Subscriber access provided by University of Sussex Library
Sustainability Engineering and Green Chemistry
Membrane distillation biofouling: Impact of feed water temperature on biofilm characteristics and membrane performance Anne Bogler, and Edo Bar-Zeev Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b02744 • Publication Date (Web): 06 Aug 2018 Downloaded from http://pubs.acs.org on August 7, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 35
Environmental Science & Technology
1
Membrane distillation biofouling: Impact of feed water temperature
2
on biofilm characteristics and membrane performance
3
Anne Bogler* and Edo Bar-Zeev*
4
Zuckerberg Institute for Water Research, The Jacob Blaustein Institutes for Desert
5
Research, Ben-Gurion University of the Negev, Sede Boker, 84990, Israel
6
* Corresponding Author: Dr. Edo Bar-Zeev;
[email protected], Anne Bogler;
7
[email protected] 8
Key words
9
Membrane distillation, feed water temperature, wetting, biofouling, biofilm characteristics
10
ABSTRACT
11
Membrane distillation (MD) is a temperature driven membrane separation technology
12
that holds great potential for decentralized and sustainable wastewater treatment
13
systems. Yet, similarly to all membrane based systems, microbial fouling (biofouling)
14
might be a critical hurdle for MD wastewater treatment applications. In this study we
15
determined the impact of increasing feed water temperatures (47 °C, 55 °C and 65 °C)
16
on biofilm growth and MD performance via dynamic biofouling experiments with
17
Anoxybacillus sp. as a model bacterium. Our results indicated that cell growth was
18
reduced at 47 °C, resulting in moderate distillate water flux decline (30%). Differently,
19
extensive growth of Anoxybacillus sp. at feed water temperature of 55 °C caused severe
20
distillate water flux decline (78%). Additionally, biofouling induced membrane wetting,
1 ACS Paragon Plus Environment
Environmental Science & Technology
21
which facilitated the passage of bacteria cells and endospores through the membrane
22
structure into the distillate. Although bacteria growth was impaired at feed water
23
temperature of 65 °C, excessive production of EPS (compared to bacterial abundance)
24
crippled membrane separation due to severe pore wetting. These results underline the
25
importance of optimized operating conditions and development of anti-biofouling and
26
anti-wetting membranes for successful implementation of MD in wastewater treatment
27
with high biofouling propensity.
28
GRAPHICAL ABSTRACT
29 30
INTRODUCTION
31
Membrane distillation (MD) is a promising technology for sustainable fresh water
32
production, due to the possibility of utilizing waste heat or low-grade solar and
33
geothermal heat for separation of high quality fresh water from contaminated feed
34
solutions.1–3 Distillate water flux is driven by a water vapor pressure gradient across a
35
water-excluding microporous hydrophobic membrane.4,5 Air-filled membrane pores only
36
allow the passage of vapors that evaporate from the heated feed water and condense 2 ACS Paragon Plus Environment
Page 2 of 35
Page 3 of 35
Environmental Science & Technology
37
into the cold distillate. Therefore, all non-volatile contaminants are rejected at the vapor-
38
liquid interface.1,4 MD operates at ambient hydraulic pressure, which reduces system
39
complexity and facilitates the application of MD in decentralized wastewater treatment
40
systems.2,3,6 Such MD systems may be powered by an unlimited supply of solar energy,
41
thus provide remote communities that are not connected to the national water and
42
energy grids with potable water and reduce critical water scarcity.
43
Similarly to all membrane systems, treatment of contaminated feed water via MD
44
induces the challenge of fouling and especially microbial fouling (herein termed
45
biofouling).3,7,8 Biofouling develops as bacteria and organic material, ubiquitous in natural
46
and engineered systems, form biofilms on membrane and spacer surfaces.9–11 Biofilm
47
formation follows a phased paradigm, in which individual cells or planktonic bacterial
48
clusters (protobiofilms) adhere to an organic conditioning film.12 Following irreversible
49
attachment, bacteria and protobiofilms develop into a multi-layered biofilm structure,
50
comprising live and dead cells encased in a protective matrix of self-produced
51
extracellular polymeric substances (EPS).11,13 Despite substantial research on
52
membrane biofouling, very little is known about biofilm formation and community
53
composition in MD systems.14,15 In addition to factors that influence biofilm formation in
54
other membrane systems, the elevated feed water temperature is unique to MD.8,14,16
55
Feed water temperature of MD systems can range between 40 °C to 80 °C,3,17 leading to
56
a shift in bacterial community composition due to thermally induced succession, after
57
which thermophilic and heat resistant bacteria will constitute the majority of the bacteria
58
population.14,15,18,19
59
MD biofilm can reduce the vapor pressure driving force by hindering mass and heat
60
transport, consequently enhancing temperature and salt concentration at the membrane
61
surface.15,20,21 The hydrophilic nanoporous structure of the biofilm can further affect the 3 ACS Paragon Plus Environment
Environmental Science & Technology
62
driving force via water vapor pressure depression.21,22 Concomitantly, depending on the
63
biofilm properties (i.e., thickness, density, organic composition, etc.), hydraulic
64
resistance can also impair water transfer through the biofilm.21,23 Another fouling effect,
65
that is unique to MD, is membrane wetting, which refers to the penetration of
66
contaminated feed water through the pore structure into the distillate.15,24,25 Among other
67
mechanisms, wetting can be caused by a reduction of membrane surface hydrophobicity
68
via attachment of surface active compounds in the feed water.15,20,24 Some surface
69
active compounds, termed biosurfactants, can be actively secreted to the feed water as
70
amphiphilic EPS components by bacteria.15,25
71
In this study we identified the effects of feed water temperature on MD biofouling with a
72
monoculture of a thermophilic Anoxybacillus sp. at three feed water temperatures (47
73
°C, 55 °C and 65 °C) in a direct contact MD system containing feed and distillate
74
spacers. System performance was analyzed in real time over three days via distillate
75
water flux and distillate water quality measurements. In addition, biofilm properties were
76
characterized at the end of each experiment. Our results highlight that selection of the
77
appropriate feed water temperature is crucial for the success of MD application in
78
wastewater treatment.
79
MATERIALS AND METHODS
80
Bench-scale MD system. A custom-made bench-scale direct contact MD system with a
81
transparent polycarbonate cross flow cell was used for dynamic biofouling experiments
82
(Figures S1 and S2). Feed and distillate temperature were controlled by a recirculating
83
heater (Polyscience) and chiller (KNF Lab). Flow conditions were held constant using
84
two gear pumps with Masterflex® pump heads (Cole-Parmer). Distillate water flux was
85
monitored by weighing the distillate solution on a digital scale (Phoenix instrument)
86
interfaced with a data acquisition system. Additionally, values of feed (Sensorex) and 4 ACS Paragon Plus Environment
Page 4 of 35
Page 5 of 35
Environmental Science & Technology
87
distillate conductivity (Eutech Instruments, Thermo Scientific) as well as temperature
88
(DS18B20 Digital Temp Sensor, Adafruit) in the reservoirs and at in-/outlets of the cross
89
flow cell were collected every minute via a microprocessor (Arduino UNO). Further
90
details and experimental parameters are available in the Supporting Information.
91
Hydrophobic Durapore® membranes with nominal pore size 0.45 µm (HVHP, Merck
92
Millipore, Ireland) were used as MD membranes. The active membrane area in the cross
93
flow cell was 6.5 x 3.5 cm2 and spacers (46 mils = 1.2 mm, Conwed®, USA) determined
94
channel height on feed and distillate side. The spacer had a diamond shaped pattern
95
comprised of slightly hydrophobic (contact angel of ~94°) polypropylene filaments,
96
molded together in two layers. Filaments were oriented at a 45° angle towards the flow
97
direction (Figure S1, insert).
98
Bacteria strain and feed water medium. Thermophilic, Gram-positive, spore forming
99
Anoxybacillus sp. was isolated from effluent of an operational wastewater membrane
100
bioreactor (MBR). An MD biofouling experiment (see details below) was carried out with
101
effluent from the MBR for one day at a feed water temperature of 55 °C. Bacteria were
102
removed from the biofouled MD membrane with mild sonication and plated on Luria–
103
Bertani broth agar (LB, Becton, Dickinson and Company). Single colonies were isolated,
104
grown overnight (~12h) in LB and plated again. Isolation process was repeated three
105
times before genomic DNA was extracted from sub-samples using a phenol chloroform
106
approach.14 Amplification of the V1–V3 regions of the bacterial 16S rRNA was achieved
107
with a 28F and 519R primer pair 26 and the bacterium strain was identified by sanger-
108
sequencing (Hy-labs, Israel), with high confidence (99% similarity to the Silva database)
109
as an Anoxybacillus sp. (and most likely Anoxybacillus gonensis).
5 ACS Paragon Plus Environment
Environmental Science & Technology
110
Artificial sterile wastewater medium was prepared fresh for each experiment according
111
to Bar-Zeev et al. 2014,27 comprising 8.0 mM NaCl, 0.15 mM MgSO4, 0.50 mM NaHCO3,
112
0.20 mM CaCl2, 0.20 mM KH2PO4, 0.40 mM NH4Cl, and 0.6 mM Na3C6H5O7 (sodium
113
citrate). Each salt was dissolved in 60 mL double distilled water (DDW), individually
114
autoclaved and then added to sterile DDW (10 L). The resulting feed water medium had
115
a conductivity of 1 mS cm-1 and pH 7.
116
Biofouling procedure. Prior to each experiment, the MD system was thoroughly
117
cleaned in three steps by passing 10% bleach, Ethylenediaminetetraacetic acid solution
118
(EDTA, 5 mM) at pH 9 and 90% ethanol through the system for 30 min each. Between
119
steps, the system was rinsed twice for 5 min with DDW. Finally, the system was
120
thoroughly rinsed five times with DDW to remove any traces of the cleaning agents
121
before the membrane and spacers were inserted into the cross flow cell.
122
Initially, baseline (control) experiments were carried out for three days with artificial
123
sterile wastewater (10 L) as feed water and without bacterial inoculation to determine the
124
distillate water flux and salt passage at the different temperatures. Distillate temperature
125
was adjusted to reach the same initial distillate water flux of 23 ± 2 L m-2 h-1 for all three
126
feed water temperatures (further details on the operating parameters are provided in
127
Table S1). Every 24 h, collected distillate water was removed from the distillate
128
reservoir.
129
For each biofouling experiment a fresh and concentrated bacterial suspension was
130
prepared by inoculating 300 mL of LB broth with Anoxybacillus sp. and letting it grow
131
overnight to a mid-exponential stage (OD600 of ~0.5). Supernatant was removed and
132
replaced with sterile artificial wastewater after short centrifugation (4000 rpm for 20 min).
133
The feed reservoir was inoculated with the bacteria culture (~7x107 cells mL-1) once the
6 ACS Paragon Plus Environment
Page 6 of 35
Page 7 of 35
Environmental Science & Technology
134
initial distillate water flux was stable at 23 ± 2 L m-2 h-1. The feed water was constantly
135
aerated by a small air pump and sampled every 24 h to monitor pH and bacterial
136
concentration. At the same time, glucose (Sigma-Aldrich) was supplemented as an
137
additional carbon source to maintain a constant concentration of ~0.55 mM.
138
Real time observation with optical coherence tomography. OCT is an approach that
139
has recently gained importance for in situ non-invasive fouling observation in membrane
140
systems 28–30 and was used in this study to characterize biofilm formation and membrane
141
wetting. The scan lens (Thorlabs, LSM03-BB) fitted on a spectral domain Optical
142
Coherence Tomograph (Ganymede-II, Thorlabs, Germany) was placed above the cross
143
flow cell to allow observation through the viewing window of the top cell half. Cross
144
sectional scans were taken every 24 h between spacer filament crossings and along
145
filaments with a resolution of 2 µm/pixel. Sensitivity of image acquisition was set to
146
medium at an A-scan rate of 15 kHz with averaging over six A-scans.
147
Biofilm characterization. After three days of biofouling membrane and feed spacer
148
were removed from the cross flow cell, avoiding separation or shifting, and dissected into
149
1 cm2 coupons for subsequent analysis.27,31
150
Confocal laser scanning microscopy (CLSM). Imaging was carried out directly at the end
151
of each experiment. SYTO® 9, propidium iodide (LIVE/DEAD BacLight™, molecular
152
probes) and concanavalin A (Alexa Fluor® 633, molecular probes) were used to stain
153
membrane and spacer coupons for identification of live cells, dead cells and extracellular
154
polysaccharides as part of biofilm EPS. Stained coupons were mounted in an
155
unconfined biofilm characterization chamber 27 for image acquisition with an inverted
156
CLSM (LSM 510 META, Carl Zeiss, Inc.). Random Z stacks (7 to 14) with slice thickness
157
of 2 µm were captured for each membrane coupon using an LD C-Apochromat 40x
7 ACS Paragon Plus Environment
Environmental Science & Technology
158
magnification objective. Images were analyzed to obtain biovolume and biofilm thickness
159
of the three biofilm components. For spacer samples, a Plan-Neofluar 10x magnification
160
objective was used to capture CLSM images with 50 µm slice thickness at and between
161
each filament crossing. The resulting eight adjacent images were combined into one
162
image showing one square spacer element. Further details can be found in the
163
Supporting Information.
164
Bacterial abundance (BA). At the end of each experiment membrane and spacer
165
subsamples were re-suspended in 1 mL of sterile wastewater with 2.5 mM EDTA. Cells
166
were removed from samples by short (6 min) bath sonication and were diluted with
167
sterile wastewater. Dilutions were stained according to the manual with LIVE/DEAD™
168
BacLight™ Bacterial Viability and Counting Kit (molecular probes). BA in the distillate
169
was measured by staining with SYBR® Green II RNA gel stain (molecular probes) for
170
total bacteria count. All stained samples were measured at a flow speed of 25 µL min-1
171
using an Attune NxT flow cytometer (life technologies). Representative flow-cytometry
172
plots that exhibit the distinction between live and dead cells as well as total bacteria
173
count can be found in the Supporting Information (Figure S3).
174
Total organic carbon (TOC). Measurements were obtained from membrane and spacer
175
samples placed in acid-cleaned glass vials with 7 mL DDW and 40 µL 2 M HCl. Probe
176
sonication (Sonicator Model CL-188, QSONICA) for 3 min (six 30 s cycles) was required
177
to disperse cells from membrane and spacer surface. A TOC analyzer (Multi N/C,
178
Analytik-Jena, Germany) measured TOC using a five point calibration procedure.
179
Biofilm density. Organic carbon density was calculated by dividing TOC per area by total
180
biovolume per area obtained from the biofilm components in CLSM and is given in fg
8 ACS Paragon Plus Environment
Page 8 of 35
Page 9 of 35
Environmental Science & Technology
181
µm-3. Additionally, cell density was calculated by dividing BA per area by total biovolume
182
per area, given in cells µm-3.
183
Transmission (TEM) and scanning (SEM) electron microscopy. Samples were fixed,
184
stained and imaged according to Bar-Zeev et al. 2015 32. Briefly, membrane and spacer
185
coupons were fixed at 4 °C overnight with 2.5% glutaraldehyde fixation buffer (0.1 M
186
Cacodylate, 0.1% CaCl2, pH 7.2). Staining of bacterial cells was done with osmium
187
tetroxide followed by sequential dehydration steps. For TEM, membrane samples were
188
then infiltrated with epoxy resin at room temperature and cut into 70 nm slices before
189
imaging with an FEI Tecnai T12 G2 TWIN transmission electron microscope operating at
190
120 kV. For SEM, membrane samples after dehydration were freeze-fractured and then
191
both membrane and spacer samples were sputter coated with gold and imaged with an
192
JSM-7400F (JEOL) ultrahigh resolution cold FEG-SEM at an acceleration voltage of 3
193
KeV. Further details are available in the Supporting Information.
194
Statistical analysis. Distillate water flux values collected every minute were aggregated
195
into 4 h rolling averages. Averages were normalized to the last hour before inoculation,
196
as well as to the corresponding hour of two averaged baseline experiments (Figure S4).
197
Distillate salinity was averaged in the same manner, but normalized to the initial salinity
198
of the distillate, equaling DDW. TOC and BA were displayed as means with
199
corresponding standard deviations, obtained from three or four biologically different
200
experiments. Biofilm parameters were normalized to the area of membrane or spacer
201
coupon, measured with ImageJ software. The spacer area was measured as the area
202
covered and not the surface area of the spacer filaments, which the supplier estimated
203
to be 0.98 m2 m-2. Statistical analysis was conducted in Excel with the XLSTAT Add-In
204
using ANOVA pairwise comparison at a significance level of p