possible to compare the impact of different compounds.
Dividing the measured odor concentration by their specific OTV yields their odor
active value (OAV). Despite that the physiology behind odor perception is complex
and still not fully understood, OAV is considered a simple way to describe the impact
of a particular compound. By comparing OAV from different compounds it is
possible to distinguish which of the compounds that contributes most to the odor
sensation.
Figure 1: Bar plot of Odor Active Values calculated for different compounds from the exhaust air of pig farms. Both studies indicate, that H2S and MT have the highest odor active values.
In two studies by Feilberg et al. 2010 and Hansen et al 2012, the odors in exhaust air
from pig farms was measured using PTR-MS and the OAV of the identified
compounds was calculated. The results showed, that H2S followed by MT were the
largest contributors of odor sensation in the air (see figure 1). By introducing the
PTR-MS as a new technique to measure odor emissions from pig farms, it is now
possible to identify key odor components. Next step is to design measures to reduce
odor emissions with emphasis on the removal of H2S and MT.
D!
-D!
1D!
8D!
ED!
,DD!
,-D!
,1D!
,8D!
,ED!
-DD!
A92'!B:105+!C&>$+!
D2%?2$69!
9'*:F'%;!')!.:!-D,D! @.+7'+!')!.:!-D,-!
15
<5'(5%+0#+5'*!
Air treating biofilters have been introduced in farming industry as an approach to
reduce odor emissions (Oneill and Phillips 1992). As mentioned previously, the odors
represent an energy source for bacteria in an oxic environment, which is also the case
for sulfur compounds such as H2S and MT. This can be demonstrated by calculating
the potential energy from oxidation of H2S, which yields -796 KJ mol-1. MT has a
similar energy yield and given that MT also contain a methyl group MT also represent
a carbon source for the bacteria:
H2S + 2O2 !H2SO4 "G0 ' = #796KJmol#1( )
Figure 2: Schematic representation of concentration of a compound in the liquid boundary layer and biofilm. The concentration of the particular compound in the air determines the concentration in the water phase (liquid boundary layer) by the compounds water solubility. Concentration drops by diffusion in boundary layer, while diffusion and consumption is applied in the biofilm. Modified from Devinny and Ramesh (Devinny and Ramesh 2005).
!"#$%&'(
)"*+",$-'+.,/#0$&/01#2"'3"&4
%"&51#$4/51#"/&
61758
9'.:1.5#/5"'.
16
=;+#>&!'(!9'/;'.*8,!(0'/!#50!
In order for bacteria to utilize compounds from the air phase of a biofilter, the
compounds must first be dissolved in the water phase and diffuse into the biofilm (see
figure 2). The concentration of a compound in the water phase is determined by its
water solubility or Henry's law constant (Kh), which describes the relation between a
compounds concentration in water (c) and the concentration in the air phase (p)
(Sander 1999):
The flux (J) of a compound into the biofilm layer is determined by its concentration
difference (!C) over a distance (!x) and the compounds diffusion coefficient (Ds) as
described by Fick's first law of diffusion:
From the two equations the flux is defined by the compounds concentration in the air,
its solubility, its diffusion coefficient and thickness of the biofilm. It can be assumed
that the compounds concentration is zero in the proximity of bacteria, since the
compounds will be oxidized as soon as they are available for them. Solubility and
diffusion coefficients are physical parameters that cannot be altered, but air
concentration and diffusion distance, !x, can be taken in account in order to optimize
the biofiltration process.
?5%+&0!8&,5@*!
Constructing a biofilter is to provide a suitable environment for the bacteria to thrive,
and actively take up odors from the air. This is done by providing a large surface,
where the bacteria can attach to, and ensuring that the filter is kept humid.
Humidification may be provided by irrigation of the bacterial surface or by
humidification of the air coming into the filter. Under optimal conditions a biofilm
will start to develop on the surfaces, provided that bacteria are introduced into the
filter. Biofiltration can be categorized into three main types of filters: (1) Biofilter,
where the incoming air is humidified before reaching the surface of the filter, (2)
Trickling biofilter, where the filter is irrigated with water to be kept humid, (3) Bio-
scrubber, where bacterial activity is restricted to the water humidifying the incoming
Kh =cp
J = !Ds!C!x
17
air and filter area is kept to a minimum, if any filter is included (Devinny et al. 1999).
Several studies have investigated the removal of NH3 and odors from pig farms by the
use of biofilters (Sheridan et al. 2002) or trickling biofilters (Melse et al. 2012).
However there are few studies on H2S or MT removal by biofiltration, which could be
ascribed to the low air concentrations. Furthermore the efficiency of conventional
biofilters is difficult to maintain and there is a need for new types of biofilters (Riis et
al. 2008). Some biofilters combine different types of biofiltration for better removal of
different compounds from the air. One example is SKOV filter, which is sectioned
into three separate filters (Figure 3).
Figure 3: Schematic representation of the 3-step biofilter by SKOV A/S. Filter 1 and 2 have constant irrigation, while filter 3 is humidified by the water saturated air. Measuring points with PTR-MS in the third section and sampling points for 16S and 18S RNA gene analysis; before filter at 0 cm (square), 20 cm (diamond), 40 cm (triangle) and after filter at 60 cm (circle).
The SKOV filter is divided into three sections, which have the aim to take up and
metabolize different compounds in each section. The first section acts as a dust trap,
where NH3, dust particles and soluble organic compounds are trapped in the water
that is constantly irrigated over the filter surface. The second section acts as a
trickling biofilter, where irrigation is restricted to ensure better uptake and removal of
less soluble organic compounds by minimizing the liquid boundary layer on the
surface of the biofilm. The last section has no irrigation, and acts as a biofilter with
the aim of removing odorous sulfur compounds. Previous studies investigated the
SKOV filter for uptake and turnover of NH3 (Juhler et al. 2009; Ottosen et al. 2011)
!"#
$"%&'#() $"%&'#(* $"%&'#(+
,(-. *,(-. /,(-. 0,(-.
18
and organic compounds (Feilberg et al. 2010; Kristiansen et al. 2011) in the first two
sections. Product inhibition of ammonia oxidizing bacteria was minimized by
controlling water supply and drainage according to conductivity of the water irrigated
over the two sections (Ottosen et al. 2011). It was assumed that the restricted variety
of compounds, that reach the third section of the filter, would lead to a selection, in
the third section, for bacteria specialized in metabolizing sulfur compounds. The
PTR-MS was installed to continuously measure concentrations between each of the
filter sections over several weeks, with an average removal efficiency of H2S was 9
%, 15% and 68 % for the three sections respectively, see table 1 (Hansen et al. 2012).
Removal efficiency of 68% in the third section was extraordinarily high compared to
other sulfur compounds, such as MT and DMS that were not taken up in the last
section. Despite differences in air concentration and solubility, physical properties
alone could not explain the high uptake of H2S, suggesting that there were something
in the third section, that was specifically related with the uptake of H2S in the third
section.
Table 1: Concentrations of selected compounds measured with PTR-MS before and after each section of the SKOV filter. From (Hansen et al. 2012).
Concentration (ppb) Compound Before
section 1 After section 1 After section 2 After section 3
Hydrogen sulfide 353 322 272 86 Acetaldehyde 6.8 4.4 9.4 0.5 Methane thiol 12 12 10 10 Acetone 7.1 4.8 5.4 0.8 Trimethylamine 12 4.3 3.2 0.8 Acetic acid 314 49 54 1.2 Dimethyl sulfide 3.0 2.9 2.7 2.7 2-butanone 3.3 2.0 2.4 0.6 Propanoic acid 67 16 19 0.5 2,3-butanedione 1.2 0.7 1.1 0.2 Butanoic acid 42 11 9.4 0.3 Phenol + dimethyl disulfide 2.0 0.7 1.9 0.2 C
5 carboxylic acids 11 3.2 2.7 0.2
p-cresol 9.5 2.5 2.6 0.2 Indole 0.7 0.2 0.1 - 4-ethylphenol 1.2 0.4 0.5 - Dimethyl trisulfide 0.1 0.06 0.1 - 3-methylindole 0.4 0.1 0.07 -
Few studies have measured the removal efficiency of H2S online at filters operated
under similar conditions. For comparison, Gabriel et al. reported outlet concentrations
of H2S that were similar to inlet concentrations of the SKOV filter at half the
ventilation rate (Gabriel and Deshusses 2003). Both filters emphasize the difficulties
19
of removing all H2S from the air phase. The two filters differ in design and in filter
material, which also influence the efficiency and running costs of the filter.
?5%+&0!/#+&05#%!
Choosing the right filter material is a major challenge in the development of biofilters,
when keeping material and running costs to a minimum. There are three main
parameters influencing the choice of filter material: the price of material, the lifetime
of the material and also the structure of the material. Typically used are inexpensive
materials such as wood chips, straws or peat (Iranpour et al. 2005). The low cost in
acquisition is counter balanced by the materials being degraded over time, which
gives a relative short lifetime. They also have a dense structure, which increases when
the material collapses during degradation. The ventilation rate in pig farms is
normally regulated in order to maintain a constant indoor temperature. With
increasing temperature the ventilation rates increases, which leads to higher energy
expenditure due to the air resistance in the biofilter, caused by filter material or the
clogging by dust and biomass (Andreasen et al. 2012). Some materials may be harder
to degrade, such as oyster shells, lava rock or clay pellets. These materials still have a
high air resistance and may be little more expensive, but less degradable. Polymer
structures (Gabriel and Deshusses 2003; Melse et al. 2012) or cellulose pads
introduced in SKOV filters have lower air resistance compared to the previously
described filter materials. These materials may be more expensive and for the
cellulose pads will be degraded with time.
A&9B#*5,/,!('0!CD-!.;+#>&!
The previously mentioned results from the SKOV filter lead to an investigation of the
last section, where the high removal of H2S was observed (Manuscript I). The setup of
PTR-MS was changed to measure air concentrations at 20 cm intervals in the last
section, where a core with 10 cm in diameter was taken out and Teflon tubes installed
inside the filter section and connected to the PTR-MS (Figure 4 and 5).
20
Figure 4: Setting up PTR-MS for measuring odor compounds inside the third section of SKOV filter. Taking out a core from the front of filter (left). Tubes and filters for air samples were inserted in the filter core (insert). Teflon tubes connected to PTR-MS protruding out from the backside of the filter (right).
PTR-MS measurements were recorded continuously over two days. The odor
concentration in the air varied according to the ventilation rate in the pig stable
(Figure 5). The variation in concentration was due to variation in ventilation rates,
which was dependent on the stable temperature. The ventilation rate in the stable
acted as dilutant for the constant production of odor from the slurry, with highest
ventilation rates during afternoons.
21
Figure 5: Measured H2S concentrations over two days at four different depths of the third section of SKOV filter. Highest concentrations were measured during nighttime, when the ventilation rate was lowest.
Removal efficiency of H2S in the third section varied between 43% and 79% with an
average of 66%. Based on the PTR-MS measurements and the dimensions of the
biofilter, the flux of H2S into the biofilm could be calculated. Assuming uniform
distribution and consumption of H2S in the biofilm and no diffusive boundary layer,
then following equation was developed to calculate H2S concentration for each depth
(Cx) of the biofilm (Nielsen et al. 1990):
Where x is the depth into the biofilm, J0 is the flux of H2S into the biofilm, while C0 is
the concentration on the biofilm surface and D is diffusion coefficient of H2S in
water. H2S concentration reached zero at 2.9 µm in the first 20 cm, and 2.7 µm in the
following 20 cm interval of the third section. Assuming that bacterial cell size to be
approximately 1 - 2 µm in diameter would restrict H2S metabolism to one or two cell
layers. Considering the assumptions made in the calculation, it was not possible to
explain the uptake of H2S. There had to be another explanation and the surface of the
biofilm was investigated with different imaging techniques.
?.*@#%!,.%(58&!'E58#+5'*!
Photographs were taken from the sectioned core of the third filter and fungal aerial
hyphae were observed (Figure 6). Image analysis showed that fungi had the largest
cover on the front of the filter, and combining measurements with cryo scanning
electron microscopy fungal density was calculated. It was then possible to correlate
0
50
100
150
200
250
300
350
12:00 18:00 0:00 6:00 12:00 18:00 0:00 6:00 12:00
Con
cent
ratio
n, p
pbv
Time
0 cm 20 cm 40 cm 60 cm
Cx =J02
4DC0x2 ! J0
Dx +C0
22
fungal density and removal of H2S, which suggested that fungi could be responsible
for the removal of H2S (Figure 7). The 50 % of H2S uptake took place in the first 20
cm of the filter, which also had the highest density of fungi (Figure 6). It has
previously been suggested that fungi can oxidize elemental sulfur to thiosulfate
(Grayston and Wainwright 1988), but no previous study have reported fungi
metabolizing H2S.
Figure 6: Photo of cross section of the third section of SKOV filter with inlet side on the left. Fungal aerial hyphae were observed, with highest density on the inlet side of the filter. Close-up of the hyphael structure by cryo scanning electron microscopy from the third section of SKOV filter (insert).
Fungi have previously been used in biofiltration (Kennes and Veiga 2004; Estevez et
al. 2005), where they were found to metabolize anthropogenic toxins. Fungi have a
diverse metabolic capacity, but are generally considered heterotrophs so it seemed
unlikely that they could growth on H2S alone. It was also speculated that fungi could
be detoxifying H2S, since H2S is known to inhibit cytochrome c oxidase in the
respiratory chain (Cooper and Brown 2008). However the measured H2S
concentrations in the filter were far from inhibitory concentrations. To support the
23
correlation between fungi and H2S removal further measurements were conducted on
site at periods with and without fungal aerial hyphae (Manuscript II).
Figure 7: Removal efficiency of H2S relative to inlet concentration (black line) compared to fungal density (open bars). Values were plotted against filter depth.
Measurements of H2S concentration before and after the third filter were compiled
with observations of fungi from 2010 till 2012 (Figure 8).
Figure 8: Removal efficiency of H2S plotted against ventilation rate for third filter section. Measurements with fungi present in 2010 (closed squares), measurements with fungi present in 2011 (open triangles), measurements with fungi present in 2012 (open squares) and measurements without fungi present in 2011 (closed diamonds).
There was a significant difference (P < 0.001) between removal efficiency of H2S
with or without fungal aerial hyphae present in the third filter, which seemed to
confirm that fungi were the reason for the high removal of H2S. However it was still
D!
DGDH!
DG,!
DG,H!
DG-!
DG-H!
DI!
-DI!
1DI!
8DI!
EDI!
,DDI!
,-DI!
D! ,DD! -DD! JDD! 1DD! HDD! 8DD!
K%$L'()'#!46+;.:!.%'.!&'%!F*$M*:5!.%'.!
N65-!65O-P!
@-A!.)!*+:'#!($+('+)%.)*$+Q!I!
9*:)'%!#'&)?!R55S!
K%$L'()'#!46+;.:!.%'.!
@-A!%'5$/.:!
DI!
-DI!
1DI!
8DI!
EDI!
,DDI!
D! ,DDDD! -DDDD! JDDDD! 1DDDD!
TUQ!I!
V'+)*:.)*$+Q!>J!?$6%O,!
W!96+;*!-D,-! W!96+;*!-D,,! O!96+;*!-D,,! W!96+;*!-D,D!
24
unclear, how H2S was being oxidized and it was speculated whether fungi was
metabolizing H2S or perhaps facilitating it to symbiotic bacteria. Regarding the uptake
of H2S, Matai et al concluded, that there were no need for specific adaptations for
taking up H2S over cell membranes, since H2S could pass over the cell membrane
freely (Mathai et al. 2009). This was further supported by Cuevasanta et al., which
showed that cell membranes did not limit diffusion of H2S, since H2S solubility in
lipids were not markedly different, than H2S solubility in water (Cuevasanta et al.
2012). Both studies focused on H2S due to its importance as a signal molecule in
eukaryote cells (Szabo 2007; Olson 2012). Further research into the therapeutic
potential of H2S lead to the proposal of a general enzymatic pathway for H2S
oxidation in mitochondria (Hildebrandt and Grieshaber 2008). Therefore fungal H2S
oxidation could no longer be ignored, since it was establish that fungal cell walls and
membranes were no obstacle for H2S, and that fungal mitochondria could oxidize
H2S. However it was still unclear, whether fungi would gain energy from the
oxidation of H2S. The pathway proposed by Hildebrandt and Grieshaber suggested
ATP production from H2S oxidation and this had also been stated previously by
Goubern et al. (Goubern et al. 2007). However excretion of metabolites may also
demand energy, and an open question still remains, whether the fungi can sustain its
metabolic demand in the filter by H2S oxidation alone.
A5+'9B'*805#%!CD-!'E58#+5'*!
Fungi could explain the removal of H2S in SKOV filter; however H2S was also
removed at periods without fungal aerial hyphae. Could this removal be ascribed to
fungal or bacterial activity? The biofilm had a thickness of approximately 60 µm,
when no fungal aerial hyphae were present (Figure 9). The biofilm depth could be
compared with H2S penetration depth calculated from PTR-MS at times were no
fungal hyphae were observed. Penetration depth was calculated as described above, to
be 115 µm, indicating that H2S oxidation could take place through the whole biofilm.
As seen in figure 9 a dense biofilm was attached to the filter material and blue stains
from DAPI indicated presence of eukaryotes, such as fungi.
25
Figure 9: Fluorescence in situ hybridization (FISH) image of biofilm in profile from the third section in absence of fungal aerial hyphae. Sample was hybridized with general bacterial probe (red) and stained with DAPI (blue). Auto fluorescence from the cellulose filter material was colored green.
This lead to a further investigation of the third section of SKOV filter with the aim of
describing the bacterial and eukaryotic community in the third section of SKOV filter
and to identify the fungal strain that were taking up H2S (Manuscript III). Samples of
filter material were collected from the third section in presence and absence of fungal
aerial hyphae. The samples were collected from four depths of the filter (Figure 3),
since fungi were observed to be unevenly distributed (Manuscript I). The comparison
of eukaryotic and bacterial diversity between periods with and without fungal
presence showed no significant difference. There were no differences between
samples at different depths at the same time of sampling. Thus changes in H2S
removal could not be ascribed to changes in bacterial diversity and no strict sulfide
oxidizing bacteria were found in any of the samples. The fungal strain
Tremellomycetes was identified in the third section of the filter. This strain had
previously been identified as degrader of 4-ethylphenol and acetic acid in sections one
26
two in a SKOV filter, however they were not associated with H2S oxidation
(Kristiansen 2010). Being a yeast type, Tremellomycetes could not be identified as the
producer of the aerial hyphae. Several strains of basidiomycota were also identified in
the third section, which could form aerial hyphae. May be it was not important which
fungi were oxidizing H2S as long as fungal aerial hyphae could be maintained in the
filter, when considering mitochondrial H2S oxidation as a general trait. In essence,
H2S oxidation is taking place when H2S reacts with ferric iron in the heme group of
cytochrome c oxidase (Cooper and Brown 2008), suggesting that ferric iron could
catalyze H2S oxidation regardless of what form ferric iron is present in the filter.
70'*!9#+#%1F&8!CD-!'E58#+5'*!!
Oxidation of H2S with metal ions, such as iron, has been used in oil industry to
remove H2S from natural gas (de Angelis 2012). This process oxidize H2S to
elemental sulfur:
H2S + 2Fe3+ ! S0 + 2Fe
2+ + 2H +
This reaction is pH dependent due to the formation of protons, and it has also been
investigated in marine environments (Chen and Morris 1972). Ferric ions has also
been introduced as catalyst in biofilters treating waste air from waster water treatment
plants (Goncalves and Govind 2008). Assuming that oxygen is present the ferrous ion
could oxidize to ferric ion and H2S oxidation could continue. The question still
remained, whether iron catalyzed H2S oxidation was still taking place in a running
biofilter. To address this questing filter material containing iron oxides was sampled
from a trickling biofilter that had been running for more than a year (Manuscript IV).
27
Figure 10: Photo of two trickling biofilters investigated at a farm. Insert: close-up of light expanded clay aggregates (LECA) sampled from one of the trickling biofilters.
The filter material used was light expanded clay aggregates (LECA), which contain
high amounts iron oxides (Tabase et al. Unpublished). Previous studies have
investigated clay pellets, which sometimes are referred to as lava rock, however they
did not relate H2S oxidation with iron oxides (Li et al. 2005; Akdeniz et al. 2011). The
iron oxides are noticed by its distinct red color (Figure 10). A laboratory experiment
was set up, using a small scale setup, where H2S, MT and DMS concentrations and air
speeds were equivalent to concentrations measured by PTR-MS at the two pilot plants
(Liu et al. Unpublished). With new LECA material, with no biofilm present, the
small-scale setup showed a removal H2S by 46 % and 14% of MT. When using the
same setup and same working parameters, the filter material was substituted with
material taken from the pilot filter in operation, the removal efficiency increased to 78
% for H2S and 35 % for MT. The increased removal efficiency could only be ascribed
to the presence of an active biofilm, which was confirmed by pasteurizing the material
and repeating the experiment. Pasteurized material had the same or lower removal
efficiency as observed for new material without biofilm. There was some removal of
DMS from the LECA material, however with removal efficiency ranging between -2
28
% and 8 %. The negative removal indicates formation of DMS, which was ascribed to
conversion of MT to DMS (Liu et al. Unpublished; Tabase et al. Unpublished).
Release of DMS instead of MT would be preferable, when considering that human
odor threshold is 84 times lower for MT than DMS (van Gemert 2003). The increased
H2S oxidation in biofilter material could be ascribed to bacteria directly oxidizing
H2S, or to an indirect effect by oxidizing elemental sulfur. Since no elemental sulfur
was observed in the running biofilter it could be expected that bacteria were oxidizing
elemental sulfur to more soluble products, such as sulfate.
)'*9%.,5'*,!#*8!;&0,;&9+56&,!!
This study has shown how the presence of iron in biofilters can improve removal of
H2S from the ventilation air. Iron was present in the form of ferric ions in heme-
containing enzymes in fungi or iron oxides in LECA material. In future both fungi
and iron oxides could be incorporated in filter material in biofilters treating H2S. Iron
oxides have previously been used in biofiltration, but there have been no previous
records of their use in biofiltration in farming industry, or testing its effect in an
operating biofilter. There are several considerations for the use of iron oxides in future
designs of biofiltration. The thickness of the biofilm should be kept to a minimum to
promote the contact between H2S from the air and iron oxides in the filter material,
however the biofilm should be dense enough to further oxidize sulfur products to
sulfate. It could be speculated that bacterial H2S oxidation would compete with iron
catalyzed sulfide oxidation as it was observed in the material from the operating
biofilter. The structure of the filter material should also be considered, where LECA
material showed a high pressure drop due to the densely packed clay pellets, other
filter materials with iron oxides incorporated may be more preferable.
Provided that fungi is supported by a carbon source it could be possible in the future
to use and control fungi as an alternative to bacterial oxidation of H2S. It could be
assumed that fungal H2S oxidation is a general feature for fungi, however it should be
tested by growing different strains of fungi on nutrient plates and exposing them to
different H2S concentrations. Comparing results with controls, which have not been
exposed to H2S it could then be confirmed or denied whether H2S oxidation is an
energetically cost or benefit for the fungi by measuring the removal rate of H2S and
comparing to fungal growth rate under nutrient limitations. As mentioned in
Manuscript II, it could be expected that mitochondrial H2S oxidation has an upper
29
limit due to H2S inhibition of cytochrome c oxidase (Cooper and Brown 2008) which
limits the use fungal biofilters to treat H2S at high concentrations. The inhibition
concentration could be determined by measuring fungal removal rate of H2S, while
increasing H2S concentration could determine the inhibition concentration.
30
G&*&0#%!85,9.,,5'*!'(!/#*.,905;+,!The experiments conducted during this PhD have resulted in four manuscripts that are
in preparation for publishing. In the following section the results and conclusions are
summarized and discussed.
A#*.,905;+!7!
Few studies have investigated the efficiency of biofiltration of odorous sulfur
compounds from pig farms. Despite its low water solubility, H2S was removed in a
biofilter (Hansen et al. 2012), which lead to a further investigation of the filter in
order to discover the mechanisms involved in the uptake and removal of H2S. Image
analysis of the surface structure of the biofilm was combined with online
measurements of H2S removal in the biofilter and a correlation between fungal
density and H2S removal was found. This correlation suggested that fungi were taking
up H2S, despite no previous observations of fungi metabolizing H2S. Although it was
not clear how or why fungi were removing H2S from the air, but lead to the
hypothesis that fungi could oxidize H2S. The correlation between fungal aerial hyphae
and H2S removal lead to the hypothesis that fungi could oxidize H2S. Fungal strains
have previously been found to degrade odorous compounds without assimilating the
degraded compounds (Kennes and Lema 1994). The correlation between fungal aerial
hyphae and H2S removal could not explain how or why H2S was removed. H2S was
not considered toxic at the low concentrations observed in the filter (Reiffenstein et al.
1992).
A#*.,905;+!77!!
The correlation found in Manuscript I between fungal aerial hyphae and removal of
H2S in a biofilter led to the hypothesis that fungi could oxidize H2S. Further
measurements of H2S removal were recorded in the presence and absence of fungal
aerial hyphae. A comparison between filter efficiency in the presence or absence of
fungi showed an increase in H2S removal by a factor of four, when fungi were
observed in the biofilter. This result supported the hypothesis, despite that oxidation
of H2S had not been studied in fungi before. Due to its toxicity, H2S has been studied
in other eukaryotic organisms, where it was found to reversibly inhibit mitochondrial
31
cytochrome c oxidase (Cooper and Brown 2008). An enzymatic pathway has been
proposed in the mitochondria to oxidize H2S, which could also explain how fungi
oxidize H2S, provided that H2S enter the fungal cell (Hildebrandt and Grieshaber
2008). Matai et al. showed that no facilitator was required for H2S to cross the cell
membrane, and Cuevasanta et al. showed that cell membranes were no limitation for
diffusion of H2S (Mathai et al. 2009; Cuevasanta et al. 2012). Based on these studies
it seemed likely that fungal aerial hyphae could explain the increased H2S oxidation
rates measured with PTR-MS in the biofilter. Fungi have previously been used in
biofiltration and this new discovery could lead to the use of fungi in biofilters
removing H2S, provided that the aerial hyphael structure could be maintained (Kennes
and Veiga 2004).
A#*.,905;+!777!!
The aim of this study was to describe bacterial and eukaryotic community in a
biofilter, where fungal aerial hyphae had been observed. There were no significant
differences in bacterial or eukaryotic diversity, when comparing samples taken in the
presence and absence of fungal aerial hyphae, suggesting that changes in bacterial
diversity could not account for the increased H2S oxidation rates observed in the
presence of fungal aerial hyphae. The bacterial community was significantly different
in the third section, when comparing it to the bacterial communities described in
sections one and two in similar biofilters, suggesting a strong selection in each section
by the incoming compounds (Kristiansen et al. 2011; Kristiansen et al. 2011). This
finding correlates well with PTR-MS measurements showing difference in uptake for
each section (Hansen et al. 2012). 40 % of the identified sequences from the
eukaryotic clone library were tremellomycetes, which had previously been identified
in a similar biofilter to metabolize acetic acid and p-cresol (Kristiansen 2010).
However tremellomycetes are not known to form aerial hyphae. Trichocomaceae
were found in five of the eight samples and despite not being found in all the samples
seemed more likely to be oxidizing H2S due to its aerial hyphael structure. It seemed
more relevant which structure the fungi could have, to facilitate contact between H2S
and mitochondria, than which fungi were oxidizing H2S. Especially if considering
fungal H2S oxidation in the mitochondria to be a general feature.
32
A#*.,905;+!7H!
Clay pellets have been introduced as filter material in trickling biofilters. Clay pellets
contain iron oxides, such as hematite, which have been found to catalyze oxidation of
H2S and MT (Tabase et al. Unpublished). The aim of this study was to identify
whether iron oxides still contributed to the oxidation of H2S and MT in a trickling
biofilter that had been operating for more than a year. Clay pellets that had not been
previously used in biofiltration showed removal of both H2S and MT as expected.
Clay pellets, which had been used for biofiltration, showed higher removal rates of
H2S and MT. The increased removal could be a result of bacterial activity, which was
tested by pasteurizing the clay pellets. Pasteurized clay pellets showed similar results
to clay pellets that had not been used in biofiltration. These results showed that iron
oxides were still contributing to the oxidation of H2S and MT, despite the presence of
biofilm on the clay pellets. Catalysis of H2S with iron oxides would lead to formation
of elemental sulfur, which could clog the biofilter. Nevertheless elemental sulfur was
not observed in the trickling biofilter, which suggested that bacterial activity were
oxidizing the catalyzed products to sulfate. In future the use of iron oxides in
biofilters treating waste air from pig farms could contribute to the reduction of odor
emissions, since H2S and MT has been identified as the key odor compounds emitted
from pig farms. Iron oxides could either be introduced in running biofilters or
incorporated in the design of new biofilters treating H2S and MT.
33
2&(&0&*9&,!
Akdeniz, N., K. A. Janni, et al. (2011). "Biofilter performance of pine nuggets and
lava rock as media." Bioresource Technology 102(8): 4974-4980.
Andreasen, R. R., R. E. Nicolai, et al. (2012). "Pressure drop in biofilters as related to
dust and biomass accumulation." Journal of Chemical Technology and Biotechnology
87(6): 806-816.
Cai, L. S., J. A. Koziel, et al. (2006). "Characterization of volatile organic compounds
and odorants associated with swine barn particulate matter using solid-phase
microextraction and gas chromatography-mass spectrometry-olfactometry." Journal of
Chromatography A 1102(1-2): 60-72.
Chen, K. Y. and J. C. Morris (1972). "Kinetics of Oxidation of Aqueous Sulfide by
O2." Environmental Science & Technology 6(6): 529-&.
Cooper, C. E. and G. C. Brown (2008). "The inhibition of mitochondrial cytochrome
oxidase by the gases carbon monoxide, nitric oxide, hydrogen cyanide and hydrogen
sulfide: chemical mechanism and physiological significance." J Bioenerg Biomembr
40(5): 533-539.
Council, D. A. F. (2012). "Statistics 2011 Pigmeat."
Cuevasanta, E., A. Denicola, et al. (2012). "Solubility and Permeation of Hydrogen
Sulfide in Lipid Membranes." PLoS One 7(4).
de Angelis, A. (2012). "Natural gas removal of hydrogen sulphide and mercaptans."
Applied Catalysis B-Environmental 113: 37-42.
Devinny, J. S., M. A. Deshusses, et al. (1999). Biofiltration for air pollution control.
Boca Raton, Fla., Lewis Publishers.
Devinny, J. S. and J. Ramesh (2005). "A phenomenological review of biofilter
models." Chemical Engineering Journal 113(2-3): 187-196.
34
Estevez, E., M. C. Veiga, et al. (2005). "Biofiltration of waste gases with the fungi
Exophiala oligosperma and Paecilomyces variotii." Appl Microbiol Biotechnol 67(4):
563-568.
Feilberg, A., A. P. S. Adamsen, et al. (2010). "Evaluation of Biological Air Filters for
Livestock Ventilation Air by Membrane Inlet Mass Spectrometry." J. Environ. Qual.
39(3): 1085-1096.
Feilberg, A., D. Liu, et al. (2010). "Odorant Emissions from Intensive Pig Production
Measured by Online Proton-Transfer-Reaction Mass Spectrometry." Environmental
Science & Technology 44(15): 5894-5900.
Feilberg, A., T. Nyord, et al. (2011). "Chemical Evaluation of Odor Reduction by Soil
Injection of Animal Manure." Journal of Environmental Quality 40(5): 1674-1682.
Gabriel, D. and M. A. Deshusses (2003). "Performance of a full-scale biotrickling
filter treating H2S at a gas contact time of 1.6 to 2.2 seconds." Environmental
Progress 22(2): 111-118.
Gabriel, D. and M. A. Deshusses (2003). "Retrofitting existing chemical scrubbers to
biotrickling filters for H2S emission control." Proc Natl Acad Sci U S A 100(11):
6308-6312.
Goncalves, J. J. and R. Govind (2008). "H2S Abatement in a biotrickling filter using
iron(III) foam media." Chemosphere 73(9): 1478-1483.
Goubern, M., M. Andriamihaja, et al. (2007). "Sulfide, the first inorganic substrate for
human cells." FASEB J 21(8): 1699-1706.
Grayston, S. J. and M. Wainwright (1988). "Sulfur oxidation by soil fungi including
some species of mycorrhizae and wood-rotting basidiomycetes." Fems Microbiology
Ecology 53(1): 1-8.
Hansen, M. J. (2011). Significance of reduced sulphur compunds in relation to odour
from pig production, Dept. of Engineering, Aarhus University.
35
Hansen, M. J., A. P. S. Adamsen, et al. (2011). "Stability of Odorants from Pig
Production in Sampling Bags for Olfactometry." Journal of Environmental Quality
40(4): 1096-1102.
Hansen, M. J., D. Liu, et al. (2012). "Application of Proton-Transfer-Reaction Mass
Spectrometry to the Assessment of Odorant Removal in a Biological Air Cleaner for
Pig Production." Journal of Agricultural and Food Chemistry 60(10): 2599-2606.
Hildebrandt, T. M. and M. K. Grieshaber (2008). "Three enzymatic activities catalyze
the oxidation of sulfide to thiosulfate in mammalian and invertebrate mitochondria."
FEBS J 275(13): 3352-3361.
Iranpour, R., H. H. J. Coxa, et al. (2005). "Literature review of air pollution control
biofilters and biotrickling filters for odor and volatile organic compound removal."
Environmental Progress 24(3): 254-267.
Johnson, R. C., R. G. Cooks, et al. (2000). "Membrane introduction mass
spectrometry: Trends and applications." Mass Spectrometry Reviews 19(1): 1-37.
Juhler, S., N. P. Revsbech, et al. (2009). "Distribution and Rate of Microbial
Processes in an Ammonia-Loaded Air Filter Biofilm." Applied and Environmental
Microbiology 75(11): 3705-3713.
Kennes, C. and J. M. Lema (1994). "Simultaneous biodegradation of p-cresol and
phenol by the basidiomycete Phanerochaete chrysosporium." J Ind Microbiol 13(5):
311-314.
Kennes, C. and M. C. Veiga (2004). "Fungal biocatalysts in the biofiltration of VOC-
polluted air." Journal of Biotechnology 113(1-3): 305-319.
Kristiansen, A. (2010). Composition and metabolic significance of microorganisms in
biological airfilters. Section of Biotechnology, Chemical and Environmental
Engineering. Aalborg, Aalborg University. PhD.
Kristiansen, A., S. Lindholst, et al. (2011). "Butyric Acid- and Dimethyl Disulfide-
Assimilating Microorganisms in a Biofilter Treating Air Emissions from a Livestock
Facility." Applied and Environmental Microbiology 77(24): 8595-8604.
36
Kristiansen, A., K. H. Pedersen, et al. (2011). "Bacterial community structure of a
full-scale biofilter treating pig house exhaust air." Systematic and Applied
Microbiology 34(5): 344-352.
Le, P. D., A. J. Aarnink, et al. (2007). "Effects of dietary crude protein level on odour
from pig manure." Animal 1(5): 734-744.
Li, H., D. R. Luelking, et al. (2005). "Biogeochemical analysis of hydrogen sulfide
removal by a lava-rock packed biofilter." Water Environment Research 77(2): 179-
186.
Lim, T. T., A. J. Heber, et al. (2001). "Characteristics and emission rates of odor from
commercial swine nurseries." Transactions of the Asae 44(5): 1275-1282.
Liu, D., A. Feilberg, et al. (Unpublished). "Performance evaluation of two air
biofilters in treating odor pollutants from intensive pig production."
Markfoged, R., L. P. Nielsen, et al. (2011). "Transient N2O accumulation and
emission caused by O-2 depletion in soil after liquid manure injection." European
Journal of Soil Science 62(4): 541-550.
Mathai, J. C., A. Missner, et al. (2009). "No facilitator required for membrane
transport of hydrogen sulfide." Proc Natl Acad Sci U S A 106(39): 16633-16638.
Melse, R. W., J. P. M. Ploegaert, et al. (2012). "Biotrickling filter for the treatment of
exhaust air from a pig rearing building: Ammonia removal performance and its
fluctuations." Biosystems Engineering 113(3): 242-252.
Merchant, J. A. and R. F. Ross (2002). Iowa Concentrated Animal Feeding
Operations Air Quality Study, Iowa State University and The University of Iowa
Group.
Nielsen, L. P., P. B. Christensen, et al. (1990). "Denitrification and Oxygen
Respiration in Biofilms Studied with a Microsensor for Nitrous-Oxide and Oxygen."
Microbial Ecology 19(1): 63-72.
Olson, K. R. (2012). "A practical look at the chemistry and biology of hydrogen
sulfide." Antioxid Redox Signal 17(1): 32-44.
37
Oneill, D. H. and V. R. Phillips (1992). "A Review of the Control of Odor Nuisance
from Livestock Buildings .3. Properties of the Odorous Substances Which Have Been
Identified in Livestock Wastes or in the Air around Them." Journal of Agricultural
Engineering Research 53(1): 23-50.
Ottosen, L. D. M., S. Juhler, et al. (2011). "Regulation of ammonia oxidation in
biotrickling airfilters with high ammonium load." Chemical Engineering Journal
167(1): 198-205.
Reiffenstein, R. J., W. C. Hulbert, et al. (1992). "Toxicology of hydrogen sulfide."
Annu Rev Pharmacol Toxicol 32: 109-134.
Riis, A. L. and M. Lyngbye (2005). Grisenes indflydelse på lugtemissionen,
LANDSUDVALGET FOR SVIN OG DANSKE SLAGTERIER.
Riis, A. L., M. Lyngbye, et al. (2008). Afprøvning af vertikalt biofilter efter
amerikansk princip, DANSK SVINEPRODUKTION OG VIDENCENTER FOR
SVINEPRODUKTION, DEN RULLENDE AFPRØVNING.
Sander, R. (1999). "Compilation of Henry's Law Constants for Inorganic and Organic
Species of Potential Importance in Environmental Chemistry (Version 3)." from
http://www.henrys-law.org.
Schiffman, S. S., J. L. Bennett, et al. (2001). "Quantification of odors and odorants
from swine operations in North Carolina." Agricultural and Forest Meteorology
108(3): 213-240.
Schiffman, S. S., E. A. S. Miller, et al. (1995). "The Effect of Environmental Odors
Emanating from Commercial Swine Operations on the Mood of Nearby Residents."
Brain Research Bulletin 37(4): 369-375.
Sheridan, B., T. Curran, et al. (2002). "Biofiltration of odour and ammonia from a pig
unit - a pilot-scale study." Biosystems Engineering 82(4): 441-453.
Szabo, C. (2007). "Hydrogen sulphide and its therapeutic potential." Nat Rev Drug
Discov 6(11): 917-935.
38
Tabase, R. K., D. Liu, et al. (Unpublished). "Chemisorption of hydrogen sulphide and
methanethiol by light expanded clay aggregates."
Trabue, S., K. Scoggin, et al. (2008). "Field sampling method for quantifying volatile
sulfur compounds from animal feeding operations." Atmospheric Environment
42(14): 3332-3341.
Trabue, S. L., K. D. Scoggin, et al. (2008). "Field sampling method for quantifying
odorants in humid environments." Environmental Science & Technology 42(10):
3745-3750.
van Gemert, L. J. (2003). Compilations of odour threshold values in air, water and
other media. Huizen, The Netherlands Boelens Aroma Chemical Information Service
Warneke, C., J. Kuczynski, et al. (1996). "Proton transfer reaction mass spectrometry
(PTR-MS): Propanol in human breath." International Journal of Mass Spectrometry
154(1-2): 61-70.
Wing, S., R. A. Horton, et al. (2008). "Air pollution and odor in communities near
industrial swine operations." Environ Health Perspect 116(10): 1362-1368.
39
Correlation between fungi and removal of hydrogen sulfide in air treating
biofilter
Claus Lunde Pedersen1,*), Michael Jørgen Hansen2), Louise Helene Søgaard Jensen3),
Lise Bonne Guldberg4), Anders Feilberg2), and Lars Peter Nielsen1)
1) Department of Bioscience, Microbiology, Aarhus University, Ny Munkegade 116,
8000 Aarhus, Denmark; 2) Department of Engineering, Aarhus University, Blichers
Allé 20, 8830 Tjele, Denmark; 3) Centre for Electron Nanoscopy, Technical
University of Denmark, Fysikvej, building 307, 2800 Kgs. Lyngby, Denmark; 4)
SKOV A/S, Department of Research and Development, Hedelund 4, 7870 Glyngøre,
Denmark
*Corresponding author: Tel.: 0045 8715 4336, Fax: 0045 8715 4326, E-mail address:
40
Abstract:
Biological air filters used for reducing emissions of odorants are often challenged by
low solubility of reduced sulfur compounds (RSC) when these compounds are present.
In a sectioned biofilter treating exhaust air from pig production the removal
efficiencies of hydrogen sulfide (H2S) were found to be too high to be explained by
conventional rates of bacterial biofilm metabolism and diffusion. Measurements of
the specific uptake of RSC through the filter matrix using online Proton-Transfer-
Reaction Mass Spectrometry (PTR-MS) were compared with cryo scanning electron
microscopy of the biofilm. There was a positive correlation between the uptake of
H2S and presence of aerial fungal hyphae. It is therefore suggested that the observed
fungi catalyze H2S oxidation and may play a major role in biofilters due to their
expansion of the active surface area by hyphae.
Keywords: Biofiltration, sulfur gasses, fungi, sulfide oxidation, proton transfer
reaction mass spectrometer,
Abbreviations:
(RSC) reduced sulfur compounds, (PTR-MS) Proton-Transfer-Reaction Mass
Spectrometry, (H2S) hydrogen sulfide, (cryo-SEM) cryo scanning electron
microscopy, (NH3) ammonia, (MT) methanethiol, (DMS) dimethyl sulfide, (DOF)
depth of field, (sem) standard error of the mean, (ppbv) parts per billion volumetric
41
1 Introduction:
Emissions of odorous gasses and ammonia (NH3) from intensive pig houses are
causing nuisance to nearby rural communities (Wing et al. 2008) and eutrophication
to the environment (Hutchings et al. 2001). The odorous gasses are formed in anoxic
liquid manure and emitted from the slurry pits in the pig production facility. Some of
the gasses are highly volatile, especially reduced sulfur compounds (RSC) such as
hydrogen sulfide (H2S), methanethiol (MT) and dimethyl sulfide (DMS) (Schiffman
et al. 2001). These compounds are emitted in the ppbv range (Blanes-Vidal et al.
2009; Feilberg et al. 2010; Hansen et al. 2012), and may be the major contributors to
odor nuisance due to low odor threshold values (Devos et al. 1990; van Gemert 2003).
RSC also represent an energy source to aerobic bacteria in biotrickling filters
(Sheridan et al. 2003; Chen and Hoff 2009). In general a biofilter works by promoting
microbial growth on a large surface area exposed to compounds, which the bacteria
can metabolize. To make sure the biofilm does not dry out, the biofilm is either
irrigated directly or the air is humidified before it passes over the biofilm. This study
investigated a biofilter, where hydrogen sulfide removal rates were too high to be
explained by bacterial metabolism (Hansen et al. 2012).
2 Material and methods:
2.1 Description of the biofilter
A three-step biofilter from SKOV A/S was installed at a pig production facility with
350 growing-finishing pigs. The pig production facility and the biofilter have
previously been described in details by Hansen et al. (Hansen et al. 2012). The third
step was designed as a vertical filter with the aim of removing RSC. Dimensions of
the third step were 2 m in height, 5 m wide and 60 cm in depth. Cellulose was used as
42
filter material with a surface to volume ratio of 384 ± 4% m2 m-3. The humidified air
from step 1 and 2 served as source for humidification of the filter in step 3.
2.2 Sampling:
A 10 cm wide, cylindrical core was cut and taken out from the center of the third step
and filter material was sampled from the core in depths of 20 cm and 40 cm from the
front. Heated Teflon tubes with a 5 !m Teflon filter was inserted at the two sampled
depths before the core were put back into the filter. Further more filter materials were
sampled on the front (0 cm) and on the backside of the filter (60 cm) and heated
Teflon tubes were placed before and after the filter.
2.3 PTR-MS
A PTR-MS (Ionicon Analytik, Innsbruck, Austria) was applied for measuring the
removal of odorants in the biofilter. All four Teflon tubes from the filter were
connected to the PTR-MS through a heated switchbox with a five-way PEEK valve
(Bio-Chem Valve Incorporated, Boonton, New Jersey, USA). The switchbox was
controlled by the PTR-MS software. Specific details on the operating conditions of
the PTR-MS can be found in Hansen et al. (Hansen et al. 2012). PTR-MS and
switchbox were placed in an insulated room next to the biofilter. The removal of
odorants was measured continuously for 45 hours with the PTR-MS. before step 3
(0cm), at the depth of 20 cm and 40 cm and after step 3 (60 cm) in the biofilter (See
figure 1). The PTR-MS measured for 10 minutes at each depth in a continuous cycle.
2.4 Fungal cover on biofilter surface:
43
With the core section taken out of the filter as described above in section 2.2, photos
were taken at each sampling depth from random angles. Fungal covers were then
calculated from the photos as the fraction of filter surface covered by moldy patches.
Image calculations were made with software program ImageJ, National Institute of
Health, USA (http://rsbweb.nih.gov/ij/index.html).
2.5 Cryo scanning electron microscopy (cryo-SEM):
To determine the fungal density, sampled material with moldy patches from each
depth were kept at 0-5 °C until they were attached to sample holder with cryo media
(Tissue-Tek O.C.T.) and plunge- frozen in slush liquid nitrogen. After freezing, the
sample was quickly moved under vacuum to the microscope in a preparatory
chamber, which was kept a -120 °C. Surface water sublimation was done by
increasing temperature from -120 °C to -90 °C at different time lengths, between 0
and 15 min. After this step, samples were coated with platinum for 60 s (at 5 mA and
2E-2 mbar) before moving the samples into the microscope for evaluation. Samples
were evaluated under high vacuum and -120 °C at varying spot size and electron
acceleration voltage (1-5 kV) in a Quanta 200 ESEM FEG, (FEI, Eindhoven,
Netherlands) fitted with a Quorum PP2000 Cryo System (Quorum Technologies Ldt,
East Sussex, UK). In some cases samples were fractured after freezing. The fracture
was perpendicular with the biofilm surface to attain a cross section image of the
sample.
2.6 Fungi surface area calculations
44
To determine the effect of fungal presence on the biofilm, the specific surface area of
hyphae were calculated on cryo-SEM images from a line intercept method developed
by Newman (Newman 1966) and modified to find
[Specific hyphae surface area per biofilm area] = !2N2H
D
Where H is the length of a series of transects, N the number of intersects between
hyphae and transects and D is the diameter of hyphae. Transects and intersects were
measured using software program ImageJ, National Institute of Health, USA
(http://rsbweb.nih.gov/ij/index.html).
3 Results:
Ventilation rate varied daily between 12696 to 29733 m3 hour-1 corresponding to an
air speed between 0.35 and 0.83 m s-1 respectively with highest ventilation rates
during afternoons, reflecting that ventilation rate was controlled by air temperature in
the pig stable (figure 2). RSC load remained more stable over the period with a
decrease in inlet concentrations when ventilation rate was rising (Figure 2 & 3). On
average ± standard error of the mean (sem) 66 ± 1 % of H2S was removed over the
45-hour measurements, with a significant difference in inlet and outlet concentrations
(P < 0.001 using student's t-test) (Figure 3). 73% of the removed H2S was taken up in
the first 20 cm and the rest was taken up between 20 cm and 40 cm depth. There were
no significant changes between inlet and outlet concentrations, when applying
student’s t-test for MT (P > 0.66) or DMS (P > 0.48).
The highest percentage of fungal cover (average ± sem) were calculated from
photographs to be on the front side and decrease with depth: 31 ± 3 % for 0 cm, 24.82
45
± 1 % for 20 cm, 2.56 ± 0.5 % for 40 cm and 1.39 ± 0.5 % for 60 cm. A high contrast
between the white patches of fungi and the dark filter material ensured good accuracy
of determining the fungal cover.
From cryo SEM images average diameters (± sem) of the fungal hyphae were found
to be 1.3 ± 0.04 µm (n = 94), see Figure 4. From the image analysis the fungal density
in moldy patches was determined to be highest at 20 cm (1.77 µm2
µm2) see figure 6.
Combining the information from the fungal cover and the fungal density estimates of
the air exposed fungal surface area was calculated for each depth and highest area was
found in the front of the filter (0.47 ) and decreasing with depth, see figure 6.
The image analysis was potentially limited by the depth of field (DOF). In cases
where fungi were extending out of the biofilm with several micrometers (figure 5),
then DOF would be too small at high magnification. To avoid for this potential error
every high magnification image was accompanied by a series of low magnification
images for the same biofilm area. With lower magnification, DOF increased and
verified that no hyphae were excluded from the DOF. Freeze fractured profiles (figure
5) showed fungal growth extending out from the biofilm by 60µm at the front of step
3. Preparations for the cryo-SEM images proved to be at suitable method for
evaluating delicate biological material. In a few cases fractures of fungal hyphae was
observed, but it was difficult to verify if it had happened naturally or during handling
of the sample, so areas with broken hyphae were avoided for image analysis. Few
images contained artifacts in the form of ice crystals, but they were dismissible due to
their distinguishable features (Echlin 1991). Platinum coating was estimated to be 17
µm2
µm2
46
nm, based on previous experiments (data not shown). In a few cases, the coating was
ripped off the samples possibly due to temperature changes in the preparatory
chamber or during microscopy. However, these areas were dismissed due to charging
and samples would appear white and without details in structure where the platinum
was gone. Still cryo-SEM images were of high quality with low distortion even at
high magnification. Drift was observed in some images at high magnification, but no
image analysis was made from high magnification microscopy.
4 Discussion:
As observed previously (Hansen et al. 2012) H2S was efficiently removed in step 3 of
the biofilter, while neither MT nor DMS were removed to any significant degree.
From a physical and biological point of view MT and DMS were actually expected to
be removed at higher rates than H2S, since MT and DMS have higher water solubility
than H2S (Coquelet and Richon 2005; Iliuta and Larachi 2006) and contain methyl
groups which sulfide oxidizers are capable of separating and utilizing as electron
donors in addition to the sulfide (Friedrich et al. 2001; Meyer et al. 2007). Activity of
conventional sulphide oxidizing bacteria was therefore not sufficient to explain the
observations.
To evaluate if biofilm activity alone could at all explain the high removal rate of H2S
a one dimensional model based on Fick’s first law was used to calculate sulfide
penetration in the biofilm based on the surface flux and surface concentration
assuming the same H2S consumption capacity throughout the biofilm (Nielsen et al.
1990). The flux, J, at any depth is given by:
J = !D "C"x
47
Where D is diffusion coefficient for H2S in water: 1.59E-05 cm2 s-1 (Broecker and
Peng 1974). With a known flux in the surface, J0 (H2S removal rate divided by filter
surface area), and an aqueous surface concentration C0 (calculated from Cair and
Henry's law's constant: H2S solubility at 17 °C in pure water was calculated to be
1.244E-1 atm M-1 (Liu et al. 2012)) concentration Cx at the depth x of the biofilm
becomes:
A penetration depth of 2.9 µm was calculated in the biofilm of the first 20 cm section
having a surface flux of 2.74 pmol cm-2 s-1 at an average air concentration of 200 ppb
(Figure 7). In the 20 cm to 40 cm section, the surface flux was calculated to be 0.99
pmol cm-2 s-1 yielding a penetration depth of 2.7 µm. With a typical bacterial cell size
of 1-2 µm, metabolism would thus have to take place in a single cell layer exposed
directly to the air. This is not realistic as very few cells were even visible by cryo-
SEM on the surface due to EPS (extracellular polymeric substances).
The biological and physical considerations above indicated that other agents than the
bacterial biofilm must be important for H2S removal in the biofilter. A positive
correlation between H2S uptake and presence of fungi strongly suggested that fungal
hyphae could be the missing agents (Figure 6). With the cells directly exposed to the
air they were evidently in a better position than bacteria embedded in the biofilm
below, but fungal oxidation of H2S has so far not been reported in the literature. It is
known, however, that fungi can oxidize other sulfurous compounds like elemental
sulfur (Grayston and Wainwright 1988) and thiosulfate (Jones et al. 1991). Biofilters
Cx =J02
4DC0x2 ! J0
Dx +C0
48
with cultivated fungi can also efficiently take up and metabolize other volatile gasses
like toluene (Kennes and Veiga 2004). Further studies are needed to determine
whether H2S oxidation is part of the fungal metabolism, a necessary detoxification
mechanism or a process spontaneously catalyzed by specific components on the
surface or inside the cells.
Acknowledgements:
This study was funded by Danish Council for Strategic Research (DSF), funding
number (2104-08-0017).
References:
Blanes-Vidal, V., M. N. Hansen, et al. (2009). "Characterization of odor released
during handling of swine slurry: Part I. Relationship between odorants and perceived
odor concentrations." Atmospheric Environment 43(18): 2997-3005.
Broecker, W. S. and T. H. Peng (1974). "Gas-Exchange Rates between Air and Sea."
Tellus 26(1-2): 21-35.
Chen, L. and S. J. Hoff (2009). "Mitigating odors from agricultural facilities: A
review of literature concerning biofilters." Applied Engineering in Agriculture 25(5):
751-766.
Coquelet, C. and D. Richon (2005). "Measurement of Henry's Law Constants and
Infinite Dilution Activity Coefficients of Propyl Mercaptan, Butyl Mercaptan, and
Dimethyl Sulfide in Methyldiethanolamine (1) + Water (2) with w1 = 0.50 Using a
Gas Stripping Technique." Journal of Chemical & Engineering Data 50(6): 2053-
2057.
Devos, M., F. Patte, et al. (1990). Standardized human olfactory thresholds. New
York, Oxford University Press
49
Echlin, P. (1991). "Ice crystal damage and radiation effects in relation to microscopy
and analysis at low-temperatures." Journal of Microscopy-Oxford 161: 159-170.
Feilberg, A., D. Liu, et al. (2010). "Odorant Emissions from Intensive Pig Production
Measured by Online Proton-Transfer-Reaction Mass Spectrometry." Environmental
Science & Technology 44(15): 5894-5900.
Friedrich, C. G., D. Rother, et al. (2001). "Oxidation of Reduced Inorganic Sulfur
Compounds by Bacteria: Emergence of a Common Mechanism?" Applied and
Environmental Microbiology 67(7): 2873-2882.
Grayston, S. J. and M. Wainwright (1988). "Sulfur oxidation by soil fungi including
some species of mycorrhizae and wood-rotting basidiomycetes." Fems Microbiology
Ecology 53(1): 1-8.
Hansen, M. J., D. Liu, et al. (2012). "Application of Proton-Transfer-Reaction Mass
Spectrometry to the Assessment of Odorant Removal in a Biological Air Cleaner for
Pig Production." Journal of Agricultural and Food Chemistry 60(10): 2599-2606.
Hutchings, N. J., S. G. Sommer, et al. (2001). "A detailed ammonia emission
inventory for Denmark." Atmospheric Environment 35(11): 1959-1968.
Iliuta, M. C. and F. Ø. a. Larachi (2006). "Solubility of Total Reduced Sulfurs
(Hydrogen Sulfide, Methyl Mercaptan, Dimethyl Sulfide, and Dimethyl Disulfide) in
Liquids." Journal of Chemical & Engineering Data 52(1): 2-19.
Jones, R., S. M. Parkinson, et al. (1991). "Oxidation of thiosulfate by fusarium-
oxysporum grown under oligotrophic conditions." Mycological Research 95: 1169-
1174.
Kennes, C. and M. C. Veiga (2004). "Fungal biocatalysts in the biofiltration of VOC-
polluted air." Journal of Biotechnology 113(1-3): 305-319.
Liu, D., A. Feilberg, et al. (2012). "PTR-MS measurement of partition coefficients of
reduced volatile sulfur compounds in liquids from biotrickling filters."
Chemosphere(0).
50
Meyer, B., J. F. Imhoff, et al. (2007). "- Molecular analysis of the distribution and
phylogeny of the soxB gene among sulfur-oxidizing bacteria – evolution of the Sox
sulfur oxidation enzyme system." - 9(- 12): - 2977.
Newman, E. I. (1966). "A Method of Estimating Total Length of Root in a Sample."
Journal of Applied Ecology 3(1): 139-&.
Nielsen, L. P., P. B. Christensen, et al. (1990). "Denitrification and Oxygen
Respiration in Biofilms Studied with a Microsensor for Nitrous-Oxide and Oxygen."
Microbial Ecology 19(1): 63-72.
Schiffman, S. S., J. L. Bennett, et al. (2001). "Quantification of odors and odorants
from swine operations in North Carolina." Agricultural and Forest Meteorology
108(3): 213-240.
Sheridan, B. A., T. P. Curran, et al. (2003). "Biofiltration of n-butyric acid for the
control of odour." Bioresource Technology 89(2): 199-205.
van Gemert, L. J. (2003). Compilations of odour threshold values in air, water and
other media. Huizen, The Netherlands Boelens Aroma Chemical Information Service
Wing, S., R. A. Horton, et al. (2008). "Air pollution and odor in communities near
industrial swine operations." Environ Health Perspect 116(10): 1362-1368.
51
Figures:
Figure 1: Schematic drawing of the 3 step biofilter with sampling points; 0 cm (square), 20 cm (diamond), 40 cm (triangle) and 60 cm (circle).
Figure 2: Ventilation rate, m3 h-1 (thick line) and RSC load indexed at starting level for hydrogen sulfide:
179.58 mmol h-1 (square), methane thiol: 9.46 mmol h-1 (diamond) and dimethyl sulfide: 2.73 mmol h-1
(triangle).
!"#
$"%&'#() $"%&'#(* $"%&'#(+
,(-. *,(-. /,(-. 0,(-.
!"
#!!!"
$!!!!"
$#!!!"
%!!!!"
%#!!!"
&!!!!"
!'"
%!'"
(!'"
)!'"
*!'"
$!!'"
$%!'"
$(!'"
$)!'"
$%+!!" $*+!!" !+!!" )+!!" $%+!!" $*+!!" !+!!" )+!!" $%+!!"
,-./"01234"5&"6
7$"
89:3;3:"<
=>"-.1:4"'"
?53"
@%=" A?" BA=" C-./"0123"
52
Figure 3: Concentration (ppbv) measurements in four depths of section 3 of hydrogen sulfide (A), methane
thiol (B) and dimethyl sulfide (C) respectively (0 cm; square, 20 cm; diamond, 40 cm; triangle and 60 cm;
circle).
!"
#!"
$!!"
$#!"
%!!"
%#!"
&!!"
&#!"
$%+!!" $*+!!" !+!!" )+!!" $%+!!" $*+!!" !+!!" )+!!" $%+!!"
'()*+),-.,/()0"112
3"
4/5+"
!"D5"" %!"D5" (!"D5" )!"D5"
6"
!"
#"
$!"
$#"
%!"
$%+!!" $*+!!" !+!!" )+!!" $%+!!" $*+!!" !+!!" )+!!" $%+!!"
'()*+),-.,/()0"112
3"
4/5+"
!"D5"" %!"D5" (!"D5" )!"D5"7"
!"
$"
%"
&"
8"
#"
9"
:"
$%+!!" $*+!!" !+!!" )+!!" $%+!!" $*+!!" !+!!" )+!!" $%+!!"
'()*+),-.,/()0"112
3"
4/5+"
!"D5"" %!"D5" (!"D5" )!"D5"'"
53
Figure 4: cryo-SEM image of spore forming hyphae with apical spherical spores. Image made from a
sample taken from the step 3 at 20 cm depth.
54
Figure 5: cryo-SEM image of step 3 at 0 cm. profile image made by freeze fracturing the biofilm surface
with spore forming hyphae. In places the fungi stretched out more than 60 µm from the biofilm surface.
Figure 6: Calculated fungal hyphae surface area average and standard error of the mean per biofilm area (µm2/µm2), open bars. Calculation of the projected area is placed above each bar by multiplying fungal cover in percentage with the fungal surface area to filter area (µm2 µm-2). Hydrogen sulfide percentage of inlet concentration between sample points 0 cm, 20 cm, 40 cm and 60 cm is shown in a black line.
!"
!#$"
!#%"
!#&"
!#'"
!#("
!#)"
!*"
%!*"
'!*"
)!*"
+!*"
$!!*"
$%!*"
!" $!!" %!!" &!!" '!!" (!!" )!!"
,-./01203"456789"8-08":0-";<.=<9>
"8-08"?5>@%"5>
@A%B"
C%D"82"<6903"1.61062-82<.6#"*"
E<920-"30:2F"G>>H"
,-./01203"456789"8-08"C%D"-0>.I89"
1.65 µm2
µm2!28.4%
1.77 µm2
µm2!24.8%
1.42 µm2
µm2!2.56%
0.87 µm2
µm2!1%
55
Figure 7: Calculated profiles for 0 cm – 20 cm (permanent) and 20 cm – 40 cm (dotted) in step 3.
!E!"
!E#"
$E!"
$E#"
%E!"
%E#"
!" #" $!" $#" %!" %#"
;+1,<0"=>5
?"@%A"'()*+),-.,/()"=15(BC*5&?"
!"7"%!"D5""
%!"7"(!"D5"
56
Fungi metabolize sulfide in an air treating biofilter
Claus Lunde Pedersen*,1), Dezhao Liu2), Anders Feilberg2) and Lars Peter Nielsen1)
1) Department of Bioscience, Microbiology, Aarhus University, Ny Munkegade 116,
8000 Aarhus, Denmark; 2) Department of Engineering, Aarhus University, Blichers
Allé 20, 8830 Tjele, Denmark
*Corresponding author: Tel.: 0045 8715 4336, Fax: 0045 8715 4326, E-mail address:
57
Abstract:
Fungal aerial hyphae were observed three times over a period of 20 months in a
biofilter treating waste air from a pig farm. Removal rates of hydrogen sulfide (H2S)
were positively correlated with fungal aerial hyphae, which suggest that the fungi take
up and incorporate H2S oxidation in the fungal metabolism. Given that fungal aerial
hyphae occurred at short intervals suggests, that H2S is not the primary energy source
for the fungi and that the main strategy for formation of fungal aerial hyphae must be
related to other purposes, such as sporulation. Presence of fungal aerial hyphae had no
significant influence on any other measured sulfurous compounds in the ventilation
air.
58
Introduction:
Biofiltration of volatile compounds has been implemented in agricultural industry,
where it has been successful in reducing emissions of ammonia and volatile carbon
compounds from exhaust air from pig stables (Juhler et al. 2009; Kristiansen et al.
2011). Sulfurous compounds however have been more difficult, which mainly has
been ascribed to their low solubility in water. Hydrogen sulfide (H2S) is considered
to be the main odor compound emitted from pig facilities due to its emission
concentration from pig facilities and its low odor threshold (Feilberg et al. 2010). In a
previous study a positive correlation was found between H2S removal and presence of
fungal aerial hyphae in the biofilter (Pedersen et al. in preparation). Fungi have been
introduced in biofiltration before, however there has been no previous studies no
fungi metabolizing H2S (Kennes and Veiga 2004; Estevez et al. 2005). H2S has been
described as a poisonous compound due to its inhibitory effect on mitochondrial
cytochrome c oxidase (Cooper and Brown 2008). Given that cell membranes have
similar H2S solubility as water, membranes are not obstacles for diffusion of H2S and
H2S can pass through the cell membrane from the air and react with cytochrome c
oxidase (Mathai et al. 2009; Cuevasanta et al. 2012). H2S has been proposed to be
oxidized in the mitochondria following an enzymatic pathway proposed by
Hildebrandt and Grieshaber (Hildebrandt and Grieshaber 2008). This pathway
incorporates electrons into the electron transport chain from the oxidation of H2S to
elemental sulfur with a potential ATP production (Olson 2012). Based on the
correlation between fungal aerial hyphae and removal of H2S, we hypothesize that
fungi oxidize H2S. To test the hypothesis, inlet and outlet concentrations of H2S were
measured in the presence and absence of fungal aerial hyphae in a biofilter treating
59
waste air from a pig farm. MT and DMS concentrations were recorded as a proxy for
microbial activity.
Materials and methods:
Filter description:
The filter and odor source has previously been described in following papers
(Pedersen et al. in preparation)(Liu et al. 2012). A three step biofilter (SKOV A/S)
treating exhaust air from a pig stable with 350 growing finishing pigs were
investigated, in which fungal aerial hyphae previously have been observed on the last
step. The biofilter was placed in a separate building next to the pig production facility.
The filter dimensions were 5 meter in with, 2 meter in height and 60 cm in depth over
the third step. Maximum ventilation air rate was set to 35.000 m3 h-1 giving a
maximum air velocity of approximately 1 meter second-1. Filter material was made of
cellulose with a surface to volume ratio of 384 m2 m-3. Humidified air from step 1 and
2 served as source for humidification of the filter wall in step 3.
Sampling:
Air samples were taken before and after step 3 of the biofilter. All air samples were
collected between 1st of March 2011 and 30th of September 2012. Sampling took
place outside the filter so that airflow was not disturbed. Air was collected in Tedlar
bags (volume = 10 L) situated in a vacuum box, which were connected with the inside
of the filter by Teflon tubes (! = 5 mm). A pump (XXX) generated negative pressure
in the vacuum box, which filled the Tedlar bag with air from the biofilter. Filling time
was 5 minutes for each bag at a rate of 2 Liter per minute. After air sampling the
biofilter was visual inspected for fungal growth and ventilation rate through the filter
60
was recorded. Air samples were analyzed within 2 hours after sampling using PTR-
MS.
PTR-MS:
A PTR-MS (Ionicon Analytik, Innsbruck, Austria) was applied for measuring the
collected air samples from the biofilter. PTR-MS was operated under standard ion
drift tube conditions applying a total voltage of 600 V and maintaining the pressure in
the range of 2.1 - 2.2 mbar. The measurements were performed as single ion
monitoring with each ion being detected for 2000 - 4000 ms. A mean was calculated
from at least 15 measurements H2S, MT and DMS respectively. Before the bag
measurements a background was measured on room air purified for hydrocarbon
contaminants via a Supelpure™ HC filter (Supelco, St. Louis, Missouri, USA). The
measurement of hydrogen sulfide with PTR-MS is humidity dependent and the
concentration of hydrogen sulfide was calculated in accordance to the method
described by (Feilberg et al. 2010).
Fungal evaluation:
After air sampling the external surfaces of the biofilter were inspected for presence or
absence of fungi.
61
Results:
There were a correlation in time between fungal presence and increased H2S removal.
15 times were air samples collected at the biofilter in 2011, where samples in March
and April (n = 6), where no fungi were observed. Fungal aerial hyphae appeared in
July and August (n = 4), but disappeared in September and October (n = 3) only to
reappear in November (n = 2) (see figure 1). Further 7 samples were collected in
September 2012, when fungal aerial hyphae were appeared for the first time since
November 2011. Average H2S loading rates (see table 1) remained similar for
samples with and without fungal presence with no significant difference (P > 0.75).
There was a significant difference in removal efficiency (P < 0.001). No difference
was observed in loading rate for MT (P > 0.71) with or without fungal presence and
no significant difference in removal efficiency (P > 0.75). There was a significant
difference in loading rate for DMS, when comparing rates with or without fungal
presence (P < 0.001), despite that there was no significant difference in removal
efficiency (P > 0.87) with or without fungal presence. Formation of elemental sulfur
was not detected at any time during the measurements, suggesting that H2S oxidation
was not due to chemical auto oxidation.
Discussion:
Y1) Little or no bacterial activity could be detected regarding oxidation of MT and
DMS supporting that H2S oxidation is related to something else than bacterial
activity. Y2) Y2B) Fungal aerial hyphae had previously been observed at the same
62
biofilter in August and September 2010. In this period the same PTR-MS used for air
samples had been measuring inlet and outlet concentrations of RSC for three
consecutive weeks, of which 13 days were included in the article by Hansen et al.
(Hansen et al. 2012). Y3) Hansen et al. had also observed high removal efficiencies
for H2S in the third step and no significant removal of MT or DMS,. Y4) There was
no difference in H2S removal in the different periods of fungal presence, however it
was not certain whether it was the same organism present in the different periods (see
figure 2). Y5) Assumed that there is a maximum rate of fungal H2S oxidation as
described in the introduction, then exceeding it would lead to a partial or fully
inhibition of complex IV in the electron transport chain, blocking the metabolic
system(Cooper and Brown 2008). However fungal H2S removal was observed to
increase with increasing H2S concentration, so an upper limit had not yet been
reached. Y6) Whether or not H2S could serve, as substrate for the fungi is
questionable. Y7) The total metabolic rate of H2S oxidation was known, however to
calculate the specific rate fungal biomass or bio-volume was needed. Previous study
showed, that the fungal distribution was heterogenic distributed in the biofilter and
hence difficult to assess (Pedersen et al., submitted). However distinguishing between
presence and absence of fungal aerial hyphae were sufficient to explain the significant
difference in H2S oxidation. Given that most of the compounds in the air has been
taken up in the first or second stage of the filter leaves few options for the fungi to
metabolize. It could be speculated, that the filter material, cellulose, served as
substrate for the fungi and that H2S oxidation were metabolized by chance, while the
fungi were releasing its spores. That could explain, why fungal aerial hyphae only
were present in short intervals during the years of observation. Fungi have been
utilized as biofilters treating hazardous substances(Kennes and Veiga 2004) and may
63
in future also be considered as effective way of reducing H2S from exhaust air using
biofiltration, provided that presence of fungal aerial hyphae can be maintained.
Tables:
Table 1:Average loading rate ± SEM and removal efficiency shown in brackets
for hydrogen sulfide, methane thiol and dimethyl sulfide, with or without fungal
presence.
With presence of fungal aerial hyphae:
Average loading rate ± SEM, mmol min-1 -
removal efficiency (%)
Without presence of fungal aerial hyphae:
Average loading rate ± SEM, mmol min-1 -
removal efficiency (%)
H2S 3.14 ± 0.82 (64.28) 2.91 ± 0.26 (17.6)
MT 0.08 ± 0.01 (7.4) 0.075 ± 0.01 (5.1)
DMS 0.065 ± 0.01 (2.53) 0.036 ± 0.01 (3.1)
Figures:
!"
#!!"
$!!"
%!!"
&!!"
'!!!"
#(#%(''" $(')(''" %(%(''" )(#%(''" *('$(''" ''(+(''"
!"#$%#&'(&)"#*+,,-+
.(&%+
,"-./01"1/234" ","-./01"5.4234" 6"-./01"1/234" 6"-./01"5.4234"/+
64
Figure 1: Hydrogen sulfide (A), Methane thiol (B) and dimethyl sulfide (C)
having concentration (ppbv) plotted against sample dates. Inlet concentration
(diamonds), outlet concentration (squares) marking presence of fungi (filled) and
without fungi (open).
!"
7"
'!"
'7"
#!"
#7"
#(#%(''" $(')(''" %(%(''" )(#%(''" *('$(''" ''(+(''"
!"#$%#&'(&)"#*+,,-+
.(&%+
,"-./01"1/234" ","-./01"5.4234" 6"-./01"1/234" 6"-./01"5.4234"0+
!"#"$"%"&"'!"'#"'$"
#(#%(''" $(')(''" %(%(''" )(#%(''" *('$(''" ''(+(''"
!"#$%#&'(&)"#*+,,-+
.(&%+
,"-./01"1/234" ","-./01"5.4234" 6"-./01"1/234" 6"-./01"5.4234"!+
65
Figure 2: Removal efficiency (%) plotted against ventilation rate (m3 hour-1).
Points from 2010 were recalculated from Hansen et al. 2012 and Pedersen et al.,
submitted.
References:
Cooper, C. E. and G. C. Brown (2008). "The inhibition of mitochondrial cytochrome
oxidase by the gases carbon monoxide, nitric oxide, hydrogen cyanide and hydrogen
sulfide: chemical mechanism and physiological significance." J Bioenerg Biomembr
40(5): 533-539.
Cuevasanta, E., A. Denicola, et al. (2012). "Solubility and Permeation of Hydrogen
Sulfide in Lipid Membranes." PLoS One 7(4).
Estevez, E., M. C. Veiga, et al. (2005). "Biofiltration of waste gases with the fungi
Exophiala oligosperma and Paecilomyces variotii." Appl Microbiol Biotechnol 67(4):
563-568.
Feilberg, A., D. Liu, et al. (2010). "Odorant Emissions from Intensive Pig Production
Measured by Online Proton-Transfer-Reaction Mass Spectrometry." Environmental
Science & Technology 44(15): 5894-5900.
!8"
#!8"
$!8"
%!8"
&!8"
'!!8"
!" '!!!!" #!!!!" +!!!!" $!!!!"
9:;"8"
<3/412=415/;">+"?5.@,'"
6"-./01"#!''" ,"-./01"#!''" 6"-./01"#!'!"
66
Hansen, M. J., D. Liu, et al. (2012). "Application of Proton-Transfer-Reaction Mass
Spectrometry to the Assessment of Odorant Removal in a Biological Air Cleaner for
Pig Production." Journal of Agricultural and Food Chemistry 60(10): 2599-2606.
Hildebrandt, T. M. and M. K. Grieshaber (2008). "Three enzymatic activities catalyze
the oxidation of sulfide to thiosulfate in mammalian and invertebrate mitochondria."
FEBS J 275(13): 3352-3361.
Juhler, S., N. P. Revsbech, et al. (2009). "Distribution and Rate of Microbial
Processes in an Ammonia-Loaded Air Filter Biofilm." Applied and Environmental
Microbiology 75(11): 3705-3713.
Kennes, C. and M. C. Veiga (2004). "Fungal biocatalysts in the biofiltration of VOC-
polluted air." Journal of Biotechnology 113(1-3): 305-319.
Kristiansen, A., S. Lindholst, et al. (2011). "Butyric Acid- and Dimethyl Disulfide-
Assimilating Microorganisms in a Biofilter Treating Air Emissions from a Livestock
Facility." Applied and Environmental Microbiology 77(24): 8595-8604.
Liu, D., M. J. Hansen, et al. (2012). "Kinetic evaluation of removal of odorous
contaminants in a three-stage biological air filter." Environ Sci Technol 46(15): 8261-
8269.
Mathai, J. C., A. Missner, et al. (2009). "No facilitator required for membrane
transport of hydrogen sulfide." Proc Natl Acad Sci U S A 106(39): 16633-16638.
Olson, K. R. (2012). "A practical look at the chemistry and biology of hydrogen
sulfide." Antioxid Redox Signal 17(1): 32-44.
Pedersen, C. L., M. J. Hansen, et al. (in preparation). "Fungi explain removal of
hydrogen sulfide in air treating biofilter." Chemical Engineering Journal.
67
Bacterial and eukaryotic community in a biofilter treating hydrogen sulfide
Claus Lunde Pedersen*, Andreas Schramm, and Lars Peter Nielsen
Department of Bioscience, Microbiology, Aarhus University, Ny Munkegade 116, 8000 Aarhus, Denmark;
*Corresponding author: Tel.: 0045 8715 4336, Fax: 0045 8715 4326, E-mail address: [email protected]
68
Abstract:
Fungal oxidation of hydrogen sulfide (H2S) was observed for short periods in a
biofilter treating waste air from a pig farm. Biofilter samples were collected in the
presence and absence of fungal aerial hyphae and 16S and 18S rRNA gene clone
libraries were made with the aim of identifying the H2S oxidizing fungal strain and to
test whether bacterial diversity were altered due to fungal presence. Presence or
absence of fungal aerial hyphae had no influence on the bacterial diversity or
eukaryotic diversity in the filter. No strictly H2S oxidizing bacteria were found in the
biofilter. Fungal structure was more important for the removal of H2S than the
phylogeny, when considering fungal H2S oxidation a general feature. So far fungi
have not been exploited as H2S oxidizer and may in future be implemented in
commercial biofilters.
69
Introduction: Emissions from farming industry comprise of odorous compounds, such as ammonia
(NH3), volatile organic compounds (VOC) and sulfur compounds such as hydrogen
sulfide (H2S). Biofiltration has been introduced in farming industry to reduce odor
emissions and the technology has been effective in the removal of NH3 and VOC
(Juhler et al. 2009; Kristiansen et al. 2011). Sulfurous compounds, however, has
proven more difficult to remove with biofiltration due to low emission concentrations
and low water solubility (Feilberg et al. 2010; Kristiansen et al. 2011). Recently a
correlation was found between presence of fungal aerial hyphae in a biofilter and
removal rates of H2S (Pedersen et al.). The investigated biofilter was sectioned into
three parts, where NH3 and VOC were taken up in the section one and two, while H2S
were removed in section three. Further studies showed that presence of fungi in the
filter increased H2S removal by a factor of 4.4 compared to periods without fungi
(Pedersen et al.). It could be speculated that fungal aerial hyphae would influence
bacterial community in such a way, that changes in the bacterial community could
explain the difference in removal efficiency of H2S. To address matters, biofilter
community was compared between periods with and without fungal presence. It was
also investigated how the bacterial community in the third section of the filter
deviated from the community of the first two section, which had been studied
previously(Kristiansen et al. 2011; Kristiansen et al. 2011).
Materials and methods:
Biofilter description:
A sectioned biofilter treating waste air from a pig filter was investigated, where
fungal aerial hyphae had been observed on previous occasions on the surface of the
last section. The filter dimensions were 5 meter in with and 2 meters in height. The
filter was divided into three sections, where first and second was 15 cm in depth and
third section was 60 cm in debt. Filter material for all three sections were cellulose
pads with a surface to volume of 384 m2 m-3. The maximum air speed was set to 1 m
s-1, equaling an air volume rate of 36,000 m3 hour-1. The air came from a pig stable
with 350 growing finishing pigs. Further details on the filter can be found in (Feilberg
et al. 2010; Hansen et al. 2012).
70
Sampling:
On October 25 2011 a core of 10 cm in diameter was taken from the third section and
filter material was sampled on the inlet side (0 cm) at 20 cm, 40 cm and the outlet side
(60 cm). No fungal hyphae had been observed in the filter since August 2011. Fungal
aerial hyphae were observed again in the beginning of November 2011 and new
samples were collected on November 15 2011 following procedure as described
above. Samples were kept on ice during transport to the lab and then stored at -18 °C
until DNA extraction.
Phylogenetic analysis:
Total genomic DNA extraction from samples (0.1±0.05 g) were done by mechanical
and enzymatic lysis (Juretschko et al. 1998) to degrade EPS and cell structures. DNA
purification was made with FastDNA® Spin Kit for Soil, (Qbiogene). DNA
concentration was determined with NanoDrop ND-1000 (Thermo Scientific). PCR
were made with two sets of primers on all 8 samples with total of 16 PCR products.
First primer sets were Bac27F and 1492R (Loy et al. 2002), which resulted in almost
full-length 16S rRNA genes. Second primer set was Euk82F (Lopez-Garcia et al.
2001) and Euk499R (Zhu et al. 2005), which gave a smaller section of the eukaryotic
18S rRNA gene. Annealing temperature for bacterial primer set were set to 57 °C and
ran for 25 PCR cycles, while annealing temperature for eukaryotic primer was 56°C
and ran at 30 PCR cycles. PCR products were purified with SIGMA GenElute! PCR
Clean-Up Kit (Sigma-Aldrich) followed by a ligation step. PCR-product containing
plasmids were transformed into competent E. coli strain JM 109 and screening of
cells containing plasmids with PCR insert was done on LB-plates containing
ampicillin, IPTG and X-Gal for "Blue/White screening". 64 colonies from each
sampling point were transferred and incubated on LB-plates without IPTG and X-Gal
following amplicon sequencing of PCR inserts (GATC Biotech, Germany).
Sequences were manually inspected for sequencing errors followed by alignment
using the SILVA SINA web-alinger (Pruesse et al. 2007). Sequences were loaded into
ARB (Ludwig et al. 2004) for phylogenetic analysis. ! YC dissimilarity matrix
between samples were calculated using open source software Mothur as described in
Schloss et al (Schloss et al. 2011).
71
Results:
There were a total of 435 partial 16S rRNA gene sequences of which 422 were
aligned to phylotypes and 337 partial 18S rRNA gene sequences of which 155 were
aligned to different groups (see table 1).
Table 1: 16S and 18S clone library overview:
Total No fungi With Fungi 0
cm 20 cm
40 cm
60 cm
0 cm
20 cm
40 cm
60 cm
16S: Number of clones 422 62 49 42 57 50 57 48 57 Actinobacteria 37 2 3 6 5 5 4 12 Alphaproteobacteria 30 4 4 7 9 5 1 Bacilli 8 1 5 2 Betaproteobacteria 43 17 5 2 7 8 1 3 Clostridia 7 2 1 1 1 2 Flavobacteria 2 1 1 Gammaproteobacteria 280 34 41 36 37 26 33 35 38 Planctomycetacia 1 1 Sphingobacteriia 13 3 1 2 3 2 Verrucomicrobiae 1 1 2 18S: Number of clones 152 49 10 4 13 48 10 5 14 Fungi: Tremellomycetes(Basidiomycota) 90
28 9 4 1 35 5 2 6
Trichocomaceae(Ascomycota) 15
4 5 2 4
Sordariomycetes(Ascomycota) 4
4
Saccharomycetales(Ascomycota) 7
3 2 2
Metazoa: Rhabditidae(nematode) 5 5 Protists:
Hypotrichia (ciliate) 11 11 Platyophryida(ciliate) 9 1 8 Oligohymenophorea(ciliate) 7 5 2 Proleptomonas(amoeba) 3 1 2 Learamoeba(Amoeba) 1 1
72
Discussion:
Fungal presence did not influence bacterial diversity, when comparing the bacterial
diversity in the presence or absence of fungal aerial hyphae (P > 0.22) and there were
no difference between sampling points, despite a previous study had shown variation
in fungal density with depth of the section (Pedersen et al.). The 8 sampling points
could then be considered one sample and compared with clone library from sections
one and two in similar biofilters investigated previously by Kristiansen et al.
(Kristiansen et al. 2011; Kristiansen et al. 2011). There were distinct differences in
the bacterial community between sections one, two and three. These differences were
ascribed to the strong selective pressure from the differences in variations of nutrients
coming in to the different sections of the biofilter (Hansen et al. 2012). The overlaps
in bacterial community between the different sections were ascribed to bacteria that
were transported with the air from one section to another.
Among the bacterial sequences from the third section, 140 were found to belonging to
Rhodanobacter and closely related to Rhodanobacter thiooxydans (Lee et al. 2007).
Together with the finding that other sequences aligned to Ottowia thiooxydans
(Spring et al. 2004) suggests, that there could be thiosulfate present in the biofilter. It
could be speculated that the fungi oxidize H2S to thiosulfate and excrete it. 120 clones
aligned to uncultured Xanthomonadales, which had also been found in sections one
and two, however they were not closely related. 90 of 18S sequences were identified
to the class tremellomycetes were found in all 8 samples. This made it a candidate as
the H2S oxidizing fungi. However Tremellomycetes is a yeast type, which previously
have been found in biofilters (Kristiansen 2010). Trichocomaceae were found in 3 of
the 4 samples taken in the presence of fungal aerial hyphae and it could be considered
a candidate for the fungi producing the aerial hyphae observed.
73
Figures:
Figure 1: Tree structure showing differences between the 8 samples using a ! YC dissimilarity matrix with a 3 % cutoff of 16S sequences (Left) and 1 % cutoff of 18S sequences (right). There were no significant differences between sampling points for either trees.
0.05
40Okt
20Okt
60Nov
60Okt
20Nov
00Nov
40Nov
00Okt
0.05
60Nov
00Okt
20Nov
60Okt
40Okt
00Nov
40Nov
20Okt
74
Figure 2: Phylogenetic tree of 18S gene sequences from the third section aligned with known sequences from Silva reference tree.
18S20o40, Arachnida, 18S October 20cm
HJCFeron, Arachnida, Histiostoma feroniarum, 93
18S60o37, Eurotiomycetes, 18S October 60cm
PenLagen, Eurotiomycetes, Penicillium lagena, 99
HemOrna2, Eurotiomycetes, Hemicarpenteles ornatus, 100
18S60n44, Eurotiomycetes, 18S November 60cm
18S60n46, Eurotiomycetes, 18S November 60cm
PenFreii, Eurotiomycetes, Penicillium freii, 99
PenTard3, Eurotiomycetes, Penicillium tardum, 99
18S60n34, Tremellomycetes, 18S November 60cm
TrhJirov, Tremellomycetes, Trichosporon jirovecii, 98
TrhLaiba, Tremellomycetes, Trichosporon laibachii, 99
18S60n41, Tremellomycetes, 18S November 60cm
18S40n43, Tremellomycetes, 18S November 40cm
CrpFragi, Tremellomycetes, Cryptococcus fragicola, 99
18S60n33, Mucoromycotina, 18S November 60cm
LbgTrans, Mucoromycotina, Lobosporangium transversale, 100
18S60n45, Fungi_Basal fungi_Basal fungi_LKM11 Rozella, 18S November 60cm
UnlEu741, Fungi_Basal fungi_Basal fungi_LKM11 Rozella, uncultured eukaryote, 87
18S40n45, Fungi_Ascomycota_Pezizomycotina, 18S November 40cm
UnlE1034, Fungi_Ascomycota_Pezizomycotina, uncultured eukaryote, 85
18S40n34, Eukarya, 18S November 40cm
UnlLabyr, Stramenopiles_Stramenopiles_Labyrinthulea_Lab, uncultured labyrinthulid, 64
18S40n44, Stramenopiles_Stramenopiles_Labyrinthulea_Lab, 18S November 40cm
UnlLab72, Stramenopiles_Stramenopiles_Labyrinthulea_Lab, uncultured labyrinthulid, 56
18S40n48, Eukarya, 18S November 40cm
18S60o33, Glissomonadida_Sandonidae, 18S October 60cm
CczSpe10, Glissomonadida_Sandonidae, Cercozoa sp. IVY21ap83t8LS, 97
A9EFaeci, Sarcomonadea, Proleptomonas faecicola, 91
18S60n36, Sarcomonadea, 18S November 60cm
18S40o31, Colpodellidae_Family Incertae Sedis, 18S October 40cm
CryStrut, Colpodellidae_Family Incertae Sedis, Cryptosporidium struthionis, 77
18S60n40, Stichotrichia, 18S November 60cm
UnlOxy30, Stichotrichia, uncultured Oxytrichidae, 97
18S40n36, Scuticociliata, 18S November 40cm
BZ4Pers3, Scuticociliata, Pseudocohnilembus persalinus, 96
18S60n43, Colpodea_Family Incertae Sedis, 18S November 60cm
So0Stoi2, Colpodea_Family Incertae Sedis, Sorogena stoianovitchae, 100
1.00
75
Figure 3a: Tree structure of sequences from the third section (Mors) set against sequences from samples in sections one and two in other SKOV filters (Kiel filter and Roslev filter) (Kristiansen et al 2011;Kristiansen et al. 2011).
ARB_5EF8B508, Rhodanobacter, Mors 40cm With FungiARB_A8D56CDA, Rhodanobacter, Mors 60 cm With Fungi
ARB_33B8A306, Rhodanobacter, Mors 40cm With FungiDQ450788, Rhodanobacter, uncultured gamma proteobacteriumEF018682, Rhodanobacter, uncultured Xanthomonadaceae bacterium
ARB_2B63D909, Rhodanobacter, Mors 00 cm No FungiHM447956, Rhodanobacter, uncultured Xanthomonadaceae bacterium
ARB_A327BBF7, Rhodanobacter, Mors 40cm No FungiARB_29CF88AD, Rhodanobacter, Roslev FilterHM238170, Rhodanobacter, uncultured gamma proteobacterium
ARB_F3B6D813, Rhodanobacter, Roslev FilterHM447914, Rhodanobacter, uncultured Rhodanobacter sp.ARB_1BDCE820, Rhodanobacter, Mors 60 cm No Fungi
JN032382, Rhodanobacter, uncultured Sporosarcina sp.JN032390, Rhodanobacter, uncultured Rhodanobacter sp.ARB_9DB40E07, Rhodanobacter, Mors 20 cm With FungiFN550151, Rhodanobacter, Xanthomonadaceae bacterium B1B
AM087226, Rhodanobacter, Rhodanobacter spathiphylliARB_CDAB6894, Rhodanobacter, Mors 00 cm With FungiARB_B5BBE50E, Rhodanobacter, Kiel FilterEU332826, Rhodanobacter, Rhodanobacter ginsenosidimutansAF039167, Rhodanobacter, Rhodanobacter lindaniclasticus
ARB_8A8B1DB, Rhodanobacter, Mors 20 cm No FungiAB100608, Rhodanobacter, Rhodanobacter fulvusARB_F3CA3687, Rhodanobacter, Mors 60 cm No FungiEF166075, Rhodanobacter, Rhodanobacter ginsengisoli
AB286179, Rhodanobacter, Rhodanobacter thiooxydansARB_59F4697F, Gammaproteobacteria_Xanthomonadales_Xanthomon, Mors 40cm No Fungi
HM163220, Gammaproteobacteria_Xanthomonadales_Xanthomon, gamma proteobacterium OR 113ARB_194A3CC2, uncultured, Mors 00cm No FungiEU440725, uncultured, uncultured Xanthomonadaceae bacterium
FJ536886, uncultured, uncultured Xanthomonadaceae bacteriumARB_72839B66, uncultured, Mors 00cm No Fungi
GU983377, uncultured, uncultured Xanthomonadales bacteriumARB_DA0A7C4E, Dokdonella, Mors 40cm With Fungi
EU685334, Dokdonella, Dokdonella sp. KIS28 6ARB_5B232D38, Dokdonella, Mors 00 cm No Fungi
ARB_45BF360A, Dokdonella, Roslev FilterARB_366BC38A, Dokdonella, Roslev Filter
AB245362, Dokdonella, Dokdonella ginsengisoliAJ969432, Dokdonella, Dokdonella fugitivaAY987368, Dokdonella, Dokdonella koreensis
ARB_C3C6DD0C, Dokdonella, Roslev Filter
Burkholderiales 39
76
Figure 3b: Tree structure of sequences from the third section (Mors) set against sequences from samples in sections one and two in other SKOV filters (Kiel filter and Roslev filter) (Kristiansen et al 2011;Kristiansen et al. 2011).
Figure 3c: Tree structure of sequences from the third section (Mors) set against sequences from samples in sections one and two in other SKOV filters (Kiel filter and Roslev filter) (Kristiansen et al 2011;Kristiansen et al. 2011).
Gammaproteobacteria_Xanthomonadales_Xanthomon 43
ARB_72F02B8F, Comamonas_1, Roslev FilterARB_D4AF61ED, Comamonas_1, Kiel FilterARB_AF66B85B, Comamonas_1, Mors 00 cm With Fungi
ARB_9255DE85, Comamonas_1, Kiel FilterARB_23467512, Comamonas_1, Mors 00 cm No Fungi
AB164432, Comamonas_1, Comamonas badiaARB_18C417C8, Comamonas_1, Roslev FilterARB_8F4C00AF, Comamonas_1, Kiel Filter
AF275377, Comamonas_1, Comamonas koreensisARB_323EFAE9, Ottowia, Mors 00 cm No Fungi
ARB_3315AF11, Ottowia, Roslev FilterHM238121, Ottowia, uncultured beta proteobacterium
ARB_33FE18E3, Ottowia, Roslev FilterHM238159, Ottowia, uncultured beta proteobacterium
AJ537466, Ottowia, Ottowia thiooxydansARB_65306BDC, Castellaniella, Mors 40cm No FungiARB_73C4AAAF, Castellaniella, Roslev Filter
ARB_A6DF3067, Castellaniella, Mors 40cm With FungiARB_ECD21331, Castellaniella, Roslev Filter
AJ005447, Castellaniella, Castellaniella defragransARB_CE53D863, Castellaniella, Kiel FilterARB_18508D58, Castellaniella, Roslev Filter
AB166879, Castellaniella, Castellaniella caeniARB_D1B8C2CF, Castellaniella, Mors 40cm With FungiU82826, Castellaniella, Castellaniella denitrificansGQ241321, Castellaniella, Castellaniella sp. MJ06
EU873313, Castellaniella, Castellaniella ginsengisoliARB_CA32D50B, Castellaniella, Mors 60 cm No Fungi
ARB_2CA444B1, Castellaniella, Kiel FilterARB_C540995E, Alcaligenaceae_2, Mors 00cm No Fungi
DQ417606, Pusillimonas_4, Pusillimonas noertemanniiARB_46D21384, Pusillimonas_4, Mors 00 cm With Fungi
GQ232740, Pusillimonas_4, Pusillimonas sp. B201ARB_ACB0764D, Pusillimonas_4, Mors 00 cm With Fungi
EF672088, Pusillimonas_4, Pusillimonas sp. DCY25TARB_70E16661, Pusillimonas_4, Roslev Filter
ARB_EEA6A41C, Herbaspirillum_3, Mors 00cm No FungiFJ812350, Herbaspirillum_3, Herbaspirillum sp. 5410S 80GQ853364, Herbaspirillum_3, uncultured beta proteobacterium
ARB_E8362000, Halothiobacillus_1, Mors 00 cm With FungiJF416645, Halothiobacillus_1, Halothiobacillus neapolitanus
JN175334, Halothiobacillus_1, Halothiobacillus neapolitanusPseudomonadales Pseudomonadaceae Cellvibrio 4
Halothiobacillus_1 3ARB_5487C2A0, Pseudomonadales_Pseudomonadaceae_Cellvibrio, Mors 20 cm With Fungi
AJ289162, Pseudomonadales_Pseudomonadaceae_Cellvibrio, Cellvibrio gandavensisCP000934, Pseudomonadales_Pseudomonadaceae_Cellvibrio, Cellvibrio japonicus Ueda107AF452103, Pseudomonadales_Pseudomonadaceae_Cellvibrio, Cellvibrio japonicus
ARB_6E51461A, Xanthomonadales_Sinobacteraceae, Mors 20 cm No FungiARB_CAAA7A62, Xanthomonadales_Sinobacteraceae, Mors 00 cm No Fungi
GQ389099, Xanthomonadales_Sinobacteraceae, uncultured bacteriumARB_9BA99A7A, Rhizobiales_1, Mors 60 cm No Fungi
FN646688, Rhizobiales_1, Mesorhizobium sp. DLS 79AY995149, Rhizobiaceae_Shinella, Shinella granuliAB187585, Rhizobiaceae_Shinella, Shinella granuli
ARB_A4615C2E, Rhizobiaceae_Shinella, Mors 00 cm With FungiJQ342936, Rhizobiales_1, Shinella kummerowiae
ARB_56606EE9, Aminobacter_1, Roslev FilterARB_38BE8567, Aminobacter_1, Mors 40cm With Fungi
JF922314, Aminobacter_1, Nitratireductor sp. WW12(2011)FQ659598, Aminobacter_1, uncultured soil bacterium
ARB_77786FE2, Phyllobacteriaceae_1, Mors 00 cm With FungiAM293857, Rhizobiales_1, Devosia subaequoris
ARB_3C98FF5, Paracoccus_5, Mors 00 cm With FungiARB_401C31D5, Paracoccus_5, Roslev Filter
Y13827, Paracoccus_5, Paracoccus alkeniferARB_28EC6DF4, Alphaproteobacteria, Mors 00 cm No Fungi
ARB_3F050EDA, Brevundimonas, Mors 20 cm With FungiEF363713, Brevundimonas, Brevundimonas lenta
CP002102, Brevundimonas, Brevundimonas subvibrioides ATCC 15264ARB_571B4E7E, Family Incertae Sedis_Rhizomicrobium, Mors 60 cm No Fungi
EU861927, Family Incertae Sedis_Rhizomicrobium, uncultured soil bacteriumAB081581, Family Incertae Sedis_Rhizomicrobium, Rhizobiales bacterium A48AB365487, Family Incertae Sedis_Rhizomicrobium, Rhizobiales bacterium Mfc52
ARB_8532476F, uncultured, Mors 20 cm No FungiARB_38416E86, uncultured, Roslev Filter
ARB_FE6087AE, uncultured, Roslev FilterARB_E2B8F2F0, uncultured, Mors 40cm With Fungi
ARB_B24C8BC1, uncultured, Roslev FilterHM238171, uncultured, uncultured Sphingobacteria bacteriumARB_D6EA2912, uncultured, Roslev Filter
ARB_E8696879, uncultured, Roslev FilterAY218571, uncultured, uncultured bacterium
FJ711766, uncultured, uncultured bacteriumJN391735, uncultured, uncultured bacterium
ARB_CE65EB8E, Isosphaera, Mors 20 cm With FungiHQ014635, Isosphaera, uncultured bacterium
CP002353, Isosphaera, Isosphaera pallida ATCC 43644DQ986200, Isosphaera, Isosphaera like str. OJF2
ARB_B33B3339, Isosphaera, Roslev Filter
Actinobacteria 48
77
Figure 3d: Tree structure of sequences from the third section (Mors) set against sequences from samples in sections one and two in other SKOV filters (Kiel filter and Roslev filter) (Kristiansen et al 2011;Kristiansen et al. 2011).
Isosphaera 5ARB_D3E0DF80, Brevibacteriaceae_Brevibacterium, Mors 20 cm With FungiY17962, Brevibacteriaceae_Brevibacterium, Brevibacterium avium
ARB_36E51754, Arthrobacter_3, Mors 60 cm No FungiGQ406811, Arthrobacter_3, Arthrobacter livingstonensisAJ640198, Arthrobacter_3, Arthrobacter stackebrandtii
ARB_DB34A464, Arthrobacter_3, Mors 00cm No FungiARB_63144188, Arthrobacter_2, Mors 60 cm With Fungi
Y15885, Arthrobacter_2, Arthrobacter rhombiEU873539, Plantibacter, Microbacterium sp. MNA2
FJ865215, Plantibacter, Microbacterium sp. CC 012309ARB_D6E23D15, Plantibacter, Mors 40cm With Fungi
AB177868, Plantibacter, Plantibacter auratusARB_875D019B, Microbacteriaceae, Mors 00 cm No Fungi
ARB_59CB4E08, Microbacterium, Mors 60 cm With FungiEF466125, Microbacterium, Microbacterium kribbense
ARB_2BAEEE3C, Microbacterium, Mors 00cm No FungiAB286021, Microbacterium, Microbacterium sediminicola
ARB_F79F9BCB, Microbacteriaceae, Mors 20 cm With FungiEF451667, Microbacteriaceae, Cryocola sp. KAR37EU861941, Curtobacterium, uncultured soil bacterium
ARB_2659BF, Curtobacterium, Mors 40cm With FungiEF587758, Curtobacterium, Curtobacterium ginsengisoli
ARB_1EC1B5AE, Luteipulveratus, Mors 60 cm No FungiAB539735, Luteipulveratus, Flexivirga albaARB_5B9A1F6E, Luteipulveratus, Mors 40cm With FungiAB245398, Luteipulveratus, Actinomycetales bacterium Gsoil 907HQ418230, Luteipulveratus, Actinomycetales bacterium SW102
GQ241687, Luteipulveratus, Dermacoccus sp. HOR6 4ARB_4421EA5D, Luteipulveratus, Mors 40cm No Fungi
AB468971, Luteipulveratus, Luteipulveratus mongoliensisARB_7EAD5756, Oryzihumus, Mors 40cm No Fungi
ARB_DAAF1BBE, Oryzihumus, Mors 60 cm With FungiAB193172, Oryzihumus, Oryzihumus leptocrescensARB_E63882F5, Mycobacteriaceae_Mycobacterium, Roslev FilterHM770865, Mycobacteriaceae_Mycobacterium, Mycobacterium terrae
ARB_4A15BB0F, Mycobacteriaceae_Mycobacterium, Mors 60 cm No FungiAY734996, Mycobacteriaceae_Mycobacterium, Mycobacterium triviale
ARB_570A9EEF, Nocardiaceae_Rhodococcus_1, Mors 60 cm With FungiARB_AF264525, Nocardiaceae_Rhodococcus_1, Mors 60 cm No Fungi
AF447391, Nocardiaceae_Rhodococcus_1, Rhodococcus aetherivoransX80625, Nocardiaceae_Rhodococcus_1, Rhodococcus ruber
ARB_33FE7908, Microlunatus, Mors 60 cm No FungiFJ807672, Microlunatus, Microlunatus soli
FN556016, Microlunatus, Microlunatus parietisEF601828, Microlunatus, Microlunatus aurantiacus
ARB_34ED6B0E, Bacteria, Mors 00 cm No FungiIncertae Sedis 10
78
Figure 3e: Tree structure of sequences from the third section (Mors) set against sequences from samples in sections one and two in other SKOV filters (Kiel filter and Roslev filter) (Kristiansen et al 2011;Kristiansen et al. 2011).
References: Feilberg, A., A. P. S. Adamsen, et al. (2010). "Evaluation of Biological Air Filters for
Livestock Ventilation Air by Membrane Inlet Mass Spectrometry." J. Environ. Qual.
39(3): 1085-1096.
Hansen, M. J., D. Liu, et al. (2012). "Application of Proton-Transfer-Reaction Mass
Spectrometry to the Assessment of Odorant Removal in a Biological Air Cleaner for
Pig Production." Journal of Agricultural and Food Chemistry 60(10): 2599-2606.
Juhler, S., N. P. Revsbech, et al. (2009). "Distribution and Rate of Microbial
Processes in an Ammonia-Loaded Air Filter Biofilm." Applied and Environmental
Microbiology 75(11): 3705-3713.
Juretschko, S., G. Timmermann, et al. (1998). "Combined molecular and conventional
analyses of nitrifying bacterium diversity in activated sludge: Nitrosococcus mobilis
and Nitrospira-like bacteria as dominant populations." Applied and Environmental
Microbiology 64(8): 3042-3051.
ARB_34ED6B0E, Bacteria, Mors 00 cm No FungiARB_A8F78006, Incertae Sedis, Mors 20 cm With Fungi
EU772915, Incertae Sedis, uncultured bacteriumARB_D50ABF5, Incertae Sedis, Mors 60 cm With Fungi
AM697593, Incertae Sedis, uncultured bacteriumARB_60CE32D9, Incertae Sedis, Mors 20 cm No Fungi
FJ367217, Incertae Sedis, uncultured bacteriumARB_30F01BA8, Incertae Sedis, Mors 60 cm With Fungi
EU460475, Incertae Sedis, uncultured bacteriumABEZ02000015, Incertae Sedis, Clostridium bartlettii DSM 16795
X73447, Incertae Sedis, [Clostridium] irregulareARB_E128CEF8, Peptostreptococcaceae, Mors 00 cm With FungiARB_5D9B2EA1, uncultured, Mors 20 cm With Fungi
EU467855, uncultured, uncultured bacteriumARB_A461393F, Clostridium_1, Mors 20 cm With FungiAM930366, Clostridium_1, uncultured bacteriumAJ458420, Clostridium_1, Clostridium butyricum
ARB_833AA23A, Clostridium_1, Mors 20 cm With FungiEU473135, Clostridium_1, uncultured bacterium
ARB_44ED5F62, Ruminococcus_1, Mors 40cm No FungiEU777253, Ruminococcus_1, uncultured bacterium
ARB_782AE5C4, Streptococcus, Roslev FilterHQ716527, Streptococcus, uncultured bacteriumARB_D49B07C4, Streptococcus, Mors 20 cm With FungiARB_DA20E7A2, Streptococcus, Mors 60 cm No Fungi
AB002519, Streptococcus, Streptococcus alactolyticusARB_D395E366, Streptococcus, Mors 20 cm With Fungi
AM183015, Streptococcus, uncultured bacteriumDI107336, Streptococcus, Streptococcus mitis
ARB_3AEE6364, Facklamia, Mors 60 cm With FungiHQ327304, Facklamia, uncultured Facklamia sp.
Y17312, Facklamia, Facklamia sourekiiARB_16DAB622, Staphylococcus_1, Mors 20 cm With Fungi
DJ491021, Staphylococcus_1, Staphylococcus sp.FJ957467, Staphylococcus_1, uncultured Staphylococcus sp.
ARB_9DD72F44, Staphylococcus_1, Mors 20 cm With FungiJN175375, Staphylococcus_1, Staphylococcus cohnii
EF512729, Staphylococcus_1, Staphylococcus cohniiARB_FCD61D51, Jeotgalicoccus_1, Mors 60 cm With Fungi
FJ386517, Jeotgalicoccus_1, Jeotgalicoccus halophilusARB_B877AE76, Armatimonadetes, Mors 20 cm No Fungi
AY218766, Armatimonadetes, uncultured bacteriumARB_D671FF6F, Armatimonadetes, Mors 20 cm With FungiJN541182, Armatimonadetes, uncultured bacteriumARB_2B15672D, Armatimonadetes, Roslev Filter
1.00
79
Kristiansen, A. (2010). Composition and metabolic significance of microorganisms in
biological airfilters. Section of Biotechnology, Chemical and Environmental
Engineering. Aalborg, Aalborg University. PhD.
Kristiansen, A., S. Lindholst, et al. (2011). "Butyric Acid- and Dimethyl Disulfide-
Assimilating Microorganisms in a Biofilter Treating Air Emissions from a Livestock
Facility." Applied and Environmental Microbiology 77(24): 8595-8604.
Kristiansen, A., K. H. Pedersen, et al. (2011). "Bacterial community structure of a
full-scale biofilter treating pig house exhaust air." Systematic and Applied
Microbiology 34(5): 344-352.
Lee, C. S., K. K. Kim, et al. (2007). "Rhodanobacter thiooxydans sp. nov., isolated
from a biofilm on sulfur particles used in an autotrophic denitrification process." Int J
Syst Evol Microbiol 57(Pt 8): 1775-1779.
Lopez-Garcia, P., F. Rodriguez-Valera, et al. (2001). "Unexpected diversity of small
eukaryotes in deep-sea Antarctic plankton." Nature 409(6820): 603-607.
Loy, A., A. Lehner, et al. (2002). "Oligonucleotide microarray for 16S rRNA gene-
based detection of all recognized lineages of sulfate-reducing prokaryotes in the
environment." Applied and Environmental Microbiology 68(10): 5064-5081.
Ludwig, W., O. Strunk, et al. (2004). "ARB: a software environment for sequence
data." Nucleic Acids Research 32(4): 1363-1371.
Pedersen, C. L., M. J. Hansen, et al. "Fungal presence in an airtreating biofilter
removing hydrogen sulfide."
Pedersen, C. L., D. Liu, et al. "Fungi explain removal of hydrogen sulfide in air
treating biofilter."
Pruesse, E., C. Quast, et al. (2007). "SILVA: a comprehensive online resource for
quality checked and aligned ribosomal RNA sequence data compatible with ARB."
Nucleic Acids Research 35(21): 7188-7196.
Schloss, P. D., D. Gevers, et al. (2011). "Reducing the effects of PCR amplification
and sequencing artifacts on 16S rRNA-based studies." PLoS One 6(12): e27310.
80
Spring, S., U. Jackel, et al. (2004). "Ottowia thiooxydans gen. nov., sp. nov., a novel
facultatively anaerobic, N2O-producing bacterium isolated from activated sludge, and
transfer of Aquaspirillum gracile to Hylemonella gracilis gen. nov., comb. nov." Int J
Syst Evol Microbiol 54(Pt 1): 99-106.
Zhu, F., R. Massana, et al. (2005). "Mapping of picoeucaryotes in marine ecosystems
with quantitative PCR of the 18S rRNA gene." Fems Microbiology Ecology 52(1):
79-92.
81
Non-biological oxidation of sulfurous compounds in an air treating
biofilter containing iron oxides.
Claus Lunde Pedersen1), Dezhao Liu2), Anders Feilberg2) and Lars Peter Nielsen1)
1) Department of Bioscience, Microbiology, Aarhus University, Ny Munkegade 116,
8000 Aarhus, Denmark; 2) Department of Engineering, Aarhus University, Blichers
Allé 20, 8830 Tjele, Denmark;
*Corresponding author: Tel.: 0045 8715 4336, Fax: 0045 8715 4326, E-mail address: [email protected]
82
Abstract:
Biotrickling filters based on clay pellets seem superior in removing sulfurous gasses
from high-flow air streams with low concentrations. The aim of this study was to test
if this could be ascribed to non-biological sulfide oxidation catalyzed by the high iron
content of the filter matrix. Clay pellets from a biofilter treating waste air from a pig
facility had an average removal efficiency of 82 %, where 56 % was ascribed as that
non-biological oxidation. At similar conditions methane thiol removal was 45 %,
where 23 % was ascribed to that non-biological oxidation. Dimethyl sulfide showed
removal efficiency of 12 % with less than 4 % non-biological oxidation corresponded
to 56 % of the H2S removal, 23 % of MT removal and 4 % of the DMS removal in the
bioactive filter. The product of non-biological oxidation, elemental sulfur, did not
accumulate, suggesting secondary biological oxidation to sulfate. These results
encourage incorporation of iron oxides in filters designed for H2S removal.
83
Introduction:
Biofiltration has been implemented to reduce odor emissions from pig facilities in
farming industry. Emissions comprise a complex matrix of volatile compounds,
where hydrogen sulfide (H2S) is considered to have the highest impact on odor
perception due to its high emission concentration relative to the low odor threshold
(Feilberg et al. 2010). Due to their low redox potential, reduced sulfurous compounds
(RSC) represents a biological energy source in aerobic environments such as a
biofilter. However the uptake of RSC may be limited by the compounds low water
solubility. Some biofilters have combined bacterial oxidation of sulfurous compounds
with iron catalyzed H2S oxidation(Pagella and De Faveri 2000; Li et al. 2005;
Goncalves and Govind 2008). Catalyzed sulfide oxidation by metals has been studied
in marine waters and sediments, where the reaction is [HS-] dependent, which again is
influenced by pH, temperature and salinity (Chen and Morris 1972; Almgren and
Hagstrom 1974; Vazquez et al. 1989; Poulton et al. 2004; Luther et al. 2011).
Oxidation of H2S leads to a drop in pH, which further complexes the reaction pathway
that is still not fully understood (Luther et al. 2011). Few studies have investigated
iron catalyzed oxidation of methane thiol (MT) (Petre and Larachi 2007). Given that
MT is present in the exhaust air from pig farms suggest that iron catalyzed MT
oxidation may also occur, however it has not been described in a biofilter before. The
aim of this study was to separate biological and non-biological oxidation of H2S and
MT in biofilter, where the filter material was made of clay pellets containing iron
oxide. Clay pellets has been tested and implemented in a number of biofilters treating
waste air containing RSC from farming industry and wastewater treatment plants (Li
et al. 2005; Ramirez-Saenz et al. 2009; Akdeniz et al. 2011). However the degree of
biological and non-biological oxidation processes has not been clarified, till now.
84
Materials and methods:
Trickling biofilter:
The filter was cylindrical shaped with a height of 2.5 m and an outer diameter of 2.2
m. Filter material was stacked at a width of 40cm leaving a void of 1.4 m in diameter
in the center from where ventilation air dissipated out through the filter material.
Light expanded clay aggregates (LECA!, Weber Saint-Gobain A/S) with a particle
size of 10 to 20 mm were used as biofilter material. The trickling biofilter was
dimensioned to treat up to 6000 m3 hour-1 of waste air, which gave a theoretical air
speed of 0.12 m s-1 in the empty filter. The filter was constantly irrigated by water,
which was circulated from a reservoir tank with a volume of 600 Liter. Water
conductivity in the reservoir tank were monitored and regulated by drainage of water
to the slurry pit followed by addition of tap water. Further details on the farm as odor
source and details on the biofilter and its operation can be found in (Lui et al, in
preparation).
Model setup:
To distinguish between biological and iron-catalyzed oxidation of sulfurous
compounds, the filter material was pasteurized to inhibit microbial activity in a small-
scale setup. A Plexiglas cylinder with an inner diameter of 7.4 cm and a length of 40
cm were filled with clay pellets from the trickling biofilter described above (se figure
1). Air speed equivalent to maximum air rate in the biofilter were calculated to be 30
L min-1 in the small-scale setup. Compressed air served as dilutant through a bubbling
flask for humidification (see figure 1), while two gas canisters (air liquid?, type
specs.?) served as odor stock. One contained a mixture (5 ppmv) of H2S, MT and
DMS in nitrogen solution and the other H2S (1000 ppmv). Three digital mass-flow
85
controllers (manufacturer, specs) kept a constant flow from the two gas canisters and
the compressed air. A Ventilation Experiment investigated the effects of varying
ventilation rate, at constant inlet concentrations. Inlet concentrations reflecting
reported concentrations from the actual biofilter (ppbv): H2S (300), MT(7) and
DMS(3), while varying ventilation rates: 0.5, 3, 6, 10, 20 and 30 L min-1. Additional
Loading Measurements evaluated the relationship between removal efficiency and
inlet loading [mmol hour-1 m-3], where H2S inlet loading rates were raised to [mmol
hour-1 m-3]; 5, 13, 23 and 38, while ventilation rate were kept at 6 or 10 L min-1. MT
and DMS concentrations were kept as in the Ventilation experiment. The
measurements were carried out with intact filter material that had been collected two
hours prior to experiments and kept at ambient temperature during transport. The
small-scale model was then pasteurized by filling it with irrigation water from the
trickling biofilter heated to 90° C. Pasteurization was kept bellow 100 °C with the aim
of inhibiting bacterial activity without disrupting the biofilm structure and irrigation
water from the trickling biofilter was used to ensure that ionic strength were not
altered during pasteurization. After 10 minutes the water was drained and the scale
model was cooled to ambient temperature by blowing compressed laboratory air
through the cylinder for 30 minutes and measuring series were repeated for
comparison in randomized order. New (unused) LECA! material, which initially
were soaked in tap water and then subjected to a flow rate of 30 L min-1 at a H2S
concentration of 300 ppbv for 30 minutes were subjected to same conditions as intact
LECA! material.
86
PTR-MS:
At each measuring point inlet and outlet concentrations were measured continuously
for 5 minutes or until measured concentrations were stable. A high-sensitivity PTR-
MS (Ionicon Analytik, Innsbruck, Austria) was selected for the experiments in this
study. The PTR-MS was run under standard conditions with the drift tube temperature
set to 60 °C, the drift tube voltage set to 600 v and the drift tube pressure controlled
within the range of 2.1 - 2.2 mbar. The E/N number was ca. 135 Td during all
experiments. The sensitivity for the measured compounds in this study was calculated
based on proton transfer reaction rate constants(de Gouw and Warneke 2007; Feilberg
et al. 2010). The transmission efficiencies were estimated through a gas mixture
cylinder containing 14 aromatic compounds with m/z (mass to charge ratio) ranging
between 21 and 181 (P/N 34423-PI, Restek, USA). Humidity dependence for
hydrogen sulfide detection was calibrated regularly during the experiments by
applying the method described by Feilberg et al. (2010).
Data analysis:
Statistical analysis and figures were made with R (version. 2.13.2) open software
(R_Development_Core_Team 2011). Student's t test and paired t test as well as
ANOVA test were made with R. Beta coefficients for ANOVA test were made by
standardizing defining variables. Removal efficiency (RE) was calculated as inlet
concentration subtracted by outlet concentration divided by inlet concentration.
Unless stated otherwise values are represented as average ± standard error of the
mean.
87
Results:
Removal efficiency (RE) for H2S with new material went from 99 % at lowest air
speed to 33 % at highest air speed (figure 2a). This removal was described as non-
biological oxidation and confirmed, when comparing results with RE on
established/pasteurized material. A student’s t test between RE on new material and
established-pasteurized material showed no significant difference for all air speeds (P
> 0.44). In both cases RE decreased for H2S with increasing air speed, while inlet
concentration remained constant. Established-pasteurized and new material showed
removal efficiencies that were 50 % lower than established material, before
pasteurization. The difference before and after pasteurization described the biological
activity. Non-biological oxidation of MT was lower than H2S. 93 % of MT was
removed at lowest air speed with new material, however RE dropped to 8 % at air
speed of 0.023 m s-1 and remained at an average of 10 % for the following air speeds.
Similar trend was observed for established-pasteurized material and a students t test
showed no significant difference in RE at various air speeds (P > 0.31) see figure 2b.
Biological activity accounted for 53 % of MT removal and 47 % was ascribed to non-
biological oxidation, when comparing RE between established and established-
pasteurized material. There was little or no removal of DMS regardless of filter
material or pasteurization treatment (see figure 2c). Removal efficiency averaged 13
% for established material, while established-pasteurized material and new material
averaged 4 %. Despite this difference in RE, there was no significant difference for
RE between established and established-pasteurized material (student's t test P >
0.15). For loading experiments (figure 3) RE for established material averaged 64 %
compared to established-pasteurized material at 31 %, indicating that non-biologically
H2S oxidation accounted for 50 % as observed in the ventilation experiment (figure
88
2a). A paired t test comparing RE between the two air speeds 0.023 m s-1 and 0.039 m
s-1 at equivalent loading rates showed no significant difference (P > 0.83).
Discussion:
The change in RE between intact and pasteurized material defined the biological
activity and at the same time verified that the pasteurization was effective. 27 % of
the removed H2S and 21 % of MT were ascribed to biological activity, when
comparing RE from new material with established material. 11 % of DMS removal
were ascribed to biological activity, however when comparing the different treatments
in figure 2c there were no clear trend, despite that there was a significant difference
between intact and pasteurized material. The difference in RE for H2S and MT
between clean material and pasteurized suggested, that there was still an effect of the
present biofilm. It could be speculated that the biofilm covered the catalyzing surface
of the filter material, since the RE was higher for clean than the pasteurized material.
There was no effect of ventilation rate measured in the loading experiment, which in
some cases have been reported as empty bed retention time (EBRT). This was
evident, when comparing RE for the two ventilation rates. Comparing RE from the
ventilation experiment at 30 L min-1 with corresponding loading rates at 6 L min-1 and
10 L min-1 in the loading experiment showed no significant difference, despite five
fold difference in ventilation rate. The correlation between RE and inlet concentration
indicated a diffusion limited removal, which have been confirmed in a number of
other biofilm studies. It is reasonable to believe that a similar relation between iron
catalyzed and biological H2S and MT removal occurs in the actual biofilter from
which the material was sampled, since air speed and inlet concentrations reflected the
actual biofilter. However the scale model was not subjected to diurnal variations in
89
temperature and other defining variable and the scale model were not subjected to the
constant load of ammonia and volatile organic compounds that the full-scale biofilter
were treating. It could be argued that since biological H2S oxidation only accounted
for 35% it would be better just to rely on chemical oxidation as it has been done in oil
industry for nearly a century. However iron catalyzed sulfide oxidation stops at
elemental sulfur, which can clog the filter. No elemental sulfur precipitation was
observed in the investigated filter, which suggests that bacteria may not be oxidizing
H2S, but the products of the iron catalyzed H2S oxidation. This is supported by a
previous study(Pedersen et al. Submitted), where it was argued that H2S concentration
under similar conditions would be too low to maintain a population of strictly sulfide
oxidizing bacteria. However a heterotrophic community of bacteria, that can
metabolize the large number of organic compounds and elemental sulfur, seems
possible. Further studies on the bacterial community may reveal this. Whether there
were a competitive inhibition between iron catalyzed MT and H2S oxidation was not
possible to confirm, since MT inlet concentration were not altered during the
experiments.
90
Tables:
Table 1: Measured flow rate during the experiments and their corresponding
flow rates in the actual biofilter.
Measured flowrate (L min-1) 0.50 3 6 10 20 30
Equivalent airspeed (m s-1) 1.94E-3 1.163E-2 2.33E-2 3.88E-2 7.75E-2 0.12
Corresponding flowrate (m3 h-1) 104.1 625.0 1250.0 2083.3 4166.7 6250.0
Figures:
Figure 1: Schematic drawing of the experimental setup with the small-scale
model (LECA Filter). A gas canister (5 ppm) with hydrogen sulfide, methane
thiol and dimethyl sulfide in a nitrogen solution served as source of MT and
DMS, while a gas canister with 1000 ppm hydrogen sulfide in nitrogen solution
served as source for hydrogen sulfide. Flow controllers adjusted inlet
concentrations and humidified lab air served as dilution for the odor
!"#$%&'()*+
,-./01/.
,-.
&(23425)+2((*+
6078/.
!9:$'+
Humidifier
&(23425)+2((*+ &(23
425)+2((*+
Exhaust
91
compounds. Inlet and outlet concentrations were recorded with a prton transfer
reaction mass spectrometer (PTR-MS). Photo insert of intact LECA material in
upper right corner.
!"#$%
#$%
"#$%
&#$%
'#$%
(#$%
)##$%
#% #*#"% #*#&% #*#'% #*#(% #*)% #*)"%
!"#$%&'(")*+,+"-,./(0
(
1+2(34""5/(#(367(
+,-./01,234% +,-./01,23456.,-3781934% :3;%
92
Figure 2: Removal efficiency plotted against ventilation rate for hydrogen sulfide
(A), methane thiol (B) and dimethyl sulfide (C), while inlet concentration were
kept constant. Intact biofilter material (diamonds), pasteurized material
(squares), new material (triangles). There was a significant lower Removal
efficiency (RE) for pasteurized material, when comparing to intact material for
all three compounds: H2S (P < 0.0001), MT (P < 0.0001) and DMS (P < 0.01).
Inlet concentrations averaged ± standard error of the mean: (H2S) 277.5 ± 6.1
ppbv, (MT) 3.87 ±0.069, (DMS) 5.26 ± 0.1.
!"#$%
#$%
"#$%
&#$%
'#$%
(#$%
)##$%
#% #*#"% #*#&% #*#'% #*#(% #*)% #*)"%
!"#$%&'(")*+,+"-,./(0
(
1+2(34""5/(#(367(
+,-./01,234% +,-./01,23456.,-3781934% :3;%
!"#$%
#$%
"#$%
&#$%
'#$%
(#$%
)##$%
#% #*#"% #*#&% #*#'% #*#(% #*)% #*)"%
!"#$%&'(")*+,+"-,./(0
(
1+2(34""5/(#(367(
+,-./01,234%
93
Figure 3: Removal efficiency plotted against inlet load (mmol hour-1 m-3) for
hydrogen sulfide. A paired t-test of RE at corresponding loading rates showed no
significant difference for H2S for intact (P > 0.9), pasteurized (P > 0.26) or clean
material (P > 0.22).
#$%
"#$%
&#$%
'#$%
(#$%
)##$%
#% <% )#% )<% "#% "<% =#% =<% &#% &<%
8.52$9"-(3:'*+5"((
2"#$%&'(")*+,+"-,./(0
(
;-'"<('$&5/(##$'(=$:267(#6>(
>?-.@-% 6.,-3781934% A03.?%
94
References:
Akdeniz, N., K. A. Janni, et al. (2011). "Biofilter performance of pine nuggets and
lava rock as media." Bioresource Technology 102(8): 4974-4980.
Almgren, T. and I. Hagstrom (1974). "Oxidation Rate of Sulfide in Sea-Water."
Water Research 8(7): 395-400.
Chen, K. Y. and J. C. Morris (1972). "Kinetics of Oxidation of Aqueous Sulfide by
O2." Environmental Science & Technology 6(6): 529-&.
de Gouw, J. and C. Warneke (2007). "- Measurements of volatile organic compounds
in the earth's atmosphere using proton-transfer-reaction mass spectrometry." - 26(-
2): - 257.
Feilberg, A., D. Liu, et al. (2010). "Odorant Emissions from Intensive Pig Production
Measured by Online Proton-Transfer-Reaction Mass Spectrometry." Environmental
Science & Technology 44(15): 5894-5900.
Goncalves, J. J. and R. Govind (2008). "H2S Abatement in a biotrickling filter using
iron(III) foam media." Chemosphere 73(9): 1478-1483.
Li, H., D. R. Luelking, et al. (2005). "Biogeochemical analysis of hydrogen sulfide
removal by a lava-rock packed biofilter." Water Environment Research 77(2): 179-
186.
Luther, G. W., 3rd, A. J. Findlay, et al. (2011). "Thermodynamics and kinetics of
sulfide oxidation by oxygen: a look at inorganically controlled reactions and
biologically mediated processes in the environment." Front Microbiol 2: 62.
Pagella, C. and D. M. De Faveri (2000). "H2S gas treatment by iron bioprocess."
Chemical Engineering Science 55(12): 2185-2194.
Pedersen, C. L., M. J. Hansen, et al. (Submitted). "Fungi explain removal of hydrogen
sulfide in air treating biofilter." Chemical Engineering Journal.
95
Petre, C. F. and F. Ø. a. Larachi (2007). "Removal of Methyl Mercaptide by
Iron/Cerium Oxide‚àíHydroxide in Anoxic and Oxic Alkaline Media." Industrial &
Engineering Chemistry Research 46(7): 1990-1999.
Poulton, S. W., M. D. Krom, et al. (2004). "A revised scheme for the reactivity of iron
(oxyhydr)oxide minerals towards dissolved sulfide." Geochimica et Cosmochimica
Acta 68(18): 3703-3715.
R_Development_Core_Team (2011). "R: A Language and Environment for Statistical
Computing." from http://www.R-project.org/.
Ramirez-Saenz, D., P. B. Zarate-Segura, et al. (2009). "H(2)S and volatile fatty acids
elimination by biofiltration: Clean-up process for biogas potential use." Journal of
Hazardous Materials 163(2-3): 1272-1281.
Vazquez, F., J. Z. Zhang, et al. (1989). "Effect of Metals on the Rate of the Oxidation
of H2s in Seawater." Geophysical Research Letters 16(12): 1363-1366.
96
Acknowledgements: I am very grateful to my supervisor Lars Peter Nielsen for guidance and inspiring
discussions. I thank all collaborators in the CEBONA project, especially Dezhao Liu
and Anders Feilberg for their support and close collaboration. For being supportive I
would like to thank Andreas Schramm and Kasper Kjeldsen in their assistance with
molecular work. In the assistance with fieldwork and lab work I would like to thank
Lise Bonne Guldberg, Anders Leegaard Riis, Tommy Andersen, Britta Poulsen and
Anne Stenteberg. Thanks to all the people at Microbiology and Geomicrobiolgy for
creating such a nice atmosphere. For invaluable help and support I would like to thank
my family and especially thanks to Laura for being so sweet and putting up with me,
while exploring and endeavoring this PhD.
I
Conference activities: CEBONA Open Workshop, Viborg 2009
" Getting poorly soluble VOSC to the cells in biofilters " (Oral presentation)
3rd International Congress on Biotechniques for Air Pollution Control, Delft 2009
(no contribution)
2010 Duke UAM Conference on biofiltrtation for air pollution control
" Surface structure of biofilm removing sulfurous compounds from air" (oral
presentation)
4th Congress of European Microbiologists- FEMS, Geneva 2011
"Properties of biofilm metabolizing hydrogen sulfide from air" (Poster)
EUROBIOFILMS, Copenhagen 2011
"Properties of biofilm metabolizing hydrogen sulfide from air" (Poster)
4th International Conference on Biotechniques for Air Pollution Control, La
Coruña 2011
"Static and dynamic measurements of sulfur gas removal in a biofilter" (oral
presentation)
14th International Symposium on Microbial Ecology -ISME 14, Copenhagen 2012
"Hydrogen sulfide oxidation in an air-treating biofilter" (Poster)
II
Curriculum Vitae: Contact information:
Work address: Department of Bioscience Aarhus University Ny Munkegade 116 8000 Aarhus Home address: Name: Claus Lunde Pedersen, Address: Grønnegade 62 4. sal, 8000 Aarhus c Phone: 61 30 17 51 Personal information:
Date of birth: June 24th 1980
Citizenship: Danish
Language skills: Danish (Native), English (Excellent), German (Basic)
Education:
2009 - 2012, Ph.D. Microbiology, Aarhus University
Thesis title: Function and structure of biofilters removing sulfur gasses
2007 - 2009, M. Sc. Zoo physiology, Aarhus Universitet
Thesis title: Effects of nitric oxide on twitch force development and oxygen consumption in myocardial tissue from trout (Oncorhynchus mykiss) and goldfish (Carassius auratus)
2004 - 2007 - B. Sc. Biology, Aarhus Universitet
2001 - 2004, Civil engineering, Medical technology, Aalborg University
1997 -2000 - Graduated from high school - Viborg Amtsgymnasium
Teaching assistant at following courses:
III
• Systems biology - Nutrient cycling, microbial activity • Terrestrial zoology, field course • Advanced zoo physiology, enzyme kinetics • Vertebrate zoology, anatomy
Courses attended during PhD studies: 2012: Bioinformatics for microbial ecologists - Mothur pipeline for data analysis,
University of Jyväskulä, Finland (2 ECTS)
2012: Workshop symbiosis imaging, University of Vienna, Austria (5 ECTS)
2010: Microsensor analysis in the environmental sciences, University of Copenhagen
and Aarhus University (5 ECTS)
2010: Practical course in 3D light and electron microscopy, ETH, Zürich and
University of Zürich, Schweiz (2 ECTS)
2010: Public Outreach, Aarhus University (5 ECTS)
2010: Scientific writing, Aarhus Universitet (5 ECTS)
2009: Electron Microscopy and Analysis for Materials Research, Technical
University of Denmark (5 ECTS)
2009: Volatile Organic Sulphur Compounds - Analytical and Physical Chemistry,
Aarhus University (5 ECTS)
IV
Abstract submitted to
CEBONA Open Workshop, Viborg 2009
Work package 2c: Getting poorly soluble VOSC to the cells in biofilters
Claus Lunde Pedersen and Lars Peter Nielsen
Department of Biological Sciences, Microbiology, Aarhus University, Ny Munkegade
116, DK-8000 Aarhus C, Denmark, E-mail: claus.pedersenbiology.au.dk
Odorous sulphur compounds also known as volatile organic sulphurous compounds
(VOSC) consist of a group of molecules, which are more or less hydrophobic. In
some filters the removal of VOSC from exhaust air have been relatively effective, but
why is not evident. The working hypothesis of the present study is that the uptake of
VOSC in the biofilm depends on the structure of the microbial community and
interactions with the filter substrate. This can be shown by the use of electron
microscopy and light microscopy combined with different staining techniques to
show surface topography and spatial distribution of bacteria and fungi in the biofilm.
The aim of the study is to contribute with knowledge on how improvements of filter
design and operation can raise the efficiency of the removal of VOSC from waste air.
V
Abstract submitted to
2010 Duke UAM Conference on biofiltration for air pollution control
Surface structure of biofilm removing sulfurous compounds from air
Claus Lunde Pedersen and Lars Peter Nielsen
Department of Biological Sciences, Microbiology, Aarhus University, Ny Munkegade
116, DK-8000 Aarhus C, Denmark, E-mail: claus.pedersenbiology.au.dk
Despite low solubility some biotrickling filters are still able to remove volatile
organic sulfurous compounds (VOSC) from the exhaust air from pig farms. Since
simple diffusion cannot explain the uptake alone, it’s been speculated that the biofilm
could either have a large surface or contain features in the surface that could
immobilize VOSC. These features could be a large concentration of active bacterial
cells in the surface layer or fungi in direct contact with the air, since fungi are less
water dependant due to their chemically composition. The biofilm surface structure
was investigated using cryo scanning electron microscopy (cryo-SEM). The images
showed little activity in the surface layer and there were no fungi present either, so
other mechanisms most explain the removal of VOSC. Sublimation techniques
indicated the presence of a hydrophobic micro porous structure in the biofilm, which
suggest that bacteria in the deeper parts of the biofilm contribute to the removal of
VOSC from the air.
Emission of offending malodorous compounds from animal farms is a nuisance for neighbors in the local surroundings and an increasing problem worldwide due to intensified animal production. Treatment of ventilation air by various types of biofilters/biotrickling filters has emerged as one of the very few competitive technologies for reduction of odor emissions from animal production facilities. However, the compounds present in farm emissions with lowest odour thresholds are volatile organic sulphur compounds (VOSC), which are only sparingly soluble in water and hardly removed by biofilters/biotrickling filters previously. This makes biological treatment challenging because of limitations in mass transfer through the air and water phase and in the biofilm. Due to difficulties in measuring low concentrations of VOSC, only limited information is available about the processes and parameters that control the removal of these compounds. Mass transfer process of sulphur compounds will be controlled by liquid phase solubility and reactions due to the low solubility of these compounds in the aqueous phase (Ref 17). Furthermore, constituents of biotrickling filter liquids that can affect air-water partitioning include ionic strength, suspended solids and acids. In order to optimize the removal efficiencies in biofilters/biotrickling filters and improve modeling process, accurate estimates of the air-liquid partitioning in biotrickling filters are needed. PTR-MS allows a low detection limit (10~200 pptv) and a fast time response (1~10 sec) when measuring odorous organic compounds in real time, and it has been becoming an essential tool for a wide variety of fields measurement and monitoring of VOC from atmosphere and animal houses (Ref 10, 12, 14, 15,16). The objectives of this study are: 1. to test if there are difference on air-liquid partitioning for interested sulphur compounds (hydrogen sulfide (H2S), methanethiol (MT) and dimethyl sulfide (DMS)) between distilled water and biotrickling liquids; 2. to test if ionic strength have influence on air-liquid partitioning for these sulphur compounds.
Methods
Determination of air-liquid partitioning of volatile organic sulfur compounds in biotrickling liquids
D. Liu, A. Feilberg*, A. P. S. Admsen, A. M. Nilsen and C. L. Pedersen Aarhus University, Denmark
Determination methods for air-liquid partitioning (KH) include calculation from the ratio of vapor pressure to water solubility and static and dynamic experimental techniques. Direct calculation is limited to those compounds for which accurate physical constants have been determined. The static technique of equilibrium partitioning in closed systems (EPICS) is based on gas chromatography (GC) and relies on the static determination of headspace concentrations. Thus sample has to be collected from an equilibrated system and adsorption of volatiles on walls, tubings or the syringe may affect the results. Dynamic determination methods, including the purge-bottle technique and the wetted-wall column, are better suited to determination of the pure water H where rapid air water equilibration can occur. For the purge-bottle technique which we adapted in this study, zero gas bubbles through a solution and strips dissolved VOSC leading to a decrease of their liquid concentration. Partition coefficient (KH) can be calculated by measuring the concomitant decrease of VOSC in the gas phase(Ref 11, 13). Fig.1 shows the set-up for the measurement in this study and Fig.2 shows an example for how the partition coefficient (KH) is gained from the model showed in formulae (1)~(3) below.
Reference 1. C.Coquelet et al., J.Chem.Eng.Data, 2008, 53, 2540-2543. 2. W.J.D.Bruyn et al., Journal of Geohyysical Research, 1995, 100(D4), 7245-7251. 3. V.P.Aneja et al., Chem.Eng.Comm., 1990, 98, 199-209. 4. M.C.Iliuta et al., J.Chem.Eng.Data, 2005, 50, 1700-1705. 5. A.Przyjazny et al., J.Chromatogr., 1983, 280, 249-260. 6. A.G.Vitenberg et al., J.Chromatogr., 1975, 112, 319-327. 7. P.Pollien et al., International Journal of Mass Spectrometry, 2003, 228, 69-80. 8. J.W.H.Dacey et al., Geophys.Res.Lett., 1984, 11, 991-994. 9. Kagaku Kogaku Ronbunshu, 1987, 13, 43-50. 10. A.Feilberg et al., Environ.Sci.Technol., 2010, 44(15), 5894-5900. 11. T.Karl et al., International Journal of Mass Spectrometry, 2003, 223-224, 383-395. 12. G.J.De et al., Mass Spectrometry Reviews, 2007, 26, 223-257. 13. P.Pollien et al., 4th International Conference on Proton Transfer Reaction Mass
Spectrometry and its Applications, 2009, 221-225. 14. C.N.Hewitt et al., Journal of Environmental Monitoring, 2003, 5, 1-7. 15. N.M.Ngwabie et al., Journal of Environmental Quality, 2008, 37, 565-573. 16. S.L.Shaw et al., Environ.Sci.Technol., 2007, 41, 1310-1316. 17. A.M.Nielsen et al., Air & Waste Manage. Assoc., 2009, 59, 155-162.
Air-liquid partitioning were measured by PTR-MS under a temperature range from approximately 3 to 45 °C. Distilled water, biotrickling liquids, and NaCl solutions were used to evaluate the effects of ionic strength and co-solutes on air-liquid partitioning. The partitioning coefficients measured from the distilled water fitted well with the values from literatures except hydrogen sulfide (e.g., see Fig.3 for DMS). However, according to the measurements in this study, no significant differences between distilled water and biotrickling liquids were observed on air-liquid partitioning for these sulphur compounds (e.g., see Fig.4 for DMS). Furthermore, no significant effect of ionic strengths up to 0.7 mol/l of NaCl were observed with a temperature of 14 °C(Fig.5).
Introduction Results
*Corresponding author, [email protected], Blichers Alle 20, DK-8830, Tjele
DMS
Conclusion According to the objectives we gave above, conclusions are: 1.Despite the relatively high ionic strength in the biotrickling filter liquids, no significant differences between pure water and filter liquids were observed on air-liquid partitioning for these sulphur compounds. 2. no significant effect of ionic strengths up to 0.7 mol/l of NaCl were observed, indicating that salting-out effects are not important for the measured sulfurous compounds within the applied temperature range. The results show that the solubility of sulfur compounds in the biotrickling filters is in the same level as in distilled water. This simplifies modeling of biotrickling filters since we can use experimental values measured in distilled water or literature values for air-water partitioning.
PTRMS
bubbling cells
Temperature control
Air flow controller
Fig.1 Measurement set-up
Fig.2 –ln(cps(t)/cps(0)) for DMS as dependent on time, yielding a partition coefficient (KH) of 6.1 MPa.
KH=6.1 MPa
DMS, T=287K, V=50 ml, f=200 ccm/min
T=287K
Fig.3 Comparison of partition coefficients between literature and this study
DMS, distilled water
Fig.5 Ionic strength dependence of partition coefficients for interested sulphur compounds
T=287K
Fig.4 Comparison of partition coefficients between distilled water and biotrickling liquids
DMS
VII
Abstract submitted to
4th Congress of European Microbiologists- FEMS, Geneva 2011
Properties of biofilm metabolizing hydrogen sulfide from air
Claus Lunde Pedersen and Lars Peter Nielsen
Department of Biological Sciences, Microbiology, Aarhus University, Ny Munkegade
116, DK-8000 Aarhus C, Denmark, E-mail: claus.pedersenbiology.au.dk
Some biofilters have been found to metabolize hydrogen sulphide and other volatile
odorous sulphur compounds (VOSC) from contaminated air streams. Given the high
hydrophobicity of VOSC’s it is very difficult to explain observed removal efficiencies
by conventional rates of metabolism and biofilm diffusion. In this study we wanted to
find out if the efficiency could be due to exceptional compression of the metabolic
activity at the biofilm surface, enhanced cell-air contact by expansion of the biofilm
surface, or some other special chemical or biological properties of the biofilm. Using
cryo scanning electron microscopy the biofilm surface was inspected to evaluate the
surface topography and density of bacterial and fungi. The results were then
compared to online odour measurements using proton mass transfer reaction mass
spectrometry (PTR-MS) to get the specific uptake of hydrogen sulphide (H2S) and
dimethyl sulphide in different areas of a sectioned biofilter treating exhaust air from
pig stables. Little microbial activity was detected in the front of the filter, where dust
was collected, however here dimethyl sulfide were taken up. The mid section was
dominated by bacteria, with 40% of the total uptake of H2S. Fungi dominated last
section, where the rest of H2S was metabolized.
From the microscopy images there were no indications of high bacterial densities or
unusual activity in the biofilm surface. However there were a correlation between the
uptake of H2S and density of aerial fungal hyphae. The later indicates that fungal mats
could be a prerequisite for efficient VOSC removal.
!"#$%&'"(&)*+),-*.*/-0$.)10-"(0"23)4-0%*5-*.*/63)7$%892):(-;"%2-&63)7$%8923)!"('$%<)=.$92>?"@"%2"(A5-*.*/6>$9>@<!
"#$%&!'%()*!+*)*,&*(!$()!'$,&!+*-*,!./*#&*(!
!"#$%"&%'(#)(*+#,-.(.%/0*#-+1+23(456"#3%2('7-,6%()"#.(0+"(
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
N*-:4)&5!!O!&*1?4(*)!8/49#-*,!-,*$?(3!*I:$%&-!$/,!<,47!@/3!&-$8#*&!L$&!&%8P*1-*)!-4!4(#/(*!7*$&%,*7*(-&!%&/(3!@,4-4(!7$&&!-,$(&<*,!,*$1?4(!7$&&!&@*1-,47*-,>!A+QRMN6D!-4!3*-!-:*!&@*1/91!%@-$2*!4<!:>),43*(!&%#9)*!AST6DE!Q:*!,*&%#-&!L*,*!-:*(!147@$,*)!L/-:!1,>4!&1$((/(3!*#*1-,4(!7/1,4&14@>E!Q4!*I$7/(*!-:*!&%,<$1*!-4@43,$@:>!4<!8$1-*,/$!$()!<%(3/!-:*!&%,<$1*!$,*$!4<!<%(3/!L*,*!1$#1%#$-*)!7/1,4&14@>!/7$3*&!%&/(3!$!#/(*!-,$(&*1-!7*-:4)E!
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
!"#$%
"#&%
$#&%
'#&%
(#&%
)#&%
*#&%
"% *% $"% $*% '"% '*% ("%
!"#$%&'()
*+'
,-.'/012"1$34501'(#*0672*8+'
"+,%!%'"+,%-./0%1234.%
'"+,%!%)"+,%-./0%1234.%
"+,%!%'"+,%35%1234.%
'"+,%!%)"+,%35%1234.%
V/3%,*!Y5!"$#1%#$-*)!ST6!@,49#*&!<4,![!17!\!T[!17!A8#%*D!$()!T[!17!\!X[!17!A,*)D!/(!W,)!&*1?4(E!O&&%7/(3!-:*!8/49#7!L4%#)!8*!$!]$-!L*-!&%,<$1*G!-:*(!ST6!L4%#)!@*(*-,$-*!!T!^7!A)$&:*)!#/(*&D!/(-4!-:*!8/49#7E!_I-*()/(3!-:*!$1?;*!8/49#7!&%,<$1*!$,*$!L/-:!-:*!<%(3$#!:>@:$*G!-:*(!ST6!@*(*-,$?4(!L4%#)!/(1,*$&*!-4!`^7!A&4#/)!#/(*&DE!!!
0
2
4
6
8
10
12
14
16
18
20
12.00 18.00 0.00 6.00 12.00 18.00 0.00 6.00 12.00
Con
cent
ratio
n, p
pbv
0
50
100
150
200
250
300
350
12.00 18.00 0.00 6.00 12.00 18.00 0.00 6.00 12.00
Con
cent
ratio
n, p
pbv
0 cm 20 cm 40 cm 60 cm
0
1
2
3
4
5
6
7
12.00 18.00 0.00 6.00 12.00 18.00 0.00 6.00 12.00
Con
cent
ratio
n, p
pbv
V/3%,*!a5!Q:*!*I@*,/7*(-$#!&>&-*7!L$&!$!-,/12#/(3!8/49#-*,!-,*$?(3!;*(?#$?4(!$/,!<,47!$!&-$8#*!L/-:!9(/&:/(3!@/3&E!V4%,!&$7@#*!&/-*&!L*,*!1:4&*(!/(!-:*!#$&-!),>!&*1?4(!4<!-:*!8/49#-*,5!A[!17b!,*)G!T[!17b!>*##4LG!X[!17b!3,**(!$()!`[!17b!8#%*E!C(#/(*!$/,!7*$&%,*7*(-&!AV/3E!TD!$()!8/49#7!&$7@#*&!L*,*!-$2*(!$-!-:*&*!X!)*@-:&E!!V/3%,*!T5!"4(?(%4%&!4(#/(*!$/,!7*$&%,*7*(-&!4<!ST6!L/-:!$!+,4-4(!-,$(&<*,!,*$1?4(!7$&&!&@*1-,47*-*,!A+QRMN6D!$-!*$1:!&$7@#/(3!&/-*E!
!
K(;*&?3$-*)!8/49#-*,5!!!!!!!!!
V/3%,*&!W!$()!X5!",>4!&1$((/(3!*#*1-,4(!7/1,4&14@>!/7$3*&!4<!-:*!8/49#7!&%,<$1*!A#*cD!$()!1,4&&M&*1?4(!A,/3:-D!L:*,*!-:*!$/,&-,*$7!*(-*,&!-:*!9#-*,!@$)!A[!17DE!6@4,*M<4,7/(3!<%(3$#!:>@:$*!14;*,!-:*!%()*,#>/(3!8$1-*,/$#!8/49#7E!
IX
Abstract submitted to
EUROBIOFILMS, Copenhagen 2011
Functional properties of biofilm removing hydrogen sulfide from air
Claus Lunde Pedersen and Lars Peter Nielsen
Department of Biological Sciences, Microbiology, Aarhus University, Ny Munkegade
116, DK-8000 Aarhus C, Denmark, E-mail: claus.pedersenbiology.au.dk
A biofilm removing hydrogen sulfide (H2S) from pig farm air was investigated.
Despite low water solubility H2S was taken up and metabolized at a constant rate.
Simple diffusion alone could not explain the uptake and it was speculated if physical
features in the biofilm surface could affect the uptake or immobilization of H2S. The
biofilm surface structure was investigated using cryo scanning electron microscopy
(cryo-SEM). The images showed a bacterial biofilm surface that in some areas were
covered with fungi. Modeled uptake of H2S indicates, that fungi may play a vital role
in the uptake of H2S since they produce aerial hyphae that increase the active surface
area of the biofilm. Surprisingly, methylated sulfur compounds, such as methane thiol
and dimethyl sulfide were not taken up in the same biofilm.
X
Abstract submitted to 4th International Conference on Biotechniques for Air Pollution Control, La
Coruña 2011
Static and dynamic measurements of sulfur gas removal in a biofilter Claus Lunde Pedersen1, Dezhao Liu2, Anders Feilberg2 and Lars Peter Nielsen1
1) Department of Bioscience, Microbiology, Aarhus University, Ny Munkegade 116,
8000 Aarhus, Denmark; 2) Department of Biosystem Engineering, Faculty of Agricultural Sciences, Aarhus
University, Blichers Allé 20, Tjele, 8830-DK
Removal of organic sulfurous gasses and hydrogen sulfide (H2S) was investigated. A
comparison was made between removal rates of in a full-scale biofilter treating waste
air from a pig stable and biofilm incubations from same biofilter. Small samples of
biofilm attached to biofilter material (20 g ±2) were sampled and incubated in 8 liter
tedlar® bags with a mixture of humidified air and sulfur gasses near in situ
concentrations, (230 ppb H2S, 40 ppb methane thiol (MT) and 40 ppb dimethyl
sulfide (DMS)). Incubations lasted 30 minutes or until the sulfurous gasses had been
reduced by 90%,while bag air was analyzed online using proton transfer reaction
mass spectrometry (PTR-MS). PTR-MS was also used to measure H2S, MT and DMS
in the full-scale biofilter at same depths of the filter as sampling for bag incubations
took place. H2S degradation was significant higher in bag incubations and in situ
measurements compared to MT and DMS degradation rates, which were almost
absent in the full-scale biofilter. Both MT and DMS had small degradation rates in the
bag incubations. Overall the removal rates of the gasses were significantly lower in
bag incubations than in situ despite the careful preservation of biofilm structure, gas
composition, humidity, and air flow. This suggests that some yet unidentified factors
could be important for efficient gas filtration.
XI
Abstract submitted to 14th International Symposium on Microbial Ecology -ISME 14, Copenhagen
2012
Hydrogen sulfide oxidation in an air treating biofilter
Claus Lunde Pedersen1) Dezhao Liu2), Michael Jørgen Hansen2), Anders Feilberg2),
Andreas Schramm1), Louise Helene Soegaard Jensen4), Lise Bonne Guldberg3) and
Lars Peter Nielsen1)
1) Department of Bioscience, Microbiology, Aarhus University, Ny Munkegade 116,
8000 Aarhus, Denmark; 2) Department of Engineering, Aarhus University, Blichers
Allé 20, 8830 Tjele, Denmark; 3) Centre for Electron Nanoscopy, Technical University
of Denmark, Fysikvej, building 307, 2800 Kgs. Lyngby, Denmark; 4) SKOV A/S,
Department of Research and Development, Hedelund 4, 7870 Glyngøre, Denmark
Observed H2S uptake rates in a biofilter treating waste air from a pig farm were too
high to be explained within conventional limits of H2S solubility, diffusion in a
biofilm and bacterial metabolism. 16S and 18S clone libraries from the biofilter found
no sulfide oxidizing bacteria but several fungal families including trichocomaceae. A
positive correlation was found between presence of sporulating fungi and H2S uptake.
However there have been no reports on fungi metabolizing hydrogen sulfide. We
hypothesize that iron containing enzymes such as cytochrome c oxidase catalyze
sulfide oxidation in a process uncoupled from energy conservation.
!"#$%&'()*%&+,%)-(%-(+(.%)$%#,'/!"#$%&'$()*&+*)*,%*(-.&/*01#2&'3$4.&5361#*"&78,9*(&:#(%*(4.&;()*,%&<*3"=*,94.&'2$3%*&
:*"*(*&>2*9##,)&7*(%*(?.&'3%*&@2((*&A$")=*,9B.&;(),*#%&>61,#CC-&D&'#,%&+*E*,&
F3*"%*(-&&&G&!"#$%&'()(*%(+,-./"/01&#$&)23&45+6*/)$76./+84
:H),29*(&%$"I3)*&3%&#&C#J2,&2)2,&62CK2(*(E&3(&
L*(E3"#E32(,&M,2C&K39&M#,C%&)$*&E2&3E%&62(6*(E,#E32(&
#()&"2N&2)2,&E1,*%12")O&&P,36Q"3(9&=32I3"E*,%&1#%&=**(&
3CK"*C*(E*)&N3E1&E1*C&E2&,*)$6*&2)2,%O&>$"I3)*&
2R3)#E32(&1#%&9*(*,#""H&=**(&#%6,3=*)&E2&=#6E*,3#"&
C*E#=2"3%C.&12N*L*,&#E&E1*&2=%*,L*)&62(6*(E,#E32(%&
%$"I3)*&6#(&2("H&)3MM$%*&#&M*N&SC&3(E2&E1*&=32I3"CO&P1*&
#3C&2M&E13%&%E$)H&N#%&E2&)*E*,C3(*&N1361&C*61#(3%C%&
E1#E&3(I"$*(6*&%$"I3)*&2R3)#E32(O&
!"#$%&'()*+,-$'+./(0(1-2#$.3-*.(45(6+4'7+-*7-8(9"#$%&'()*+,-$'+./(0(1-2#$.3-*.(45(:*;+*--$+*;8(<=-7%*+7#>()*+,-$'+./(45(1-*3#$?8(@-*.$-(54$(:>-7.$4*(A#*4'742/8(B1-2#$.3-*.(45(C-'-#$7%(#*D(1-,->423-*.8(EFGH("IE
9&4:'/+6#+(/$%47;(<.7#"4%$"=.)(4/>.)#6./+&<3"E*,&C#E*,3#"&62(E#3(3(9&1391&#C2$(E%&2M&3,2(&2R3)*%&
%12N*)&&E1#E&TUV&2M&E1*&%$"I3)*&2R3)#E32(&3%&)$*&E2&3,2(&
62(E*(E&3(&E1*&I3"E*,O&P1*,*&N#%&(2&)3MM*,*(6*&=*EN**(&(*N&
#()&K#%E*$,30*)&C#E*,3#"O
?&4!/+6*/""()4-#76(*.#"4%$"=.)(4<(6#-/".%<8/*%K3E*&62(%E#(E&K,*%*(6*&2M&%$"I3)*.&K,*%*(6*&2M&
%$"I3)*&2R3)303(9&=#6E*,3#&62$")&(2E&=*&)*E*6E*)&=H&
+!W&2(&>XY&9*(*%&#()&9*(*,#"&-Z>&CWF;O&P13%&
62,,*%K2()%&N3E1&2$,&#%%$CKE32(.&E1#E&%$"I3)*&
62(6*(E,#E32(%&#,*&E22&E2&"2N&E2&%$KK2,E&#&%$"I3)*&
2R3)303(9&=32I3"CO&
@&4:'/+6#+(/$%4%$"=.)(4/>.)#6./+4.+4#(*.#"4A$+0#"4;1';#(8+,*%*(6*&2M&M$(9#"&#*,3#"&1HK1#*&3(&=32I3"E*,&%39(3I36#(E"H&,*)$6*)&
%$"I3)*&*C3%%32(%O&;*,3#"&1HK1#*&3(6,*#%*)&E1*&62(E#6E&%$,M#6*&
=*EN**(,&#()&=32I3"CO&>$"I3)*&N*,*&2R3)30*)&3(&E1*&1HK1#*&=H&
9*(*,#"&*$Q#,H2E36&K#E1N#H%.&N13"*&%$"I3)*&62(6*(E,#E32(%&N*,*&
(2E&1391&*(2$91&E2&3(13=3E&M$(9#"&,*%K3,#E32(O&
<39$,*&4O-[&<$(9#"&#*,3#"&1HK1#*&K,*%*(E&
2(&6*""$"2%*&C#E*,3#"&3(&E1*&=32I3"E*,[
<39$,*&4O?[&!,H2G>\5&3C#9*&2M&E1*&
1HK1#*&&K,*%*(E&3(&E1*&=32I3"E*,O
<*]^^^_`
<*]^^_`
:4>&`&4X4&&&&&&&&&&&>XB4G&`&4:`
;3,
<3"E*,&- <3"E*,&4 <3"E*,&?
<39$,*&-O-[&>61*C#E36&),#N3(9&2M&#&?&%E#9*&E,36Q"3(9&=32I3"E*,O&<$(9#"&
#*,3#"&1HK1#*&N*,*&2=%*,L*)&2(&E1*&),H&?,)&I3"E*,O
<39$,*&BO4[&F*N"H&62""*6E*)&I3"E*,&
C#E*,3#"&62(E#3(3(9&3,2(&2R3)*%&#%&
(2E*)&=H&E1*&,*)&62"2,#E32(O&@32I3"C&
3%&K,*%*(E&2(c,*9#E*%&2(&E1*&
,391E
<39$,*&BO-[&!#"6$"#E*)&
,*C2L#"&*MI363*(6H]W\_&
K"2EE*)	#3(%E&"2#)3(9O&
>#CK"*)&I3"E*,&C#E*,3#"&
62(E#3(3(9&3,2(&2R3)*%&
M,2C&#&,$((3(9&=32I3"E*,&
C*#%$,*)&3(&#&%C#""&%6#"*&
%*E$K]@"$*_.&%#C*&
C#E*,3#"&K#%E*$,30*)
]W*)_&62CK#,*)&N3E1&
(*N&C#E*,3#"]A,**(_O&
B&4!/+7"$%./+8P13%&%E$)H&62(6"$)*%.&E1#E&K,*%*(6*&2M&3,2(&6#E#"H0*&
%$"I3)*&2R3)#E32(.&%K**)3(9&$K&E1*&,*#6E32(&=H&
%*L*,#"&2,)*,%&2M&C#9(3E$)*O&a1*E1*,&E1*&3,2(&3%&
=32"2936#""H&=2$()&2,&K,*%*(E&3(&E1*&I3"E*,&C#E*,3#"&
2)2,&,*C2L#"&3%&2M&(2&3CK2,E#(6*&#%&"2(9&#%&62(E#6E&
E2&E1*,&3%&Q*KEO&
<39$,*&4O4[&>$"I3)*&W*C2L#"&*MI363*(6H&]W\_&
K"2EE*)	#3(%E&L*(E3"#E32(&,#E*&3(&E1*&%#C*&
=32I3"E*,&N3E1&M$(93&K,*%*(E]@"$*_&#()&
#=%*(E]A,**(_O&!2(6*(E,#E32(&N*,*&
C*#%$,*)&62(E3($2$%"H&3(&M$""G%6#"*&
=32I3"E*,.&N13"*&K,*%*(6*&2,&#=%*(6*&2M&
M$(9#"&#*,3#"&1HK1#*&N*,*&,*62,)*)O
!"#
$!"#
%!!"#
!# %$# &!# '$#
!"#$%
$
&'()*$(+,-#..+($/+0123$.24$
()*+,*#
-+.*/01(2/3#
)/4#
0%
50%
100%
0 20000 40000
RE, %
Ventilation, M3 hour-1
With Fungi No Fungi