Desalination and Water Treatment

ISSN: 1944-3994 (Print) 1944-3986 (Online) Journal homepage: http://www.tandfonline.com/loi/tdwt20

The effect of UV pre-treatment on biofouling of

BWRO membranes: A field study

Tali Harif , Hila Elifantz , Eli Margalit , Moshe Herzberg , Tovit Lichi & Dror
Minz

To cite this article: Tali Harif , Hila Elifantz , Eli Margalit , Moshe Herzberg , Tovit Lichi & Dror
Minz (2011) The effect of UV pre-treatment on biofouling of BWRO membranes: A field study,
Desalination and Water Treatment, 31:1-3, 151-163, DOI: 10.5004/dwt.2011.2377

To link to this article: http://dx.doi.org/10.5004/dwt.2011.2377

Published online: 03 Aug 2012.

Submit your article to this journal

Article views: 93

View related articles

Citing articles: 2 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=tdwt20

Download by: [Ryerson University Library] Date: 04 June 2016, At: 18:39
 Desalination and Water Treatment 31 (2011) 151–163
July
www.deswater.com
1944-3994 / 1944-3986 © 2011 Desalination Publications. All rights reserved.
doi: 10.5004/dwt.2011.2377

The effect of UV pre-treatment on biofouling of BWRO membranes:

A field study

Tali Harifa*, Hila Elifantzb, Eli Margalita, Moshe Herzbergc, Tovit Lichia, Dror Minzd
a
Atlantium Technologies Ltd. Beit Shemesh, 99100, Israel
Tel. +972 (2) 9925001 ext. 229, +972 (54) 6490227; Fax +972 (2) 9925005; email: [email protected]
b
Department of Plant and Environmental Sciences, The Institute of Life Sciences, The Hebrew University of Jerusalem,
Downloaded by [Ryerson University Library] at 18:39 04 June 2016

Givat Ram, Jerusalem, 91904 Israel

c
Ben Gurion University of the Negev, The Zuckerberg Institute for Water Research, Sede Boqer Campus,
Midreshet Ben-Gurion, 84990 Israel
d
Institute for Soil, Water and Environmental Sciences ARO, Volcani Research Center P.O. Box 6, Beit Dagan, 50250 Israel

Received 2 August 2010; Accepted 24 March 2011

ab s t r ac t
The use of biocides, particularly chlorine, in reverse osmosis (RO) desalination is widely practiced
despite documented evidence that although biocides may be advantageous in controlling micro-
bial counts in the water, in some cases they can actually exacerbate biofouling of the membranes.
The adverse effects, associated with widespread biocide use, have spurred the need for finding
alternative RO pre-treatment disinfection methods. Ultraviolet (UV) disinfection, and specifically
medium pressure ultraviolet (MP-UV) disinfection has been considered a possible alternative and
is now gaining recognition as a viable disinfection method applicable to RO desalination; however
documentation on the effects of UV pre-treatment on RO membrane biofouling is scarce. This
paper reports the findings from a four month field study conducted at a brackish water reverse
osmosis (BWRO) desalination plant, treating groundwater, in the north of Israel, in which MP-UV
was applied as a pre-treatment disinfection step prior to RO desalination. The plant contains two
double stage desalination trains that operate in parallel — one train served as a reference, while in
the other an Atlantium HOD™ MP-UV system was installed. Both trains were run in parallel, and
for the duration of the study, all normalized performance parameters were collected and microbial
counts monitored. At the end of the run previously replaced sacrificial membranes, situated in the
front of the first stages, were autopsied and various biofilm analyses were conducted to elucidate
cell/extracellular polymeric substances (EPS) content and microbial speciation. The overall results
suggest that MP-UV pre-treatment prolonged the train performance, which manifested itself in a
lower relative normalized permeate flux decline vs. the train which received water without MP-UV
pre-treatment (11% vs. 17%, respectively). Significantly less EPS was found on the RO membrane
which received MP-UV treated water. The differences in biofilm thickness and cell density counts
(cells/cm2) between the two membranes were notable, in favor of UV pre-treatment, yet less signifi-
cant. The MP-UV pre-treatment also had a substantial effect on biofilm community composition;
the RO membrane that received MP-UV disinfected water exhibited a biofilm in which the diversity
was reduced by 30% and more, and did not contain certain phylogenetic groups that were detected
on the RO membrane that received water without MP-UV pre-treatment. It can be concluded that

* Corresponding author.

Presented at EuroMed 2010 — Desalination for Clean Water and Energy: Cooperation among Mediterranean Countries of Europe and
MENA Region, 3–7 October 2010, David Intercontinental Hotel, Tel Aviv, Israel. Organized by the European Desalination Society.
 152 T. Harif et al. / Desalination and Water Treatment 31 (2011) 151–163

pre-treatment disinfection using MP-UV may be a promising option for combating biofouling of
RO membranes and prolonging operation of the trains between cleaning regimes.
Keywords: Ultraviolet; Reverse osmosis; Membranes; Biofouling; EPS; Flux

1. Introduction of many RO membranes. It does not take into account

permeation velocity influences and additional factors
1.1. Biofouling of RO membranes
relevant to bacterial deposition, such as extracellular
Desalination has gained popularity as a technology for polymeric substances (EPS), excreted by bacteria, which
drinking water production, in an era where depletion of are known to play a primary role in bacterial adhesion
fresh water sources is on the global agenda. The advances to surfaces [12,14]. Following initial attachment, EPS
in membrane technology which include development of secretion is increased, enabling the formation of a robust
membrane materials and membrane elements that are polymeric matrix binding the bacteria to the surface and
able to produce potable water cost effectively have made shielding them from exterior dangers such as shear forces
Downloaded by [Ryerson University Library] at 18:39 04 June 2016

desalination a viable solution [1]. Nonetheless, reverse os- and biocides. In effect, the EPS determines the immediate
mosis (RO) desalination is prone to operational challenges conditions of life of biofilm cells, enabling the sustainment
which include fouling of the membrane elements. Among of a highly organized micro-environment [15]. It has been
the types of fouling existing in RO desalination, biofoul- suggested that the components of the biofilm layer, each
ing is a major topic of concern due to its crippling effect induce permeate flux decline differently. The EPS matrix
on operation of RO plants and the poor understanding in which the cells are embedded increases hydraulic re-
of the mechanisms leading to its formation [2–4]. Para- sistance to permeate flow while deposited bacterial cells,
doxically, the operational protocols implemented by RO exhibiting a more porous structure, hinder back diffusion
plants to address mineral fouling, by using antiscalants, of salt from the membrane surface, a phenomenon termed
may actually exacerbate biofouling of the membranes [5]. “biofilm-enhanced osmotic pressure” [8,14,16]. Hence,
The adverse effects biofouling causes include membrane a biofilm exhibiting less EPS should maintain a higher
flux decline, increased differential pressure and feed pres- permeate flux.
sure, membrane biodegradation, increased salt passage,
and decrease in boron rejection [3,4,6–9]. These lead to a 1.2. Microbial communities in environmental biofilm
decrease in module lifetime due to cleaning procedures
The microbial communities comprising biofilm have
and more frequent replacement costs, production loss,
been determined in recent years in a variety of environ-
product quality loss, and increased energy costs.
ments. In the marine environment, bacteria affiliated with
RO plants that implement microbial monitoring would
Alphaproteobacteria are the dominant bacterial group
use viable microbial counts in the feed water as indicators
followed by Gammaproteobacteria and Bacteroidetes [17–20].
of the biofouling potential of the water [10] although it
In freshwater biofilm the dominant group is Bacteroide-
has been hypothesized that only a few initial colonies on
tes, followed by various members of the Proteobacteria
the membrane surface suffice to initiate a mature biofilm
[21,22]. Biofilm that formed on MBR membranes and
[9]. This highlights the complexity of biofouling mecha-
on RO membranes in wastewater treatment facilities
nisms, as viable microbial counts in the feed water cannot
is composed of Bacteroidetes and various members of
indicate the susceptibility of RO desalination systems to
the Proteobacteria also, and their relative abundance is
undergo biofouling, and to what severity. Initial cell at-
depended on the flux rate, duration time and membrane
tachment and micro-colony formation occurs in two major
type [23,24]. The most dominant bacterial group in the
stages — the first stage, transport and attachment, takes
biofilm formed on an RO membrane, treating fresh sur-
minutes to hours and is governed by physicochemical
face water, is affiliated with the Sphingomonas genus of the
interactions deriving from operating conditions, water
Alphaproteobacteria [25] while additional bacteria in that
chemistry/quality, temperature, membrane properties
biofilm are affiliated with Beta- and Gammaproteobacteria,
and membrane module type. The second stage, takes days
Bacteroidetes, Planctomycetes and Verrucomicrobia [26].
to weeks, depending on nutrient availability, hydrody-
namic conditions and initially deposited cells [11,12]. The
1.3. Biocide application in RO desalination plants
classical Derjaguin–Landau–Verwey–Overbeek (DLVO)
theory has been applied to explain initial bacterial attach- Combating biofilm formation in RO plants is based
ment onto the membrane surface, with attachment found primarily upon oxidizing agents that are injected prior
to be most favorable on hydrophobic, non-polar surfaces to the membrane modules. Biocidal efficacy is measured
[13]. However, the DLVO theory fails to support deposi- in terms of percentage kill where 99.9% is considered a
tion on hydrophilic surfaces, which are representative guideline [10]. The most common biocides include chlo-
 T. Harif et al. / Desalination and Water Treatment 31 (2011) 151–163 153

rine, chloramines, chlorine dioxide and ozone [27]. Being that microorganism spectral sensitivity peaks at 254
oxidizing agents, these must be scavenged prior to the nm, may actually be erroneous [35]. The superiority of
RO membranes as the polyamide matrix that comprises MP-UV lamps has also been proven in repressing repair
the membrane is sensitive to oxidation, moreover, most mechanisms [33,36,37]. In certain circumstances repair
thin-film composite spiral wound membranes are also mechanisms have been found to exist leading to reactiva-
sensitive to certain chlorinated by-products that may tion of the microorganisms. In general, it is thought that
form in the treated water. Therefore, in practice, a re- photo-reactivation utilizes the enzyme photolyase by us-
sidual is not maintained in proximity to the membrane ing the energy of near-UV light (310–480 nm) [38] while
surface hence after-growth cannot really be inhibited at another reactivation mechanism occurs independent of a
the membrane surface. The use of biocides, specifically light source [33]. In essence, the ability of a microorgan-
chlorine, is controversial, from a process perspective, for ism to undergo reactivation will depend on the amount
various reasons, including 1) the formation of by-product of irradiation the microorganisms were exposed to and
carcinogenic compounds such as trihalomethanes and on the type of UV lamp used [36]. A successful design of
halo-acetic acids 2) the need to introduce reducing agents a UV pre-treatment system, like any other biocide appli-
into the system, i.e. sodium bisulphite, that both serves cation design, must consider both dosage requirements
as a nutrient and creates anaerobic conditions on the and pre-treatment location.
Downloaded by [Ryerson University Library] at 18:39 04 June 2016

membrane surface ideal for supporting specific bacterial This paper presents the results from a full scale field
communities 3) the increase in assimilable organic carbon study in which the effect of MP-UV was evaluated as
(AOC) that occurs due to oxidation of refractory organic a viable technology for reducing biofilm formation in
substances rendering them biodegradable [10]. Despite BWRO membranes. Both membrane performance and
the wide use of biocides, RO biofouling is common in biofilm characteristics, including microbial community
operation of RO desalination plants as after-growth can speciation, were analyzed to form an assessment of the
easily occur while only a few bacteria that reach the application of MP-UV prior to RO desalination. To date,
membrane can initiate colonization and rapidly develop to the authors’ knowledge, this has never been conducted,
into micro-colonies, depending on nutrient availability hence the findings can be considered novel. Nonetheless,
and temperature [9,28]. the authors are aware that the study is preliminary and
further work is required to acquire a deeper understand-
ing of this subject.
1.4. Ultraviolet irradiation as a pre-treatment option to de-
salination
The use of ultraviolet (UV) irradiation as a biocide in
RO desalination processes has scarcely been documented. 2. Materials and methods
Although UV is mentioned as an alternative biocide that
2.1. The BWRO desalination pilot site
can be introduced in a pre-treatment scheme [27], it is still
not implemented as a widespread anti-biofouling tech- The site used for this study is a small scale brackish
nology most likely because it does not provide a residual groundwater desalination plant that produces approxi-
effect. However, research has shown that UV irradiation mately 90 m3/h desalinated water, situated in the north of
has an adverse effect on bacterial recovery and after Israel. The water quality is moderately saline with TDS
growth on RO membranes [29]. Other authors reported values between 3000–4000 mg/l. Basic chemical param-
positive results when using UV prior to microfiltration eters of the water are summarized in Table 1.
(MF) membranes [30]. The plant operates at constant feed pressure (varying
Exposure to UV results in damage to the nucleic acids permeate production) and contains two identical desali-
of the microorganisms [31], which subsequently damages nation trains in parallel — each train contains two stages.
their ability to replicate. The use of MP-UV lamps, vs. The first stage is comprised of seven pressure vessels, each
traditional low pressure UV (LP-UV) lamps, has become housing six membranes and the second stage is comprised
in recent years more popular, due to the superior degree of two pressure vessels, each housing six membranes.
of photo-inactivation attainable for equivalent germicidal The membrane elements installed are 8” FILMTEC RO
doses [32,33]. MP-UV lamps emit polychromatic light membranes type BW30-400. Each train receives 65 m3/h
comprising UV-A (320-400 nm), UV-B (290–320 nm) and (the plant consumes a total of 130 m3/h). Train 1 (RO1)
UV-C (190–290 nm) wavelength ranges, while the LP-UV produces approximately 46 m3/h and train 2 (RO2) pro-
lamps emit only a monochromatic light at 254 nm. It is duces approximately 50 m3/h. The total recovery rate of
thought that the additional wavelengths emitted affect the plant is approximately 70%. Table 2 summarizes the
other biological molecules, not only nucleic acids, hence operating values of the two desalination trains.
leading to a greater inactivation impact [31,33,34] More- The source brackish water undergoes pre-treatment
over, the sensitivity of microorganisms can be wavelength that incorporates sand filtration following 5 mm cartridge
dependent and using monochromatic light, assuming filtration. An antiscalant is injected prior to the RO mem-
 154 T. Harif et al. / Desalination and Water Treatment 31 (2011) 151–163

Table 1 Table 2
Chemical analysis of BWRO plant source water Basic operating values of RO1 and RO2

Chemical parameter Concentration (mg/l) Parameters RO1 RO2

Boron 0.428 Operating pressure, m (Atm) 195 (19) 215 (21)
Calcium 184 Permeate conductivity, ms/cm 150 200
Magnesium 138 Permeate flow rate, m3/h 46 50
Iron 0.8 Minimal permeate flow rate, m3/h 42 40
Manganese 0.04
Potassium 41
Sodium 757 2.1.1. Preparation of the BRWO plant for the study
Silica 8.73 Two new 8” RO membrane elements, FILMTEC type
Chlorides 1906 BW30-400, were replaced in the first desalination stage,
Total dissolved solids 3740 one in each train — at the front of the central pressure
Total alkalinity 315 vessel, in identical parallel locations. These membrane
Downloaded by [Ryerson University Library] at 18:39 04 June 2016

Total organic carbon 0.7 elements were intended to serve the biofilm analyses,
pH 7.88 thereby termed “sacrificial” and designated to be re-
moved at the end of the run.
An Atlantium HODTM MP-UV system, type R-200 DL,
was installed in RO2 prior to the high pressure pump,
while RO1 served as a reference and received feed water
branes; biocides are not used in the daily operation of without disinfection. Fig. 1 shows the layout of RO2.
the plant. Various components were installed to enable manual

Stage 2
Membranes

S1-BPI S2-PPI S2-BPI

Stage 1
Membranes

Multimedia Filter Normally Open Pneumatic Valve

Normally Closed Pneumatic Valve

S1-PPI Diaphragm Proportional Valve

Micron Atlantium
Filter HODTM Normally Open Ball Valve
5 µm UV MP F- PI
F

Normally Closed Ball Valve

system
F Sampling Valve

Check Valve

Peristaltic Pump

Indicator Transmitter Device I-42

F Flow Meter F
Drain
Pressure Gauge

Fig. 1. The layout of RO2, including the HOD MP-UV system. The light blue membrane represents the position of the sacrificial
membrane in the train.
 T. Harif et al. / Desalination and Water Treatment 31 (2011) 151–163 155

Table 3 2.2. UV system operation and dose settings

Cleaning regimes and materials
To establish the applied dose of the MP-UV system
Cleaning regime Materials pH (also referred to as fluence), biodosimetry tests on the
source water were conducted using HPC and Pseudomo-
Basic EDTA, NaOH, detergent, 12 nas spp. counts as microbial indicators of inactivation
sodium lauryl sulphate
efficiency. UV exposure tests were performed according
Acidic Citric acid 2% 2 to the standard Collimated Beam Apparatus (CBA) test
protocol, as described by [39]. The CBA was constructed
by Atlantium Technologies and is shown in Fig. 2. The
monitoring and calculation of the train normalized per- methodology is based on the Ultraviolet Disinfection
formance. Aseptic sampling valves were installed before Guidance Manual for the Final Long Term 2 Enhanced
and after the MP-UV system for microbial sampling of Surface Water Treatment Rule [40].
the feed water. RO1 exhibits a similar layout, excluding
the MP-UV system and sampling valves. 2.2.1. Preparation of microbial cultures for biodosimetry
To clean the RO trains prior to the run, two cleaning testing
Downloaded by [Ryerson University Library] at 18:39 04 June 2016

cycles were implemented (Table 3). The first comprised

of a caustic wash, including chelating agents, the second Water samples were collected from the desalination
consisted of a 24-h mild acidic soak (permeate was used- plant, at the inlet to the RO trains, after pre-treatment fil-
for the wash cycles). tration, in laboratory grade sampling bottles under aseptic
Prior to the run, the trains were operated to wash the conditions and stored on ice. Prior to the CBA procedure
residual chemicals from the membrane surface. all samples were filtered through 0.45 mm nitrocellulose
membranes (Millipore, EZ Pac, white gridded 0.45 mm,
2.1.2. Data acquisition and RO train performance analysis 47 mm), using a Pall Gellman 47 mm magnetic filter
funnel (300 ml). Filters were washed with sterile saline
The BWRO plant selected for this study was not fully solution (0.9% w/v NaCl) prior to the sample filtration.
automated, hence manual readings were conducted ap- Following filtration of the samples, the filters were placed
proximately four times a week, to enable close monitoring into Petri dishes containing Standard Methods agar or
of the trains performance. Table 4 summarizes all the data Cetrimide agar — for selection of HPC, and Pseudomonas
collected manually from both trains required for perfor- spp. cultures respectively. The Petri dishes were incubated
mance normalization. The normalization software used for 72 h at 30°C, for both selected cultures. Following
was FTNORM by DOW. incubation the colonies on each filter were counted and
In addition, every two weeks, RO1 inlet, RO2 inlet the microbial concentration was calculated as CFU/ml or
and MP-UV system outlet water samples were taken for CFU/100 ml (CFU-colony forming unit). Mixed cultures of
microbial count analysis to evaluate heterotrophic plate the select “wild type” microorganisms were prepared by
counts (HPC) and Pseudomonas spp. counts, and to moni- re-growing in LB broth (Neogen Corporation; Acumedia)
tor the MP-UV system’s performance integrity. The run (30°C, 72 h) and then re-suspending in phosphate buffer
was conducted over a period of four months (with no saline (PBS) (Na2HPO4 1.15 g; KH2PO4 0.2 g; KCl 0.2 g;
additional cleaning performed for its entire duration), at NaCl 8 g; pH 7.4 in 1 L of RO water) solution.
the end of which the sacrificial RO membrane elements
were dismantled and autopsied.
2.2.2. Biodosimetry testing using the collimated beam
apparatus

Table 4
The CBA used for all biodosimetry testing at Atlan-
Data collected for performance normalization from RO1 and tium was constructed according to the EPA guidelines
RO2 [40] Fig. 2 shows the basic concept of a CBA unit.
The microbial indicators that were chosen for the CBA
1st stage data collection 2nd stage data collection procedure were HPC and Pseudomonas spp. Aliquots of
mixed cultures were placed in a Petri dish and exposed
Feed pressure Permeate flow rate
to UV light from an LP-UV lamp for pre-defined peri-
Feed temperature Permeate pressure
ods of time. The LP-UV dose delivered to each aliquot
Feed conductivity/TDS Permeate conductivity/TDS was calculated as the product of the intensity of the
Permeate flow rate Brine flow rate incident UV light, the UV absorbance of the water, and
Permeate pressure Brine pressure the exposure time. The dose response curve of the tar-
Permeate conductivity/TDS get microorganisms was constructed by measuring the
Brine pressure microbial concentration after each LP-UV exposure, Nt,
(2nd stage feed pressure) calculating the log inactivation of the microorganism
 156 T. Harif et al. / Desalination and Water Treatment 31 (2011) 151–163

ratories, Inc., CA). The 16S rRNA gene fragments were

amplified with the DreamTaq kit (Fermentas, Burlington,
Canada) and the general primers 11F [41] and 1392R [42].
The PCR products were resolved on 1% agarose gel and
the samples with apparent band at the 1400bp were used
for cloning. PCR products were directly cloned into the
pCR2.1 TOPO TA cloning vector (Invitrogen, Inc, Carls-
bad, CA) according to the manufacture instruction. Liga-
tion products were sent to the Genome Sequencing Center
at Washington University, St. Louis, MO, USA for further
processing (transformation into E. coli, clones picking,
and sequencing). A total of 148 and 172 sequences were
obtained for the biofilm obtained with and without UV
pre-treatment, respectively using the 907R primer [43]
for the sequencing reaction.
Downloaded by [Ryerson University Library] at 18:39 04 June 2016

Fig. 2. Collimated beam apparatus (USEPA, 2006). 2.3.1. Microbial community composition data analysis
A base calling analysis was performed on all sequences
retrieved using BioEdit (http://www.mbio.ncsu.edu/
(log (Nt/No)), and plotting this value against the respec-
BioEdit/bioedit.html). The sequences were aligned and
tive dose.
affiliation was determined based on the closest phyloge-
The biodosimetry curves obtained from HPC bacteria
netic group using the Arb software [44]. A UniFrac analy-
and Pseudomonas spp. cultures isolated from the BWRO
sis [45] of the 6 different libraries determined that there
inlet water appear in Fig. 3. The point where the graph
was no difference within each treatment and therefore
deviates from the log-linear relationship marked the
the data was compiled into two datasets; community of
minimum LP-UV dose that was applicable. In practice,
biofilm formed in UV pre-treated water and non-treated
a safety margin was applied and the MP-UV dose of the
water. Diversity indexes were calculated with Mothur
full-scale system was set at higher values than the mini-
[46] and the coverage for each library was calculated as
mum, at approximately 80 mJ/cm2.
follows: 1 – (n/N) where n is the number of singletons and
N is the total number of clones per library.
2.2.3. Microbial concentration analysis
Water samples were collected from the desalination
plant, at the inlet to the RO trains, and after the MP-UV 2.4. Fluorescent in situ hybridization analysis
system, in laboratory grade sampling bottles under
Fluorescent in situ hybridization (FISH) analysis was
aseptic conditions and stored on ice. All samples were
performed as described previously [47] using the follow-
filtered according to the procedure described in section
ing phylogenetic probes: Alf968 for Alphaproteobacteria
2.2.1. Following filtration of the samples, the filters were
[48], CF319a for Bacteroidetes [49], Bet42a combined with
placed into Petri dishes containing Standard Methods
a competitive unlabeled probe Gam42a for Betaproteo-
agar, Cetrimide agar or Vinogradov agar — for selection
bacteria [50], Ntspa662 for Nitrospira (Daims et. al, 2001)
of HPC, and Pseudomonas spp. or iron bacteria cultures
Pla46 and Pla886 for Planctomycetes [51]. RO membranes
respectively. The Petri dishes were incubated for 72 h at
sections were hybridized with the probes for 5 h at 46°C in
30°C, for HPC and Pseudomonas spp. and for 5 days at
35% formamide, and then washed in wash buffer (20 mM
30°C for iron bacteria. Following incubation the colonies
Tris-HCl pH 7.2, 10 mM EDTA, 0.01% SDS, 80 mM NaCl)
on each filter were counted and the microbial concentra-
for 1 h at 48°C to remove residual probes. The samples
tion was calculated as CFU/ml or CFU/100 ml (CFU-
were then counterstained with 0.5 ng/ml 4’,6’-diamidino-
colony forming unit).
2-phenylindole (DAPI) and kept at –20°C until data
collection. Confocal laser scanning microscopy (CLSM)
was performed using the Leica SP5 (Leica, Microsystems
2.3. Molecular analysis of microbial community composition
CMS GmbH, Mannheim Germany) with 4 channels. A 405
Three RO membrane coupons that were carefully cut diode, UV laser for the DAPI, Argon laser for the Alexa
from the center of each sacrificial membrane element 488 or 6Fam labeled probes, 516 laser for the Cy3 labeled
were placed in separate 2 ml centrifuge tubes containing probes, and 633 laser for the Cy5 probes. Twenty fields
sterile glass beads and the DNA extraction was performed of view were collected for each sample and the absolute
using the UltraClean soil DNA isolation kit (MoBio Labo- number of cells per cm2 was calculated and averaged.
 T. Harif et al. / Desalination and Water Treatment 31 (2011) 151–163 157

2.5. Cells and EPS biovolume by CLSM using the Zeiss LSM Image Browser. Gray scale images
were analyzed, and the specific biovolume (µm3/µm2)
RO membrane sections, cut from the center of the sac-
in the biofouling layer was determined by COMSTAT,
rificial membrane elements, into pieces of approximately
an image-processing software [52], written as a script in
5 mm × 5 mm, were fixed in 4% (w/v) para-formaldehyde
Matlab 6.5 (The MathWorks) and equipped with an image
(PFA) for 1 h and subsequently stored in 50% ethanol/PBS
processing toolbox. Thresholding was fixed for all image
(v/v). Prior to microscopic observation the fouled mem-
stacks. For every sample between 4–6 positions on the
brane coupons were stained with concanavalin A (ConA)
membrane were chosen and microscopically observed,
conjugated to Alexa fluor 633 and propidium iodide (PI)
acquired, and analyzed.
(Invitrogen Co.), for probing extracellular polymeric sub-
stances (EPS) and microorganisms, respectively. Briefly,
frozen (–20°C) 100 μl aliquots of 1 mg/ml labeled ConA 3. Results
stock solution were prepared and diluted to 100 μg/ml
3.1. Biodosimetry curves
prior to use in 10 mM phosphate buffer (pH 7.5). An
excess electrolyte solution was carefully drawn off from The biodosimetry curves (Fig. 3) show the inactivation
the fouled membrane by gently touching the edge of the kinetics of microorganisms that represent communities
Downloaded by [Ryerson University Library] at 18:39 04 June 2016

specimens with an adsorbing paper (Kimwipes). Then, typically found in water and also include known primary
100 μl of ConA staining solution were added to cover biofilm colonizers [53]; therefore, they are commonly used
the samples, which were then incubated in the dark at as indicators for sizing UV applications. Both graphs
room temperature for 20 min. Unbound ConA was drawn exhibit typical inactivation behavior [31]: a log-linear re-
off the specimens using a three-step wash of 10 mM lationship, and then deviation from this relationship. The
phosphate buffer. The unbound ConA solution and the HPC inactivation kinetics depicts a “worst case scenario”
washing solutions were carefully removed by gently in the UV resistance of a wide variety of microorganisms
touching the edge of the specimen with an adsorbing is examined. As previously mentioned, different microbi-
paper. Probing the microorganisms in the fouling layer al communities boast different resistance characteristics,
was performed with 30 µM PI solution that was added thus the average resistance observed is higher and the
to cover the samples, which were then incubated in the tailing effect begins at 5 log inactivation corresponding to
dark at room temperature for 20 min (prepared in 10 mM a LP-UV dose of 20 mJ/cm2. This effect is observed with
phosphate buffer, pH 7.5). Excess electrolyte solution was Pesudomonas spp. at barely 10 mJ/cm2. Hence, the MP-UV
carefully drawn off from a piece of a fouled membrane dose applied throughout the study was maintained above
in the same manner used for ConA staining. 20 mJ/cm2, and included a safety margin that accounted
Microscopic observation and image acquisition were for changes in UV water transmittance and reduction of
performed using Zeiss-Meta 510, a confocal scanning laser lamp efficiency; primary factors that affect the effective
microscope (CLSM), equipped with Zeiss dry objective UV dose. In practice, the applied MP-UV dose throughout
LCI Plan-NeoFluar (10X magnification and numerical the run was approximately 80 mJ/cm2.
aperture of 0.3). The CLSM was equipped with detec-
tors and filter sets for monitoring PI stained cells and
3.2. Microbial counts
the Alexa fluor 633 dye (excitation wavelengths of 488
and 633 nm, respectively). CLSM images were generated Microbial counts in the BWRO plant inlet water in

Biodosimetry Curve of Heterotrophic Plate Count Biodosimetry Curve of Pseudomonas spp.

6 6

5 5
Log Inactivation

Log Inactivation

4 4

3 3

2 2

1 1

0 0

0 20 40 60 80 100 120 0 5 10 15 20 25 30
UV Dose (mJ/cm2) UV Dose (mJ/cm2)

Fig. 3. The curves obtained from CBA biodosimetry tests performed on HPC (left) and Pseudomonas spp. (right) isolated from
the BWRO plant inlet water.
 158 T. Harif et al. / Desalination and Water Treatment 31 (2011) 151–163

both trains and in the outlet of the MP-UV system were CFU/ml of HPC bacteria, no iron bacteria and hardly any
performed every two weeks to ensure the integrity of the Pseudomonas spp. — although at the inlet to the MP-UV
MP-UV system and to provide an image of the microbial system, the counts heavily fluctuated. These fluctuations
concentration of the MP-UV treated water entering RO2 are representative of the microbial quality of the water
vs. the un-treated water entering RO1, for the entire dura- entering RO1, and we can assume that in the microbial
tion of the run. These analyses consisted of HPC, Pseu- load entering RO1 was higher in comparison to RO2 due
domonas spp. and iron bacteria. Results appear in Fig. 4. to implementation of the MP-UV system.
The results show that for the duration of the run, the
MP-UV system produced a microbial baseline in the water
3.3. RO train performance
entering RO2, containing no more than a few hundred
Biofouling phenomena is more prevalent in the first
stage of an RO train, thus the normalized flux of stage 1
Heterotrophic Plate Count was calculated for RO1 and RO2 (Fig. 5). A normalized
1000000 a permeate relative flux was obtained by additionally
RO1 inlet
RO2 inlet normalizing the permeate flux according to the maxi-
100000 UV (RO2) outlet mum flux obtained in each train, following the recovery
Downloaded by [Ryerson University Library] at 18:39 04 June 2016

of the membranes after the cleaning procedure (Fig. 6).

10000
Membrane surface properties are highly susceptible to
CFU/ml

changes while exposed to various cleaning agents [11],

1000
and require a recovery period after cleaning, until they
regain initial permeability. In this case the recovery pe-
100
riod lasted 3 and 5 days for RO1 and RO2, respectively.
Therefore, the graphs show an increase in normalized
10
14/08/09 27/08/09 10/09/09 08/11/09 22/11/09 03/12/09
permeate flux during the initial days of the run, however,
Date after reaching a maximum value at full recovery, start to
exhibit a decline.
Iron Bacteria Count
10000 b For the first six weeks of the run, both trains showed
RO1 inlet similar performance, however from this point, the nor-
RO2 inlet malized permeate flux decline became more significant
UV (RO2) outlet
1000
in RO1. At the end of the run, a 6% difference existed
between both trains, with RO1 at 83% of the initial nor-
CFU/100 ml

malized permeate flux, and RO2 at 89%.

100
3.4. Community composition of the biofilm
The microbial community composition and diversity
10 was assessed using a culture-independent technique.
14/08/09 27/08/09 10/09/09 08/11/09 22/11/09 03/12/09 The diversity of the community that formed the biofilm
Date on RO membrane from MP-UV pre-treated water was
Pseudomonas spp. Count significantly less diverse than that from the untreated
100000 c water (Table 5). While some phylogenetic groups were
RO1 inlet completely absent from the biofilm formed in the MP-UV
RO2 inlet pre-treated water, there were a few groups present in both
10000 UV (RO2) outlet
biofilm, but with varying relative abundance (Table 5).
The alphaproteobacterial group Parvularcula was only a
CFU/100 ml

1000 minor (4%) component of the community in the biofilm

of the untreated water, but was the most dominant group
(22%) in the biofilm formed after MP-UV application.
100
Similarly, Betaproteobacteria and Chloroflexi were more
dominant in the biofilm formed on RO membranes after
10 MP-UV pre-treatment, but not in the biofilm formed from
14/08/09 27/08/09 10/09/09 08/11/09 22/11/09 03/12/09 untreated water. Two phylogenetic groups seemed to be
Date
most affected by the MP-UV pre-treatment were affiliated
Fig. 4. Microbial counts in the BWRO plant inlet water to both with Nitrospirae and Bacteroidetes as their fraction in the
trains, and after UV pre-treatment. a. HPC (CFU/ml); b. Iron community was reduced from 33 and 8% to 15 and less
bacteria (CFU/100ml); c. Pseudomonas spp. (CFU/100ml). than 1%, respectively (Table 5).
 T. Harif et al. / Desalination and Water Treatment 31 (2011) 151–163 159

Fig. 5. Normalized permeate flux of stage 1 with (RO2) and without (RO1) MP-UV pretreatment.
Downloaded by [Ryerson University Library] at 18:39 04 June 2016

Fig. 6. Relative normalized permeate flux of stage 1 with (RO2) and without (RO1) MP-UV pre-treatment.

These observations were further confirmed with

FISH-CLSM analysis. The total amount of bacteria per
area was reduced from 1.3×107 to 8.7×106 cell/cm2, a
reduction by 1.5 fold. The most dominant phylogenetic
group in the biofilm formed on a RO membrane that
received un-treated water was affiliated with Nitrospira.
This was also the group that was the most affected by the
MP-UV pre-treatment as its abundance was reduced in
24% in the biofilm that received MP-UV treated water.
The abundance of Bacteroidetes and Planctomycetes was
reduced by 10% each (Fig. 7). However, the abundance of
Alphaproteobacteria which was the second important group
in the un-treated biofilm did not change in response the
MP-UV pre-treatment.

3.5. CLSM analysis of the biofilm

Fig. 7. Phylogenetic analysis by FISH of biofilm developed
Fig. 8 shows a representative view of the effect of on BWRO membranes that received water with and without
MP-UV pre-treatment on biofouling of the BWRO mem- MP-UV pre-treatment. Total amount of cells by DAPI counts
branes. The MP-UV pre-treatment affected the EPS specif- are indicated for each treatment in cells/cm2.
 160 T. Harif et al. / Desalination and Water Treatment 31 (2011) 151–163

Table 5 ic biovolume in the biofilm which was approximately half

Microbial community composition of biofilm formed on (120 mm3/mm2) of that found on the membrane that received
BWRO membrane that received water with and without MP- water without MP-UV pre-treatment (230 mm3/mm2). Ad-
UV pre-treatment ditionally, a minor difference in the cell specific biovolume
was observed between the membranes, where a lower
Phylogenetic group Train
specific biovolume was observed on the membrane that
RO2 (UV) RO1 (No UV) received water with MP-UV pre-treatment. The CLSM
Alphaproteobacteria analysis of the biofouling layer coincides with the normal-
ized permeate flux decline in the trains with and without
Sphingomonadaceae 4.73 1.16
MP-UV pre-treatment. With MP-UV pre-treatment, sig-
Parvularcula 22.97 4.07
nificantly less biofilm growth, as well as a lower decline
Other Alphaproteobacteria 14.86 11.63
rate in membrane performance were observed.
Betaproteobacteria 16.22 5.23
Gammaproteobacteria 2.03 3.49
Deltaproteobacteria 2.03 3.49 4. Discussion
Nitrospirae 14.86 33.14 The results suggest that MP-UV pre-treatment heav-
Downloaded by [Ryerson University Library] at 18:39 04 June 2016

Chloroflexi 12.16 1.74 ily impacts the characteristics of biofilm that forms on
Bacteroidetes 0.68 7.56 RO membranes — in terms of specific biovolume, EPS
Verrucomicrobia 0.68 1.16 content and microbial community composition. The
Cyanobacteria 8.11 5.23 biofilm analyzed by CLSM (Figs. 7 and 8) correlates to
Actinobacteria 0.68 2.91 the normalized permeate flux behavior with and without
Planctomycetes 0 9.88 MP-UV pre-treatment: significantly lower EPS specific
Candidate division OP3 0 5.81 biovolumes (Fig. 8) were found on the membrane from
RO2 that received MP-UV pre-treated water, and also
Acidobacteria 0 0.58
exhibited a lower normalized permeate flux decline
Fibrobacter 0 0.58
(Fig. 5). Moreover, RO2 operated at a higher normalized
Candidate division TM6 0 0.58
permeate flux, rendering it more susceptible to fouling,
Firmicutes 0 0.58 and despite this, still maintained a higher normalized
Other bacteria 0 1.16 permeate flux over time.
OTU 42 86 Although the amount of EPS was found to be drasti-
Shannon index 3.73 ± 0.21 4.45 ± 0.14 cally different between the membranes, the cell specific
Coverage 85% 62% biovolumes were not found to differ substantially with
marginally less cells found on the membrane that received
MP-UV pre-treated water. However, total cell counts de-
creased significantly on the membrane (t-test, p < 0.05).
This can be explained by the inclusive mutagenic effect
that MP-UV has on the bacteria in the feed solution [54,55]
UV, as opposed to conventional biocides, does not oxidize
the cell membrane, but damages the reproduction abili-
ties of the bacteria, rendering them intact, but unable to
reproduce [56–58]. MP-UV additionally has enhanced
ability to target proteins and lipids and may promote
cell injury to an extent that additional basic cellular
functions become damaged [59]. In this case, deposition
of inactivated bacteria can occur on the membrane sur-
face, the most reasonable explanation for the relatively
small difference in the cell specific biovolumes between
the membranes (Fig. 7) and in comparison to the FISH
enumeration.
The significantly less content of EPS in the biofilm on
the membrane that received MP-UV pre-treated water, is
Fig. 8. (A) Biofouling layer without MP-UV pre-treatment (B)
Biofouling layer with MP-UV pre-treatment. Total biomass of most likely connected to the inability of the inactivated/
EPS (transparent light blue) and microorganisms (red) was damaged bacteria, to excrete normal amounts of EPS,
analyzed using COMSTAT biofilm software (C). The images which is produced at a later stage of the biofilm evolu-
are perspectives of 900 µm × 900 µm after 3-D reconstruction tion [60]. A number of studies have shown that mutants
of the CLSM image stack with IMARIS software. unable to synthesize EPS are unable to form biofilms,
 T. Harif et al. / Desalination and Water Treatment 31 (2011) 151–163 161

although they may still attach to surfaces and form chemical surface characteristics of the model strain are to
micro-colonies to a limited extent [61,62]. Biofilm forma- be further explored as well as the related physiological
tion stages include microbial deposition and attachment changes (changes in motility and other adhesion factors).
and then, production and secretion of macromolecules Coupling these findings with further investigation into
that are responsible for creating a three dimensional optimization of the operational parameters and location
mechanically stable visco-elastic structure. This structure of the MP-UV system could offer a sustainable solution
has been shown to reduce significantly RO permeate flux for combating biofouling in RO processes.
by increasing hydraulic resistance of the RO membrane
[14], hence the apparent effect of MP-UV pre-treatment on
5. Concluding remarks
EPS production could indeed have ramifications in main-
taining flux in RO filtration. Consequently, this would The results from this study show that pre-treatment
increase the interval between cleaning procedures and using MP-UV substantially impacts the characteristics
may even affect the type of cleaning regime implemented, of biofilm formed on RO membranes; in terms of micro-
diminishing the need for certain types of chemicals and bial community composition, specific biovolume and
thereby improving the membrane recovery properties EPS content. The diminished EPS content found on the
and lifetime. biofilm that formed following MP-UV pre-treatment has
Downloaded by [Ryerson University Library] at 18:39 04 June 2016

Another explanation for the reduction in EPS amounts operational ramifications: EPS is not only the “cement”
could indeed be the change in the microbial community of the biofilm that facilitates adhesion to the membrane
composition due to MP-UV treatment. Nitrospira was the surface, but has been found to contribute significantly to
most affected group by the UV pre-treatment. This chem- the RO permeate flux decline by increasing the hydraulic
lithoautotrophic nitrite-oxidizing group [63] is common resistance to permeate flow. Its reduction by adequate
in various environments, including wastewater treatment pre-treatment may actually reduce the need for certain
facilities [64] and in water distribution systems [65]. It is cleaning regimes and consequently improve the mem-
possible that the MP-UV caused enough damage to the brane’s recovery properties and lifetime.
DNA of this bacteria with no capability of recovery and These preliminary results open avenues for further
hence, their low abundance in the biofilm. Some nitrite- investigation into the abilities of MP-UV to target the
oxidizing bacteria are known to be sensitive to light [66] biofouling potential of microorganisms on RO mem-
and it is possible that Nitrospira share this trait. Bacteria branes, one of these is its ability to affect microorganism
affiliated with Nitrospira were found to produce large attachment properties. An understanding of the com-
amount of extracellular polymers [67], and therefore the prehensive effect of MP-UV on the biofouling potential
reduction in their abundance in the biofilm also could of microorganisms on commercial RO membranes could
explain the reduction in the EPS content. The fraction of enable efficient tailoring of MP-UV as a viable technology
Alpha-, Betaproteobacteria and Chloroflexi in the biofilm for combating biofouling in RO plants.
increased following the MP-UV pre-treatment. Isolates
of Chloroflexi have the capabilities to form spores [68,69]
Acknowledgements
that may protect them from UV damage while in the
planktonic phase. These spores may germinate to become The authors gratefully acknowledge Mr. Shaul Oren,
active bacteria in the biofilm. In contrast, marine isolates who enabled this work at the desalination plant. This
that affiliates with Parvularcula and Sphingomonadaceae study was supported by the Israel Ministry of Industry
do not have this ability [70–72] and they may have an Trade and Labor within the Magnet funding program.
alternative mechanism for protection. However, a dif-
ferent MP-UV dose may be affective in eliminating these References
bacteria also from the biofilm, but laboratory experiments
on appropriate model organisms should be performed [1] L.F. Greenlee, D.F. Lawler, B.D. Freeman, B. Marrot and P. Mou-
lin, Reverse osmosis desalination: Water sources, technology and
to confirm this. today’s challenges. Wat. Res., 43 (2009) 2317–2348.
Although EPS may promote adherence to the RO [2] K.L. Chen, L. Song, S.L. Ong and J.N. Wun, The development
membrane, the initial adhesion of the cells to the mem- of membrane fouling in full-scale RO processes. J. Membr. Sci.,
branes is highly affected by adhesion forces dictated by 232 (2004) 63–72.
[3] L.Y. Dudley and J.S. Baker, Biofouling in membrane systems – a
the physical and chemical properties of the membrane
review. Desalination, 118 (1998) 81–90.
and of the cells [6,8,12,73,74]. It has been hypothesized [4] H.-C. Flemming, Reverse osmosis membrane biofouling. Ex-
by the authors that MP-UV pre-treatment, due to its perim. Thermal Fluid Sci., 14 (1997) 382–391.
mutagenic effects, may also impact the attachment [5] J.S. Vrouwenvelder, S.A. Manolarakis, H.R. Veenendaal and D.
properties of the cells. For preliminary evaluation of this van der Kooij, Biofouling of chemicals used for scale control in
RO and NF membranes. Desalination, 132 (2000) 1–10.
hypothesis, experiments have been conducted, the results [6] M.M.T. Khan, P.S. Stewart, D.J. Moll, W.E. Mickols, M.D. Burr,
of which suggest that bacterial deposition is significantly S.E. Nelson and A.K. Camper, Assessing biofouling on poly-
affected by MP-UV. The impact of MP-UV on the physico- amide reverse osmosis (RO) membrane surfaces in a laboratory
 162 T. Harif et al. / Desalination and Water Treatment 31 (2011) 151–163

system. J. Membr. Sci., 349 (2010) 429–437. compartments of a full-scale reverse-osmosis water purification
[7] E. Huertas, M. Herzberg, G. Oron and M. Elimelech, Influence plant. Appl. Environ. Microbiol., 74 (2008) 5297–5304.
of biofouling on boron removal by nanofiltration and reverse [27] D. Kim, S. Jung, J. Sohn, H. Kim and S. Lee, Biocide application
osmosis membranes. J. Membr. Sci., 318 (2008) 264–270. for controlling biofouling of SWRO membranes – an overview.
[8] M. Herzberg and M. Elimelech, Biofouling of reverse osmosis Desalination, 238 (2009) 43–52.
membranes: Role of biofilm-enhanced osmostic pressure. J. [28] L.E. Appelgate, C.W. Erkenbrecher, Jr. and H. Winters, New
Membr. Sci., 295 (2007) 11–20. chloramines process to control aftergrowth and biofouling in
[9] H.-C. Flemming, G. Schaule, T. Griebe, J. Schmitt and A. Tamach- permasep B-10 RO surface seawater plants. Desalination, 74
kiarowa, Biofouling – the Achilles heel of membrane processes. (1989) 51–67.
Desalination, 113 (1997) 215–225. [29] A.M. Munshi, M.O. Saeed and T.N. Green, Impact of UV irradia-
[10] M. Saad, Biofouling prevention in RO polymeric membrane tion on controlling biofouling problems in NF–SWRO desalina-
systems. Desalination, 88 (1992) 85–105. tion process. Proc., International Desalination Association World
[11] A. Subramani and E.M.V. Hoek, Biofilm formation, cleaning, Congress, Singapore, 2005.
re-formation on polyamide composite membranes. Desalina- [30] M. Otaki, S. Takizawa and S. Ohgaki, Control and modeling of
tion, 257 (2010) 73–79. membrane fouling due to microorganism growth by UV pretreat-
[12] A. Subramani and E.M.V. Hoek, Direct observation of initial ment. Wat. Sci. Technol., 38(4–5) (1998) 405–412.
microbial deposition onto reverse osmosis and nanofiltration [31] W.A.M. Hijnen, E.F. Beerendonk and G.J. Medema, Inactivation
membranes. J. Membr. Sci., 319 (2008) 111–125. credit of UV radiation for viruses, bacteria and protozoan (oo)
[13] R.M. Donlan, Biofilms: microbial life on surfaces. Emerging cysts in water: A review. Wat. Res., 40 (2006) 3–22.
Downloaded by [Ryerson University Library] at 18:39 04 June 2016

Infectious Diseases, 8 (2002) 881–890. [32] K. Oguma, H. Katayama and S. Ohgaki, Photoreactivation of
[14] M. Herzberg, S. Kang and M. Elimelech, Role of extracellular Escherichia coli after low- or medium-pressure UV disinfection
polymeric substances (EPS) in biofouling of reverse osmosis determined by an endonuclease sensitive site assay. Appl. En-
membranes. Environ. Sci. Technol., 43 (2009) 4393–4398. viron. Microbiol., 68 (2002) 6029–6035.
[15] H.-C. Flemming, T.R. Neu and D. Wozniak, The EPS matrix: The [33] J.L. Zimmer and R.M. Slawson, Potential repair of Escherichia coli
“house of biofilm cells”. J. Bacteriol., 189(22) (2007) 7945–7947. DNA following exposure to UV radiation from both medium-
[16] A.S. Kim, H. Chen and R. Yuan, EPS biofouling in membrane and low-pressure UV sources used in drinking water treatment.
filtration: An analytic modeling study. J. Colloid Interf. Sci., 303 Appl. Environ. Microbiol., 68 (2002) 3293–3299.
(2006) 243–249. [34] K.G. Linden, J. Thurston, R. Schaefer and J.P. Malley, Enhanced
[17] H. Dang, T. Li, M. Chen and G. Huang, Cross-ocean distribu- UV inactivation of adenoviruses under polychromatic UV lamps.
tion of Rhodobacterales bacteria as primary surface colonizers Appl. Environ. Microbiol., 73 (2007) 7571–7574.
in temperate coastal marine waters. Appl. Environ. Microbiol., [35] H. Mamane-Gravetz, K.G. Linden, A. Cabaj and R. Sommer,
74 (2008) 52–60. Spectral sensitivity of bacillus subtilis spores and MS2 coliphage
[18] P.R. Jones, M.T. Cottrell, D.L. Kirchman and S.C. Dexter, Bacte- for validation testing of ultraviolet reactors for water disinfection.
rial community structure of biofilms on artificial surfaces in an Environ. Sci. Technol., 39 (2005) 7845–7852.
estuary. Microb. Ecol., 53 (2006) 153–162. [36] J.L. Zimmer-Thomas, R.M. Slawson and P.M. Huck, A com-
[19] H. Dang and C.R. Lovell, Numerical dominance and phylotype parison of DNA repair and survival of Escherichia coli O157:H7
diversity of marine Rhodobacter species during early colonization following exposure to both low- and medium-pressure UV
of submerged surfaces in coastal marine waters as determined irradiation. J. Wat. Health, 5(3) (2007) 407–415.
by 16S ribosomal DNA sequence analysis and fluorescence in [37] K. Oguma, H. Katayama and S. Ohgaki, Photo-reactivation of
situ hybridization. Appl. Environ. Microbiol., 68 (2002) 496–504. Legionella pneumophilia after inactivation by low- or medium-
[20] H. Dang and C.R. Lovell, Bacterial primary colonization and pressure ultraviolet lamp. Wat. Res., 38 (2004) 2757–2763.
early succession on surfaces in marine waters as determined by [38] K. Tosa and T. Hirata, Photo-reactivation of entero-hemorrhagic
amplified rRNA gene restriction analysis and sequence analysis Escherichia coli following UV disinfection. Wat. Res., 33 (1999)
of 16S rRNA genes. Appl. Environ. Microbiol., 66 (2002) 467–475. 361–366.
[21] H. Honma, R. Asano, M. Obara, K. Otawa, Y. Suyama and Y. [39] J.R. Bolton and K.G. Linden, Standardization of methods for
Nakai, Bacterial populations in epilithic biofilms along two fluence (UV dose) determination in bench scale UV experiments.
oligotrophic rivers in the Tohoku region in Japan. J. Gen. Appl. J. Environ. Eng., 129(3) (2003) 209–215.
Microbiol., 55 (2009) 359–371. [40] US Environmental Protection Agency, UV disinfection guid-
[22] W. Manz, K. Wendt-Patthoff, T.R. Neu, U. Szewzyk and J.R. ance manual for the final long term 2 enhanced surface water
Lawrence, Phylogenetic composition, spatial structure, and treatment rule, 2006.
dynamics of lotic bacterial biofilms investigated by fluorescence [41] M.D. Kane, L.K. Poulsen and D.A. Stahl, Monitoring the enrich-
in situ hybridization and confocal laser scanning microscopy. ment and isolation of sulfate-reducing bacteria by using oligo-
Microb. Ecol., 37 (1999) 225–237 nucleotide hybridization probes designed from environmentally
[23] L.-N. Huang, H. De Wever and L. Diels, Diverse and distinct derived 16S rRNA sequences. Appl. Environ. Microbiol., 59
bacterial communities induced biofilm fouling in membrane (1993) 682–686.
bioreactors operated under different conditions. Environ. Sci. [42] D.J. Lane, 16S/23S rRNA sequencing. Nucleic Acid Techniques in
Technol., 42 (2008) 8360–8366. Bacterial Systematics, E. Stackebrandt and M. Goodfellow, eds.,
[24] C.M. Pang and W.-T. Liu, Community structure analysis of John Wiley & Sons, 1997, Chichester. pp. 115–175.
reverse osmosis membrane biofilms and the significance of [43] G. Muyzer, E.C. De Waal and A.G. Uitterlinden, Profiling of
Rhizobiales bacteria in Biofouling. Environ. Sci. Technol., 41 complex microbial populations by denaturing gradient gel
(2007) 4728–4734. electrophoresis analysis of polymerase chain reaction-amplified
[25] L.A. Bereschenko, A.J.M. Stams, G.J.W. Euverink and van M.C.M. genes coding for 16S rRNA. Appl. Environ. Microbiol., 59 (1993)
Loosdrecht, Biofilm formation on reverse osmosis membranes 695–700.
is initiated and dominated by Sphingomonas spp. Appl. Environ. [44] W. Ludwig, O. Strunk, R. Westram, L. Richter and H. Meier,
Microbiol., 76 (2010) 2623–2632. ARB: a software environment for sequence data. Nucleic Acids
[26] L.A. Bereschenko, G.H.J. Heilig, M.M. Nederlof, M.C.M. van Res., 32 (2004) 1363–1371.
Loosdrecht, A.J.M. Stams and G.J.W. Euverink, Molecular [45] C. Lozupone and R. Knight, UniFrac: a new phylogenetic
characterization of the bacterial communities in the different method for comparing microbial communities. Appl. Environ.
 T. Harif et al. / Desalination and Water Treatment 31 (2011) 151–163 163

Microbiol., 71 (2005) 8228–8235. [60] M. Herzberg and M. Elimelech, Physiology and genetic traits
[46] P.D. Schloss, S.L. Westcott, T. Ryabin, J.R. Hall, M. Hartmann, of reverse osmosis biofilms: a case study with pseusomonas
E.B. Hollister, R.A. Lesniewski, B.B. Oakley, D.H. Parks, C.J. aeruginosa. Intern. Soc. Microb. Ecol., 2 (2008) 180–194.
Robinson, J.W. Sahl, B. Stres, G.G. Thallinger, D.J. Van Horn [61] P.I. Watnick and R. Kolter, Steps in the development of a Vibrio
and C.F. Weber, Introducing mothur: open-source, platform- cholera El Tor biofilm. Molec. Microbiol., 34 (1999) 586–595.
independent, community-supported software for describing and [62] D.G. Allison and I.W. Sutherland, Role of exopolysaccharides
comparing microbial communities. Appl. Environ. Microbiol., in adhesion of freshwater bacteria. J. Genet. Microbiol., 133
75 (2009) 7537–7541. (1987) 1319–1327.
[47] W. Manz, R. Amann, W. Ludwig, M. Wagner and K.H. Schleifer, [63] F. Maixner, D.R. Noguera, B. Anneser, K. Stoecker, G. Wegl,
Phylogenetic oligodeoxynucleotide probes for the major sub- M. Wagner and H. Daims, Nitrite concentration influences the
classes of Proteobacteria — Problems and solutions. Syst. Appl. population structure of Nitrospira-like bacteria. Environ. Micro-
Microbiol., 15 (1992) 593–600. biol., 8 (2006) 1487–1495.
[48] F.O. Glöckner, B.M. Fuchs and R. Amann, Bacterioplankton [64] H. Daims, J.L. Nielsen, P.H. Nielsen, K.-H. Schleifer and M. Wag-
compositions of lakes and oceans: a first comparison based on ner, In situ characterization of Nitrospira-like nitrite-oxidizing
fluorescence in situ hybridization. Appl. Environ. Microbiol., bacteria active in wastewater treatment plants. Appl. Environ.
65 (1999) 3721–3726. Microbiol., 67 (2001) 5273–5284.
[49] W. Manz, R. Amann, W. Ludwig, M. Vancanneyt and K.H. [65] A.C. Martiny, H.-J. Albrechtsen, E. Arvin and S. Molin, Iden-
Schleifer, Application of a suite of 16S rRNA-specific oligonucle- tification of bacteria in biofilm and bulk water samples from
otide probes designed to investigate bacteria of the phylum a nonchlorinated model drinking water distribution system:
Downloaded by [Ryerson University Library] at 18:39 04 June 2016

Cytophaga–Flavobacter–Bacteroides in the natural environment. detection of a large nitrite-oxidizing population associated with
Microbiol., 142 (1996) 1097–1106. Nitrospira spp. Appl. Environ. Microbiol., 71 (2005) 8611–8617.
[50] H. Daims, J.L. Nielsen, P.H. Nielsen, K.H. Schleifer and M. Wag- [66] E. Bock and M. Wagner, Oxidation of inorganic nitrogen com-
ner, In situ characterization of Nitrospira-like nitrite-oxidizing pounds as an energy source. In The Prokaryotes: Ecophysiology
bacteria active in wastewater treatment plants. Appl. Environ. and Biochemistry, M. Dworkin, S. Falkow, E. Rosenberg, K-H.
Microbiol., 67 (2001) 5273–5284. Schleifer and E. Stackebrandt, eds., Springer, vol. 2, 2007, p. 463.
[51] A. Neef, R. Amann, H. Schlesner and K.-H. Schleifer. Monitoring [67] S. Ehrich, D. Behrens, E. Lebedeva, W. Ludwig and E. Bock, A
a widespread bacterial group: in situ detection of planctomy- new obligately chemolithoautotrophic, nitrite-oxidizing bac-
cetes with 16S rRNA-targeted probes. Microbiol., 144 (1998) terium, Nitrospira moscoviensis sp. nov. and its phylogenetic
3257–3266. relationship. Arch. Microbiol., 164 (1995) 16–23.
[52] A. Heydorn, A.T. Nielsen, M. Hentzer, C. Sternberg, M. Givskov, [68] S. Yabe, Y. Aiba, Y. Sakai, M. Hazaka and A. Yokota, A life cycle of
B.K. Ersboll and S. Molin, Quantification of biofilm structures branched aerial mycelium- and multiple budding spore-forming
by the novel computer program COMSTAT, Microbiol.– UK, bacterium Thermosporothrix hazakensis belonging to the phylum
146 (2000) 2395–2407. Chloroflexi. J. Gen. Appl. Microbiol., 56 (2010) 137–141.
[53] N. Doğruöz, D. Göskay, E. Ilhan-Sungur and A. Cotuk, Pioneer [69] M.B. Stott, M.A. Crowe, B.W. Mountain, A.V. Smirnova, S. Hou,
colonizer microorganisms in biofilm formation on galvanized M. Alam and P.F. Dunfield, Isolation of novel bacteria, including
steel in a simulated recirculating cooling water system. J. Basic a candidate division, from geothermal soils in New Zealand.
Microbiol., 49(S1) (2009) S5–S12. Environl. Microbiol., 10 (2008) 2030–2041.
[54] W. Zhou, X. Tang, H. Xiao, Y. Wang and R. Wang, Response of [70] C. Chen, Q. Zheng, Y.-N. Wang, X.-J. Yan, L.-K., Hao, X. Du and
marine microalgae, heterotrophic bacteria and their relation- N.-Z. Jiao, Stakelama pacifica gen. nov., sp. nov., a novel member
ship to enhanced UV-B radiation. J. Ocean Univ. China, 8(1) of the family Sphingomonadaceae isolated from the Pacific Ocean.
(2009) 35–38. Intern. J. Syst. Evolut. Microbiol., 2010; doi 10.1099/ijs.0.018945-0.
[55] V.F. Zenoff, F. Siñeriz and M.E. Farías, Diverse responses to UV-B [71] A.B. Arun, W.-M. Chen, W-A. Lai, J.-H. Chou, P.D. Rekha, F.-T.
radiation and repair mechanisms of bacteria isolated from high Shen, S. Singh and C.-C. Young, Parvularcula lutaonensis sp.
altitude aquatic environments. Appl. Environ. Microbiol., 72(12) nov., a moderately thermotolerant marine bacterium isolated
(2006) 7857–7863. from a coastal hot spring. Intern. J. Syst. Evolut. Microbiol., 59
[56] G.P. Pfeifer, Formation and processing of UV photoproducts: Ef- (2009) 998–1001.
fects of DNA sequence and chromatin environment. Photochem. [72] J.-C. Cho and S.J. Giovannoni, Parvularcula bermudensis gen.
Photobiol., 65 (1997) 270–283. nov., sp. nov., a marine bacterium that forms a deep branch
[57] J.K. Setlow, The Effects of Ultraviolet Radiation and Photoreacti- in the a-Proteobacteria. Intern. J. System. Evolut. Microbio., 53
vation. Comprehensive Biochemistry, Chap. 5, 27 (1997) 157–209. (2003) 1031–1036.
[58] A.M. Rauth, The physical state of viral nucleic acid and the sensi- [73] W. Lee, C.-H. Ahn, S. Hong, S. Lee, Y. Baek and J. Yoon, Evalua-
tivity of viruses to ultraviolet light. Biophys. J., 5 (1965) 257–273. tion of surface properties of reverse osmosis membranes on the
[59] P. Mondon and M.M. Shahin, Comparative studies on the lethal, initial biofouling stages under no filtration conditions. J. Membr.
mutagenic, and recombinogenic effects of ultraviolet –A, –B, –C Sci., 351 (2010) 112–122.
and visible light with and without 8-Methoxypsoralen in Sac- [74] S. Kang, E.M.V. Hoek, H. Choi and H. Shin, Effect of membrane
charomyces cerevisiae. Photochem. Photobiol., 55(5) (1992) 713–721. surface properties during the fast evaluation of cell attachment.
Separ. Sci. Technol., 41 (2006) 1475–1487.