Improved method for measuring transparent exopolymer particles (TEP) and their precursors in fresh and saline water

w a t e r r e s e a r c h 7 0 ( 2 0 1 5 ) 3 0 0 e3 1 2

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/watres

Improved method for measuring transparent exopolymer particles (TEP) and their precursors in fresh and saline water Loreen O. Villacorte a,b,*, Yuli Ekowati a, Helga N. Calix-Ponce a, Jan C. Schippers a, Gary L. Amy a,c,d, Maria D. Kennedy a,c a

UNESCO-IHE Institute for Water Education, Westvest 7, 2611 AX Delft, Netherlands Wetsus Center of Excellence for Sustainable Water Technology, Agora 1, 8934 CJ Leeuwarden, Netherlands c Delft University of Technology, Stevinweg 1, 2628 CN Delft, Netherlands d Water Desalination and Reuse Research Center, 4700 Kaust, Thuwal, Saudi Arabia b

article info

abstract

Article history:

Transparent exopolymer particles (TEP) and their precursors produced by phyto-/bacterio-

Received 31 July 2014

planktons in fresh and marine aquatic environments are increasingly considered as a

Received in revised form

major contributor to organic/particulate and biological fouling in micro-/ultra-filtration

29 November 2014

and reverse osmosis membrane (RO) systems. However, currently established methods

Accepted 6 December 2014

which are based on Alcian blue (AB) staining and spectrophotometric techniques do not

Available online 16 December 2014

measure TEP-precursors and have the tendency to overestimate concentration in brackish/ saline water samples due to interference of salinity on AB staining. Here we propose a new

Keywords:

semi-quantitative method which allows measurement of both TEP and their colloidal

Transparent exopolymer particles

precursors without the interference of salinity. TEP and their precursors are first retained

(TEP)

on 10 kDa membrane, rinsed with ultra-pure water, and re-suspended in ultra-pure water

Acid polysaccharides

by sonication and stained with AB, followed by exclusion of TEP-AB precipitates by filtra-

Algal blooms

tion and absorbance measurement of residual AB. The concentration is then determined

Algal organic matter

based on the reduction of AB absorbance due to reaction with acidic polysaccharides, blank

Water treatment

correction and calibration with Xanthan gum standard. The extraction procedure allows

Membrane fouling

concentration of TEP and their pre-cursors which makes it possible to analyse samples with a wide range of concentrations (down to <0.1 mg Xeq/L). This was demonstrated through application of the method for monitoring these compounds in algal cultures and a full-scale RO plant. The monitoring also revealed that concentrations of the colloidal precursors were substantially higher than the concentration of TEP themselves. In the RO plant, complete TEP removal was observed over the pre-treatment processes (coagulationsedimentation-filtration and ultrafiltration) but the TEP precursors were not completely removed, emphasising the importance of measuring this colloidal component to better understand the role of TEP and acidic polysaccharides in RO membrane fouling. © 2014 Elsevier Ltd. All rights reserved.

* Corresponding author. UNESCO-IHE Institute for Water Education, Westvest 7, 2611 AX Delft, Netherlands. Tel.: þ31 152151788. E-mail addresses: [email protected], [email protected] (L.O. Villacorte). http://dx.doi.org/10.1016/j.watres.2014.12.012 0043-1354/© 2014 Elsevier Ltd. All rights reserved.

w a t e r r e s e a r c h 7 0 ( 2 0 1 5 ) 3 0 0 e3 1 2

1.

Introduction

The presence of transparent exopolymer particles (TEP) in aquatic environments has been extensively studied over the past two decades, especially in the field of oceanography and limnology. As the name suggests, TEPs are transparent organic substances seasonally abundant (e.g., algal blooms) in marine and fresh water environments (Passow, 2002a). They are generally amorphous and have been observed in various shapes (e.g., strings, disks, sheets or fibres) and sizes, ranging from few nanometres in diameter up to 100s of mm long (Passow, 2000). In the ocean, TEPs mainly originate from exudates and detritus of phytoplankton (micro-algae) and bacterioplankton but they may also originate from some species of macro-algae, oysters, mussels, scallops and sea snails (Passow, 2002b; McKee et al., 2005; Heinonen et al., 2007). TEPs are sticky and comprise mainly hydrophilic, negatively-charged, acidic polysaccharides (Mopper et al., 1995). Moreover, they can be associated with or they tend to absorb proteins, lipids, trace elements and heavy metals from the water (Passow, 2002a). This makes them a nutritious platform and hotspots for bacterial activity (Berman and Holenberg, 2005; Mari and Kiørboe, 1996). Such characteristics led some researchers to suspect that they may have an important role in the formation of aquatic biofilms (Alldredge et al., 1993; Passow, 2002a; BarZeev et al., 2012). In 2005, Berman and Holenberg proposed the potential role of TEP as a major initiator of biofilm leading to biofouling in reverse osmosis (RO) membranes (Berman and Holenberg, 2005). Consequently, various studies were conducted to investigate the link between TEP and biofouling (Bar-Zeev et al., 2009; Villacorte et al., 2009a,b). Moreover, various studies demonstrated that TEPs can cause organic fouling in micro-/ultra-filtration membrane systems (Kennedy et al., 2009; Villacorte et al., 2010a,b; 2013; Schurer et al. 2012, 2013;

301

Alizadeh Tabatabai et al., 2014) and membrane bioreactors (de la Torre et al., 2008). However, such studies applied different methods to quantify TEP and vary in terms of calibration with a standard and size range being considered as TEP, making comparison rather difficult (Discart et al., 2014; Filella, 2014; Villacorte et al., 2009b). Marine biologists operationally defined TEPs as particles larger than 0.4 mm considering that they were first discovered through retention on 0.4 mm pore size membrane filters (Alldredge et al., 1993). However, TEPs are not solid particles, but rather agglomeration of particulate and colloidal hydrogels which can vary in size from a few nanometres to hundreds of micrometres (Passow, 2000; Verdugo et al., 2004). They are highly hydrated, comprising more than 99% water, which means they can bulk-up to more than 100 times their solid volume (Azetsu-Scott and Passow, 2004; Verdugo et al., 2004). TEPs can be directly released through sloughing of algal cell coatings (Kiørboe and Hansen, 1993) or through disintegration of large algal colonies (Passow and Wassmann, 1994). TEP may also originate from colloidal polymers (1e10 kDa) which spontaneously form free fibrils through alignment on hydrophobic surfaces, micro-hydrogels through annealing and gelation, and eventually TEP (>0.4 mm) through aggregation (Verdugo et al., 2004; Chin et al., 1998). These sub-micron components (<0.4 mm) which have similar chemical properties as TEP are collectively known as TEP precursors (Passow, 2000). Since the discovery of TEPs two decades ago, various quantification methods have been developed, all of which are based on staining with Alcian Blue dye. This particular dye is known to be highly selective and forms insoluble complexes with target compounds that cannot be easily reversed by subsequent treatments. The dye is also widely available and has been routinely used in medical and biological research for many years. Nevertheless, despite being one of the most widely-used biological stains, the mechanisms involved during reaction of the dye with a specific substrate is still not well understood (Horobin, 1988). An Alcian Blue (AB) molecule is a

Fig. 1 e Molecular structures of Alcian blue (AB) and acidic polysaccharide Xanthan gum (XG). The inset image is an optical microscope photograph of precipitates formed after reaction between AB and XG at pH 2.5.

302

w a t e r r e s e a r c h 7 0 ( 2 0 1 5 ) 3 0 0 e3 1 2

tetravalent cation with a copper atom at the centre of its core (Scott et al., 1964; Fig. 1). In aqueous solutions without extra electrolytes, AB can specifically bind with anionic carboxyl, phosphate and half-ester sulphate groups of acidic polysaccharides, resulting in the formation of neutral precipitates (Ramus, 1977; Parker and Diboll, 1966; Fig. 1). AB can also react with carbohydrate-conjugated proteins such as acidic glycoproteins (Wardi and Allen, 1972) and proteoglycans (Bartold and Page, 1985) but not nucleic acids and neutral bio€ rnsson, 2001). polymers (Karlsson and Bjo The staining ability of AB depends on the type and density of anionic functional groups associated with the biopolymer materials in the sample. The staining selectivity largely depends on the pH and ionic strength of the AB and sample solution (Horobin, 1988). The staining ability of AB was reported to decrease when pH is lower or higher than 2.5 (Passow and Alldredge, 1995). In high ionic strength solutions, better interactions between anionic polymers and cationic AB can be expected due to compression of the electrical double layer surrounding the AB molecule. Consequently, AB molecules spontaneously aggregate in saline solutions resulting in the formation of precipitates not associated with TEP. This is considered as the main drawback of the application of AB staining for TEP measurements in seawater. To minimise measurement artefacts due to coagulation, AB staining solutions should always be pre-filtered and should not be directly applied to solutions with high salinity. Four different methods were developed to identify and quantify TEP and their precursors over the last two decades, (i.e., Alldredge et al., 1993; Passow and Alldredge, 1995; Arruda-Fatibello et al., 2004; Thornton et al., 2007). Despite these developments, their application to water treatment monitoring has not been extensively studied. As TEP is gaining more attention in water research and the membrane-based desalination industry, there is an increasing need for a reliable method to measure these substances in fresh and marine (saline) waters. The aim of this study was to develop a reliable TEP method applicable for water treatment process monitoring and membrane fouling studies. The investigation mainly focused on the acid polysaccharide method developed by Thornton et al. (2007), because it is, in principle, capable of measuring both TEP and their colloidal precursors and applicable for saline and freshwater samples. Specific investigation was conducted on the effect of salinity on AB staining and how this can be mitigated during TEP measurement. Furthermore, the application of the improved method for TEP and precursor monitoring in algal cultures and in full-scale water treatment plant was also performed.

simulate seawater impacted by severe algal bloom as it has been reported to generate significant concentrations of TEP (Myklestad et al., 1989; Passow, 2002a). A strain of C. affinis (CCAP 1010/27) was inoculated in sterilised f/2 þ Si medium (Guillard and Ryther, 1962) in artificial seawater. The culture was maintained until it reaches the stationary phase and then the AOM solution was extracted by sedimentation and then filtration of the supernatant through 5 mm pore size filter following the protocol by Villacorte et al. (2013). Various untreated samples were also collected from batch cultures of C. affinis, Alexandrium tamarense (CCAP 1119/32) and Microcystis sp. (CCAP 1450/13). Samples for TEP analysis were collected every 2e3 days within the 30e60 day monitoring period.

2.2.

Blank and model solutions

Ultrapure water (UPW) was used as the blank solution for TEP measurements and base solution of all laboratory prepared water samples. UPW was derived from tap water purified through a series of treatment steps, namely: 1 mm filtration > softening > reverse osmosis demineralisation > ion exchange > 1 mm filtration > granular activated carbon filtration > ultraviolet (UV) disinfection > 0.22 mm filtration. Artificial seawater (ASW) used in saline water blank measurements was prepared based on the typical inorganic ion concentration of coastal North Sea water (see Supplementary data S2). Minor constituents which comprised less than 1% of inorganic ions in seawater were not added. To make up the ASW solution, J.T. Baker analytical grade salts (Na2CO3, NaHCO3, CaCl2.2H2O, KCl, Na2SO4, MgCl2.6H2O, NaCl) with >99% purity were sequentially dissolved in ultrapure water. Standard solutions were prepared from purified Xanthan gum (SigmaeAldrich). The solution was prepared by adding 50 mg Xanthan to 500 mL of UPW while rapidly stirring with a magnetic stirrer. Rapid stirring was maintained for at least 30 min until no flocs were visible. The resulting solution was further homogenised 3 times using a tissue grinder (Dounce, SigmaeAldrich).

2.3.

Alcian blue dye

2.

Materials and methods

Stock solutions of the dye were prepared in acetic acid solution by dissolving 250 mg/L of Alcian Blue (Standard Fluka, SigmaeAldrich). The base solution was prepared by adding drops of acetic acid until the solution pH of 2.5 is reached. The prepared stock solutions were then stirred for 12e18 h. The working solution was prepared everyday by filtering a part of the stock solution through 0.05 mm polycarbonate membranes (Nuclepore, Whatman) before staining. The remaining stock solution was stored in the dark at 4  C and a new stock solution was prepared after 4 weeks.

2.1.

Water samples

2.4.

Water samples were collected from the untreated source water and over the different process steps of a drinking water treatment plant in the Netherlands in the summer of 2012. Algal organic matter (AOM) solution extracted from a batch culture of Chaetoceros affinis was used in the investigations to

TEP methods

This article describes the development of a new semiquantitative method to measure TEP and their precursors partially based on the principles developed by Thornton et al. (2007) for measuring acidic polysaccharides. The results based on the new method were compared with those measured

w a t e r r e s e a r c h 7 0 ( 2 0 1 5 ) 3 0 0 e3 1 2

using the established TEP method by Passow and Alldredge (1995) with some modifications. For simplicity, the new method and method by Passow and Alldredge (1995) are henceforth denoted as TEP10kDa and TEP0.4mm, respectively. By definition, TEP0.4mm measures TEP while TEP10kDa measures both TEP and their precursors.

2.4.1.

TEP0.4mm measurement

The procedure for measuring TEP0.4mm is illustrated in the supporting data (Figure S3.1). Firstly, water sample was filtered through 47 mm diameter polycarbonate filters (0.4 mm pore size) by applying a constant vacuum of 0.2 bar. Two ml of UPW was then filtered through the filter-retained TEP by applying <0.2 bar vacuum to wash-out the remaining sample moisture through the filter. Pre-filtered AB dye solution (1 mL) was subsequently applied over the filter, allowed to react with TEP for 10 s, and then the un-reacted dye was flushed through the filter by vacuum filtration (<0.2 bar). To further remove the remaining un-reacted dye, a rinsing step was performed by filtering 2 ml of ultra-pure water through the filter. The rinsed filter was transferred to a 50 ml glass beaker and soak in 6 ml of 80% sulphuric acid solution. The beaker was covered with parafilm and gently mixed on a shaker for 2 h. The acid solution was then transferred to a 1 cm cuvette and absorbance (At) was measured in a spectrophotometer (Shimadzu UV-2501PC) at 787 nm wavelength - the wavelength of maximum absorbance of Alcian blue (AB) when dissolved in sulphuric acid. The concentration of TEP in the water sample was calculated using Eq. (1).

TEP0:4mm ¼

At  Af  As m787 Vf

(1)

where TEP0.4mm is TEP concentration in terms of mg Xeq/L; the total absorbance (At) is the absorbance of the dye which reacted with TEP and those adsorbed to the filter (abs/cm); filter blank absorbance (Af) is the absorbance of the dye adsorbed to the filter (abs/cm); sample correction (As) is the absorbance of unstained sample (abs/cm); Vf is the volume of sample filtered (L) and m787 is the slope of the calibration curve [(abs/cm)/mg Xeq]. The filter blank (Af) was measured in the same way as total absorbance but filtering TEP-free blank samples (e.g., synthetic water with similar ion concentration as the water sample) instead of actual water samples. For sample correction (As), water samples were filtered in the same way as determining the total absorbance but skipping the AB staining procedure (Passow et al., 2001). The slope, m787 is derived from calibration experiments where the mass of the standard (Xanthan gum) is plotted against the corresponding absorbance of AB which reacted. We conducted several calibration experiments based on the protocol by Passow and Alldredge (1995) without reliable results. A critical analysis of the protocol (Supplementary data S1) indicates that artefacts can be introduced during the drying (e.g., dust contamination) and weighing (e.g., electrostatic forces interference) steps. Consequently, we developed a new calibration protocol as a proposed alternative. The proposed procedure is described in Section 2.4.3.

2.4.2.

303

TEP10kDa measurement

The procedure for measuring TEP10kDa is illustrated in the supporting data (Figure S3.2). Firstly, water sample was filtered through 10 kDa membranes (Millipore RC membrane, 25 mm diameter) at constant flux (60 L/m2/h) using a syringe pump (Harvard syringe pump 33). After filtering a specific volume (10e100 ml) of sample, the syringe was replaced with a clean syringe containing about 10 ml of air. Air was then injected to the filter holder (60 L/m2/h) until all the remaining water in the feed side of the membrane holder has passed through the membrane. The total filtered sample volume was then measured after collecting the filtrate. To rinse out saline moisture, 5 ml of UPW was injected to the filter holder at 60 L/ m2/h. Air is again injected until all the rinse water on the feed side of the membrane holder has passed through the membrane. The membrane was then carefully removed from the filter holder and placed feed side down in a clean disposable plastic container (40 ml Unipot with screw cap) containing 10 ml of UPW. The sample was then tightly covered, vortexed (Heidolph REAX 2000) for 10 s and sonicated (Branson 2510EMT) for 60 min. Four ml of the re-suspended TEP þ precursor solution was transferred to a clean 20 ml disposable plastic container. To adjust the sample pH to 2.5, 0.05 ml of acetic acid solution was added to the solution (verified with a pH metre). One ml of pre-filtered (through 0.05 mm PC membrane) AB solution was added to the sample, which was then mixed vigorously and left to react for 10 min. Four ml sample of the TEP-AB solution is then filtered through a 0.1 mm PC filter by vacuum filtration (<0.2 bars). The filtrate was collected in a clean plastic container (10 ml Unipot with screw cap), transferred to a 1-cm cuvette and absorbance (Ae) was measured in a spectrophotometer (Shimadzu UV-2501PC) at 610 nm wavelength e the wavelength of maximum absorbance (visible light range) of AB when dissolved in acetic acid solution (pH ¼ 2.5). The TEP10kDa concentration in mg Xanthan equivalent per litre (mg Xeq/L) was calculated as follows: TEP10kDa ¼

1 Vr ðAe  Ab Þ m610 Vf

(2)

where m610 is the slope of the calibration line [(abs/cm)/(mg Xeq/L)], Vr is total volume of re-suspended TEP þ precursor solution (10 mL), Vf is the volume of filtered sample (mL), Ae is the absorbance of the excess or un-reacted dye (abs/cm) and Ab is the absorbance of filtered blank (abs/cm). The calibration experiment to obtain the slope (m610) is described in Section 2.4.3. The blank absorbance (Ab) should be measured to correct for the amount of stain adsorbed by the polycarbonate filter. This was performed following the above-mentioned procedure but replacing the sample with UPW.

2.4.3. Integrated calibration protocol for TEP0.4mm and TEP10kDa For better comparison, the procedure to obtain the two calibration slopes (m787 and m610) to express TEP0.4um and TEP10kDa concentrations in mg Xeq/L was integrated into one experiment. Firstly, homogenised standard solutions (4 ml) containing different concentrations (0e5 mg/L) of Xanthan gum were prepared from a stock solution (Section 2.3). For pH adjustment, 0.05 ml acetic acid was added to each solution

304

w a t e r r e s e a r c h 7 0 ( 2 0 1 5 ) 3 0 0 e3 1 2

and then briefly agitated. The solution was then stained by adding 1 ml of pre-filtered AB staining solution, mixed for 10 s and incubated for 10 min. Four ml of the resulting solution is then filtered through 0.1 mm PC membrane by vacuum filtration (0.2 bars). The filtrate was collected, transferred to 1-cm cuvette and absorbance was measured at 610 nm. The PC membrane used to filter AB-stained standard solution was carefully transferred to a 50 ml beaker. Six ml of 80% sulphuric acid solution was added, covered with parafilm and mixed on a shaker for 2 h. Afterwards, the acid solution was transferred to a 1-cm cuvette and absorbance was measured at 787 nm. To determine the calibration slope (m787) for TEP0.4mm, the mass of Xanthan gum retained on the PC membrane was calculated by multiplying the volume filtered (4 ml) by the concentration of Xanthan in the stained standard solution. The calculated mass was then plotted against the corresponding AB absorbance measured at 787 nm wavelength, whereby the average linear slope is the m787. To determine the calibration slope (m610) for TEP10kDa, the Xanthan concentration of the stained dye was plotted against the corresponding absorbance measured at 610 nm wavelength (excess dye absorbance) and the average linear slope is the m610. Since concentration is inversely proportional to the excess dye absorbance, the calibration slope (m610) has a negative value.

2.5.

Membrane filters and filtration set-up

Track-etched 47 mm diameter polycarbonate membranes (Nuclepore, Whatman) with nominal pore sizes of 0.4, 0.2, 0.05, 0.03, 0.015 mm were used for experiments involving vacuum filtration. Regenerated cellulose (RC) membranes with 10 kDa and 5 kDa MWCO (Millipore) were used for constant flux syringe filtration experiments. To remove possible contaminants, the PC filters were rinsed by flushing >200 ml of UPW through it while RC membranes were soaked for at least 24 h in UPW and then flushed with 5e10 ml of UPW prior to sample filtration. For TEP0.4mm measurement, the vacuum filtration set-up comprised of a 50 mm glass filter holder (Sartorius) with fibre glass porous support and a vacuum pump (Millipore

WP612205) with pressure controller. For TEP10kDa measurement, a set-up comprising a 25 mm glass filter holder (Sartorius) with PTFE coated stainless steel mesh filter support and a vacuum pump with pressure controller was used to separate stained TEP from excess AB. In order to minimise sample contamination, the filter holder was thoroughly cleaned (after every filtration) by rinsing with ultrapure water. Constant flux filtration (60 L/m2.h) through RC membranes was performed using a syringe pump (Harvard Pump 33), 60 ml disposable syringe (BD Plastipak™) and 25 mm filter holder (Schleicher & Schuell).

2.6.

Limit of detection calculation

The lower limit of detection (LODmin) of the proposed TEP þ precursor method (TEP10kDa) was determined based on the variability of the blank absorbance. This was calculated as follows: LODmin ¼

3sb Vr $ m610 Vf

(3)

where sb is the standard deviation of 10 independently measured blank absorbance (abs/cm). The factor 3 corresponds to a significance level of 0.00135, which means that only 0.135% of blank measurements will statistically yield results that fall above the computed detection limit (Harvey, 2000). Note: The correction for the blank is already included in the concentration calculation; hence, the average blank is assumed to be zero and was not added to the LODmin.

2.7.

Dynamic light scattering measurements

The hydrodynamic size distribution of AB staining solutions was estimated based on dynamic light scattering (DLS) measurements using a Malvern Zetasizer Nano ZS. The DLS technique measures the diffusion of particles moving under Brownian motion and converted to size based on the StokeseEinstein relationship. The obtained size is the diameter of a sphere with equivalent translational diffusion coefficient as the measured particle, called the hydrodynamic diameter. All measurements were performed at 25  C.

Fig. 2 e (a) Apparent size distribution of 0.025% (m/v) Alcian blue in acetic acid solution (pH 2.5) based on serial filtration through polycarbonate membranes with different (decreasing) pore sizes and (b) hydrodynamic size distribution based on dynamic light scattering technique of AB solution pre-filtered through 0.05 mm PC membrane.

w a t e r r e s e a r c h 7 0 ( 2 0 1 5 ) 3 0 0 e3 1 2

305

Fig. 3 e (a) Absorption spectra of Alcian blue 8GX (16 mg AB/L) dissolved in sulphuric acid and acetic acid solution and (b) peak absorbance values of Alcian blue (AB) solutions at various concentrations.

2.8. Liquid chromatography organic carbon detection (LC-OCD) Selected water samples were analysed using liquid chromatography e organic carbon detection (LC-OCD) at DOC-Labor (Karlsruhe, Germany) as described by Huber et al. (2011). LCOCD analyses were performed without in-line 0.45 mm prefiltration to cover part of the TEPs (>0.4 mm). Based on the pore size of the sinter filters in the chromatogram column, the theoretical maximum chromatographable size of organic substances without sample pre-filtration is 2 mm (S. Huber, per. com.). TEP and their precursors are a type of biopolymers, the specific organic carbon content of which may vary substantially in different water sources. Hence, biopolymer concentration is not considered as a direct measure of TEP þ precursors but was used in this study as a relative indicator of the amount of TEP and their colloidal precursors in water samples.

3.

Results and discussion

3.1.

Alcian blue solution pre-treatment

Alcian blue (AB) is not a stable dye and tends to form colloidal suspension when dispersed in water (Passow and Alldredge, 1995). The size distribution of a stock AB solution (250 mg/L at pH 2.5) was determined by serial filtration through different pore size membranes (5e0.015 mm PC filters) and dynamic light scattering technique. It was found that about 50% of AB was smaller than 0.2 mm and about 44% was smaller than 0.05 mm (Fig. 2a). Further filtration of the dye through 0.03 mm and 0.015 mm pore size filters only allowed passage of about 30% and 2% of the initial dye concentration, respectively. Based on the results, using a 0.03 mm filters for the pretreatment of AB dye solution is promising. However, it may be necessary to increase the initial dye concentration to avoid possible under-staining of TEP and their precursors. Moreover, filtration through such filters can be very time consuming and prone to membrane damage/leakage during vacuum filtration. Consequently, pre-filtration of AB staining solutions through 0.05 mm filters was adopted. Dynamic light

scattering analysis confirmed all AB suspensions after prefiltration through 0.05um PC filter were smaller than 0.05 mm, about 70% of which were between 20 and 40 nm (Fig. 2b). Considering that AB solutions are pre-filtered through 0.05 mm filters, the dye pigments used for staining can be as large as 0.05 mm. To minimise the possibility of retaining these materials, membrane filters used for separating TEP-AB precipitates and excess (unreacted) stain should have pore size much larger than 0.05 mm.

3.2.

Absorption spectra of Alcian blue

The two spectrophotometric methods investigated in this study are based on the absorbance of AB dissolved in different matrices, namely: sulphuric acid solution for TEP0.4mm and acetic acid solution for TEP10kDa. Spectral scans of AB in these matrices were performed and results are presented in Fig. 3a. The maximum absorbance of AB in 80% sulphuric acid solution was at 787 nm wavelength while the maximum absorbance of AB in acetic acid solution (pH 2.5) within the visible light spectrum was at 610 nm wavelength. The AB spectra are consistent with what was reported previously (Ramus, 1977; Passow and Alldredge, 1995). For comparison, the absorbance value of AB in sulphuric acid at 787 nm was about twice that of AB of similar concentration dissolved in acetic acid solution at 610 nm (Fig. 3b).

3.3.

Effect of sample salinity and rinsing

In the new method, AB staining is applied after extracting acid polysaccharides from 10 kDa membranes in ultra-pure water by sonication. This is done in order to minimise interference by salts in the water sample. To demonstrate the effect of salinity, pre-filtered AB solution (1 ml) was added to solutions (4 ml) prepared from different dilutions of ASW (see Section 2.3). Since the dye is strongly cationic, it forms flocs instantaneously after application by reacting with anions in the saline solution. The formed flocs were then removed by filtering the solution through 0.1 mm PC filter. The residual AB was measured based on absorbance of the filtrate solution at 610 nm wavelength.

306

w a t e r r e s e a r c h 7 0 ( 2 0 1 5 ) 3 0 0 e3 1 2

Fig. 4 e (a) Effect on the residual absorbance of Alcian blue (AB) when added to solution with different ionic strength. Residual absorbance is measured after filtering AB stained solution through 0.1 mm polycarbonate filter while net absorbance is the absolute difference between the residual absorbance and the residual absorbance at zero salinity. (b) Relative comparison of TEP10 kDa concentration in saline AOM solution measured with and without ultra-pure water (UPW) rinsing in the measurement procedure.

As presented in Fig. 4a, the residual absorbance of AB was reduced substantially with an increase in ion concentration. For salinities higher than 2 g/L, the measured residual absorbance reduced by at least 50%. These substantial reductions, even at lower salinities, may suggest that AB flocs can form when moisture from saline samples remains on the RC membrane after retention of TEP and their precursors. To remove retained saline moisture from the RC membrane, it is proposed to filter UPW through the membrane before extracting the TEP þ precursors for subsequent AB staining. A comparison of the results with and without rinsing during measurement of AOM samples is shown in Fig. 4b. The concentration was overestimated by a factor of 3 when rinsing was not performed. This illustrates that the rinsing procedure is an essential step to minimise interference of dissolved salts, especially in seawater samples.

3.4.

Method verification

The proposed method (TEP10kDa) is partially based on the method introduced by Thornton et al. (2007). What distinguishes the new method from the latter, specifically when

analysing saline samples, is the extraction of TEP and their precursors, whereby filtration through low molecular weight cut-off (MWCO) UF membrane is used for TEP and precursor retention and then followed by re-suspension in ultrapure water by sonication instead of dialysis treatment. The objective of these steps is primarily to concentrate TEP and their precursors to within detectable levels and to minimise the effect of salinity during the subsequent AB staining. The maximum volume of samples that can be filtered is theoretically unlimited while the retained material is re-suspended in 10 ml solution. Hence, the TEP and precursor level in the sample can be concentrated substantially prior to AB staining to allow reliable measurement in samples with low concentration of these substances. Selecting the optimal membrane pore size to collect TEP and precursors from water samples is an important aspect in the development of the new method as it indicates which size fraction of TEP precursors can be measured. To determine this, measurements were performed using membranes with different pore sizes (0.4 mm, 0.1 mm, 10 kDa and 5 kDa) to retain TEP þ precursors in AOM samples. A comparison of the results is presented in Fig. 5a. As expected, the concentrations

Fig. 5 e (a) TEP measurement in AOM solution from Chaetocers affinis culture with different pore size membranes; (b) rejection of biopolymers in AOM and Xanthan solutions by 0.4 mm, 0.1 mm and 10 kDa membranes. For (a), the TEP result with 0.4 mm membrane was set as the baseline for relative comparison.

w a t e r r e s e a r c h 7 0 ( 2 0 1 5 ) 3 0 0 e3 1 2

307

Fig. 6 e Calibration graphs for two ranges of concentrations of Xanthan gum standard.

increased when using membranes with smaller pore size due to retention of colloidal TEP precursors. The retention of a 0.1 mm PC membrane was about 32% higher than 0.4 mm membrane while 10 and 5 kDa membranes showed about 100% higher retention. LC-OCD analyses were also performed for selected membranes to measure retention of biopolymers from AOM and Xanthan gum solutions. As shown in Figure 5b, 10 kDa membranes retained about 95% of biopolymers from Chaetoceros affinis, which was about twice the retention of a 0.4 mm membrane. Moreover, the retention of Xanthan on 10 kDa UF membrane was about 88% as compared to only about 20% retention on the 0.4 mm membrane. The results of the retention experiments demonstrated that the concentration of colloidal TEP precursors (<0.4 mm) was substantially higher than TEP (>0.4 mm) itself. Consequently, the concentration TEP þ precursors based on the new method is expected to be higher than those measured using the method by Passow and Alldredge (1995) which does not cover the TEP precursors. Considering that no substantial difference in retention was observed between 10 kDa and 5 kDa membranes as well as the high feed pressure required to filter samples through 5 kDa membranes, 10 kDa was eventually selected as the standard membrane pore size for TEP þ precursor measurements based on the new method. However, in principle, sample filtration through different pore size membranes can be used using this method to determine the apparent size distribution of TEP and their pre-cursors in the sample. Further experiments were conducted to establish and verify the TEP þ precursor extraction protocol, staining time, retention of TEP-AB precipitates and effect of sample storage time. The results and findings of these experiments are discussed in the Supplementary data S4.

3.5.

Calibration with Xathan gum

Calibration of TEP10kDa was performed for two ranges of Xanthan concentrations (Fig. 6). The average calibration factor (1/m610) of Xanthan measured for concentration range of 0e5 mg/L was about 30 (mg Xanthan/L)/(abs/cm). The calibration results from 3 experiments performed with different

batches of AB and Xanthan solution showed similar slopes (Fig. 6a). Although the three calibration experiments showed variations in blank absorbance at 0 mg/L, the calibration factors were similar. The calibration factor was observed to decrease to 34 (mg Xeq/L)/(abs/cm) when the concentration range of Xanthan used in the calibration was increased to 0e10 mg/L (Fig. 6b). Linear regression between absorbance and standard concentration showed a coefficient of determination (R2) higher than 0.97. The calibration factors (with Xanthan standard) determined in this study were 5e34% lower than what was observed by Thornton et al. (2007). This might be due to variation of the staining capacity of the AB staining solution, differences in procedure for AB staining, exclusion of TEP-AB precipitates and the membranes used to retain them. Calibration was performed using a 0.1 mm PC membrane instead of a 0.2 mm cellulose acetate used by Thornton et al. (2007). As shown in the Supplementary data Figure S4.2a, a 0.1 mm pore size membrane can retain up to 40% more TEP-AB precipitates than a 0.2 mm pore size membrane; hence, a decrease in calibration factor is expected when shifting to lower pore size membrane.

3.6.

Lower limit of detection

The lower limit of detection of TEP10kDa method was calculated based on the standard deviation of 10 independently measured blanks using the same batch of staining solution. The average absorbance of 10 blank measurements was 0.413 abs/cm and the standard deviation (sb) was 0.01 abs/cm. The LODmin of TEP10kDa is dependent on the filtered sample volume. The typical filtered volume used for natural surface water samples is 50e100 ml, but a much smaller volume is normally sufficient (10e30 ml) for algal culture samples where TEP þ precursor concentration is usually much higher. The LODmin for a wide range of filtered sample volume was calculated according to Eq. (3) in Section 2.7 and the results are presented in Fig. 7. The LODmin for un-concentrated samples (10 ml filtered volume) is 0.91 mg Xeq/L. For concentrated samples (>10 ml filtered volume), the LODmin can be decreased to 0.09 mg Xeq/L for 100 ml filtered volume or 0.05 mg Xeq/L for 200 ml filtered volume. For water samples with very low

308

w a t e r r e s e a r c h 7 0 ( 2 0 1 5 ) 3 0 0 e3 1 2

Fig. 7 e The calculated LODmin of TEP10 kDa as a function of filtered sample volume. The maximum LODmin corresponds to the minimum allowable filtered volume (10 ml).

concentrations (<0.05 mg Xeq/L), filtering larger volume of samples is necessary to ensure concentration is above the LODmin.

3.7.

Method application

The application of the new method has practical advantages over the four existing methods to measure TEP and its precursors (see Supplementary data S1). Microscopic

enumeration can provide information on the size-frequency distribution of TEP in the sample (Alldredge et al., 1993), but it is rather laborious, complicated, time consuming and is not always feasible, especially for samples with low concentration and smaller size range (<2 mm) of TEP. All the succeeding methods based on semi-quantitative spectrophotometric techniques were able to address these issues. The method by Passow and Alldredge (1995) has been widely used in various scientific investigations, but additional time-consuming pretreatment techniques (e.g., bubble adsorption, laminar shear) are needed to measure TEP precursors (Zhou et al., 1998; Passow, 2000). The more recent methods by Arruda-Fatibello et al. (2004) and Thornton et al. (2007) are capable of measuring both TEP and their precursors in one single analysis. However, the former is only applicable in freshwater samples while the latter requires a dialysis step for saline samples. Furthermore, the method introduced by Thornton et al. (2007) is only accurate for samples with high concentration of TEP and their precursors. The proposed method (TEP10kDa) does not only address the major practical issues associated with the previous methods (e.g., salinity, exclusion of TEP precursors) but it also allows measurement of low concentration of TEP with the introduction of a concentration step (i.e., filtration through 10 kDa membrane). As such, it allows analyses of samples with a wide range of TEP (including their precursors) concentrations (down to <0.1 mg Xeq/L). In principle, this method also enables size fractionation of acidic polysaccharides in the water by making use of membranes with different pore sizes during the extraction step. From the perspective of water quality monitoring, the method proposed in this study (TEP10kDa) is not redundant

Fig. 8 e Monitoring of algae, TEP (TEP0.4 mm) and TEP þ precursor (TEP10 kDa) concentrations in batch cultures of (a) Alexandrium tamarense, (b) Chaetoceros affinis and (c) Microcystis sp., and (d) relationship between concentration of TEP and TEP þ precursors.

w a t e r r e s e a r c h 7 0 ( 2 0 1 5 ) 3 0 0 e3 1 2

but rather complimentary to the more established and widely accepted method (TEP0.4mm) by Passow and Alldredge (1995). TEP0.4mm measures TEP while TEP10kDa can measure both TEP and most (if not all) of their colloidal precursors. TEP0.4mm is a more rapid method than TEP10kDa which means it is ideal for routine TEP monitoring in untreated (raw) water. However, to better assess the removal of TEP and their precursors over the treatment processes, TEP10kDa measurement is more appropriate because it covers both TEP and their colloidal precursors. In this study, the integrated calibration method described in Section 2.4.3 was adopted to compare the applications of TEP0.4mm and TEP10kDa methods. The following sections illustrate how the two methods can be integrated and applied for water quality monitoring studies.

3.7.1.

TEP monitoring in algal cultures

TEP and precursor monitoring was performed in batch cultures of two common species of bloom-forming marine algae (Alexandrium tamarense and Chaetoceros affinis) and one species of freshwater cynobacteria (Microcystis sp.). Several samples were collected during the incubation period to measure algal cell concentration (direct cell counting) TEP and their precursors. TEP and precursor concentrations vary substantially for the three algal species at different growth phases (Fig. 8), whereby C. affinis produced up to about 3 times more than A. tamarense and 11 times more than Microcystis. In the three cultures, TEP10kDa concentrations were substantially higher

309

than TEP0.4mm. On average, TEP0.4mm concentrations were about 1311% of TEP10kDa in A. tamarense culture, 2115% in C. affinis culture and 5023% in Microcystis culture. For A. tamarense culture, this percentage varies from 6% during the lagexponential phase to 35% in the stationary-death phase. Such apparent variation was also observed for A. tamarense culture (6e35%) but not with C. affinis (16e19%) culture. These results show that TEP forms mainly in the senescence phase of A. tamarense and Microcystis cultures possibly due to sloughing of cell coatings or disintegration of algal cells, while C. affinis forms TEP mainly through exudation of TEP precursors over different phases (Passow, 2002a). In general, the results obtained from the algal culture monitoring were consistent with what was reported by Thornton et al. (2007), where they observed TEP to comprise between 9 and 60% of acidic polysaccharides (TEP þ precursors) in 2 algal cultures and various samples collected from the Gulf of Mexico. According to Passow (2000), colloidal precursors freshly released by algae are likely fibrillar which tend to form larger colloids and eventually TEP within hours to days. In principle, the proposed method can be used to measure the concentration of the different size fractions of acidic polysaccharides in the water by making use of membranes with different pore sizes. Hence, the proposed method can be used as a tool to better understand the role of TEP precursors in the formation of particulate organic matter and hydrogels in aquatic systems (Chin et al., 1998; Verdugo et al., 2004; Verdugo, 2012).

Fig. 9 e Biopolymer, TEP (TEP0.4 mm) and TEP þ precursor (TEP10 kDa) concentrations (a) in samples collected over the treatment processes of a drinking water treatment plant and (bed) linear regressions between measured parameters. Note: coag ¼ coagulation þ flocculation; sed ¼ sedimentation; RSF ¼ rapid sand filtration.

310

3.7.2.

w a t e r r e s e a r c h 7 0 ( 2 0 1 5 ) 3 0 0 e3 1 2

TEP monitoring in water treatment processes

TEP10kDa measurements were performed in the source water (shallow lake) and over the treatment processes of a full-scale water treatment plant (capacity ¼ 55,200 m3/day). For comparison, TEP0.4mm and biopolymer (LC-OCD analysis) measurements were also performed. Fig. 9a shows the profiles of TEP, TEP þ precursors and biopolymers through the treatment process train of the plant. A substantial reduction of TEP þ precursors and biopolymers were observed after coagulation-sedimentation-sand filtration and ultrafiltration treatments, while RO completely removes all the remaining components. In general, TEP0.4mm removals over the treatment processes were higher compared to removals of TEP10kDa and biopolymer. This can be attributed mainly to the size of TEP0.4mm which, by definition, covers mainly the particulate size range while TEP10kDa and biopolymers cover both the particulate and colloidal size ranges. Moreover, the observed similar reductions of TEP10kDa and biopolymers may indicate that TEP precursors were a significant fraction of biopolymers in the raw water of the plant. Based on linear regression analysis performed using the results from the plant, an indication on the relationship between the three parameters were observed. As shown in Fig. 9, a better correlation was observed between biopolymers and TEP10kDa than with TEP0.4mm. Although biopolymers in surface water may comprise a diverse mixture of compounds including acidic polysaccharides (TEP and precursors), neutral polysaccharides, proteins and lipopolysaccharides, the observed correlation may indicate that TEP and their precursors were the dominant fraction of biopolymers (Hung et al., 2003). On average, TEP10kDa was more than 20 times higher than TEP0.4mm (Fig. 9d), which is a lot higher than what was observed in algal cultures (Fig. 8d). TEP precursors are believed to be fibrillar which can be highly flexible and even reported to pass through membrane pores two orders of magnitude smaller than their apparent size (Passow, 2000). So, it is not surprising to observe less than 100% removal of TEP precursors and colloidal biopolymers even after passing through ultrafiltration membranes with MWCO of about 150 kDa. In desalination plants, it is crucial to remove these precursors from the RO feedwater as they can adhere to the membrane and spacers, enhancing/promoting the formation of biofilm and may eventually cause operational issues in the system due to biofouling (Berman and Holenberg, 2005; Bar Zeev et al., 2012). Hence, the application of this method to investigate the performance of pre-treatment processes for RO can be crucial in developing strategies to minimise biofouling issues in RO plants.

2. A new semi-quantitative method which can measure TEP and their precursors down to 10 kDa (TEP10kDa) in seawater and freshwater without the interference of salinity was developed. This method has an extraction step which makes it possible to analyse samples with a wide range of TEP (including their precursors) concentrations (down to <0.1 mg Xeq/L). In principle, this method also enables size fractionation of acidic polysaccharides in the water by making use of membranes with different pore sizes. 3. The application of the new method (TEP10kDa) was demonstrated in monitoring TEP and precursor concentrations in three batch cultures of bloom-forming algae. The monitoring revealed that concentrations of the colloidal precursors in algal-derived organic matter were substantially higher than the concentration of TEP themselves. 4. Further application of the method was conducted to monitor the fate of TEP and their precursors over the pretreatment processes of a full-scale RO plant. Although TEP was completely removed by the pre-treatment processes (coagulation-sedimentation-filtration) their colloidal precursors were only partially removed, emphasising the importance of measuring this colloidal component to better understand the role of TEP and acid polysaccharides in RO membrane fouling.

Acknowledgements This study was performed at UNESCO-IHE Institute for Water Education with the support of Wetsus, Centre of Excellence for Sustainable Water Technology. Wetsus is funded by the Dutch Ministry of Economic Affairs, the European Union European ^ n, the city Regional Development Fund, the Province of Frysla of Leeuwarden and by the EZ-KOMPAS Program of the “Samenwerkingsverband Noord-Nederland”. The authors would like to thank Evides and PWN for their assistance in the sample collection, Ramesh Duraisamy for his analytical assistance and the participants of the Wetsus research theme “Biofouling” for the fruitful discussions and their financial support. We also thank the two anonymous reviewers for their constructive comments.

Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.watres.2014.12.012.

4.

Summary and conclusions

1. The application of Alcian blue (AB) staining to measure acid polysaccharides, including transparent exopolymer particles (TEP), in brackish and saline waters is mainly limited by the low stability of the dye and its sensitivity to water salinity. Such problems can be minimised by prefiltering AB solution through 0.05 mm membrane and filtering ultra-pure water through TEP gels to rinsed-off inorganic ions before AB staining.

references

Alizadeh Tabatabai, S.A., Schippers, J.C., Kennedy, M.D., 2014. Effect of coagulation on fouling potential and removal of algal organic matter in ultrafiltration pretreatment to seawater reverse osmosis. Water Res. 59, 283e294. Alldredge, A.L., Passow, U., Logan, B.E., 1993. The abundance and significance of a class of large, transparent organic particles in

w a t e r r e s e a r c h 7 0 ( 2 0 1 5 ) 3 0 0 e3 1 2

the ocean. Deep Sea Res. Part I Oceanogr. Res. Pap. 40 (6), 1131e1140. Arruda-Fatibello, S.H.S., Henriques-Vieira, A.A., FatibelloFilho, O., 2004. A rapid spectrophotometric method for the determination of transparent exopolymer particles (TEP) in freshwater. Talanta 62 (1), 81e85. Azetsu-Scott, K., Passow, U., 2004. Ascending marine particles: significance of transparent exopolymer particles (TEP) in the upper ocean. Limnol. Oceanogr. 49 (3), 741e748. Bartold, P.M., Page, R.C., 1985. A microdetermination method for assaying glycosaminoglycans and proteoglycans. Anal. Biochem. 150 (2), 320e324. Bar-Zeev, E., Berman-Frank, I., Girshevitz, O., Berman, T., 2012. Revised paradigm of aquatic biofilm formation facilitated by microgel transparent exopolymer particles. PNAS 109 (23), 9119e9124. Bar-Zeev, E., Berman-Frank, I., Liberman, B., Rahav, E., Passow, U., Berman, T., 2009. Transparent exopolymer particles: potential agents for organic fouling and biofilm formation in desalination and water treatment plants. Desalin. Water Treat. 3, 136e142. Berman, T., Holenberg, M., 2005. Don't fall foul of biofilm through high TEP levels. Filtr. Sep. 42 (4), 30e32. Chin, W.-C., Orellana, M.V., Verdugo, P., 1998. Spontaneous assembly of marine dissolved organic matter into polymer gels. Nature 391, 568e572. de la Torre, T., Lesjean, B., Drews, A., Kraume, M., 2008. Monitoring of transparent exopolymer particles (TEP) in a membrane bioreactor (MBR) and correlation with other fouling indicators. Water Sci. Technol. 58 (10), 1903e1909. Discart, V., Bilad, M.R., Vankelecom, I.F.J., 2014. Critical evaluation of the determination methods for transparent exopolymer particles (TEP), agents of membrane fouling. Crit. Rev. Environ. Sci. Technol. http://dx.doi.org/10.1080/10643389.2013.829982. Filella, M., 2014. Understanding what we are measuring: standards and quantification of natural organic matter. Water Res. 50, 287e293. Guillard, R.R., Ryther, J.H., 1962. Studies on marine planktonic diatoms. I. Cyclotella nana Hustedt and Detonula confervacaea (Cleve) Gran. Can. J. Microbiol. 8, 229e239. Harvey, D., 2000. Modern Analytical Chemistry. McGraw-Hill, USA. Heinonen, K.B., Ward, J.E., Holohan, B.A., 2007. Production of transparent exopolymer particles (TEP) by benthic suspension feeders in coastal systems. J. Exp. Mar. Biol. Ecol. 341, 184e195. Horobin, R.W., 1988. Understanding Histochemistry: Selection, Evaluation and Design of Biological Stains. Wiley, New York. Huber, S.A., Balz, A., Abert, M., Pronk, W., 2011. Characterisation of aquatic humic and non-humic matter with size-exclusion chromatography e organic carbon detection e organic nitrogen detection (LC-OCD-OND). Water Res. 45 (2), 879e885. Hung, C.-C., Guo, L., Santschi, P.H., Alvarado-Quiroz, N., Haye, J.M., 2003. Distributions of carbohydrate species in the Gulf of Mexico. Mar. Chem. 81 (3e4), 119e135. € rnsson, S., 2001. Quantitation of proteoglycans in Karlsson, M., Bjo biological fluids using Alcian blue. Proteoglycan Protoc. Methods Mol. Biol. 171, 159e173. ~ oz-Tobar, F.P., Amy, G.L., Schippers, J.C., Kennedy, M.D., Mun 2009. Transparent exopolymer particle (TEP) fouling of ultrafiltration membrane systems. Desalin. Water Treat. 6 (1e3), 169e176. Kiørboe, T., Hansen, J.L., 1993. Phytoplankton aggregate formation: observations of patterns and mechanisms of cell sticking and the significance of exopolymeric material. J. Plankton Res. 15 (9), 993e1018. Mari, X., Kiørboe, T., 1996. Abundance, size distribution and bacterial colonization of transparent exopolymer particles (TEP) in the Kattegat. J. Plankton Res. 18, 969e986.

311

McKee, M.P., Ward, J.E., MacDonald, B.A., Holohan, B.A., 2005. Production of transparent exopolymer particles by the eastern oyster Crassostrea virginica. Mar. Ecol. Prog. Ser. 183, 59e71. Mopper, K., Zhou, J., Sri Ramana, K., Passow, U., Dam, H.G., Drapeau, D.T., 1995. The role of surface-active carbohydrates in the flocculation of a diatom bloom in a mesocosm. DeepSea Res. Part II 42 (1), 47e73. Myklestad, S., Holm-Hansen, O., Varum, K.M., Volcani, B.E., 1989. Rate of release of extracellular amino acids and carbohydrates from marine diatom Chaetoceros affinis. J. Plankton Res. 11 (4), 763e773. Parker, B.C., Diboll, A.G., 1966. Alcian stains for histochemical localization of acid and sulfated polysaccharides in algae. Phycologia 6 (1), 37e46. Passow, U., 2000. Formation of transparent exopolymer particles (TEP) from dissolved precursor material. Mar. Ecol. Prog. Ser. 192, 1e11. Passow, U., 2002a. Transparent exopolymer particles (TEP) in aquatic environments. Prog. Oceanogr. 55 (3), 287e333. Passow, U., 2002b. Production of transparent exopolymer particles (TEP) by phyto-and bacterioplankton. Mar. Ecol. Prog. Ser. 236, 1e12. Passow, U., Alldredge, A.L., 1995. A dye-binding assay for the spectrophotometric measurement of transparent exopolymer particles (TEP). Limnol. Oceanogr. 40 (7), 1326e1335. Passow, U., Wassmann, P., 1994. On the trophic fate of Phaeocystis pouchetii (hariot): IV. the formation of marine snow by P. pouchetii. Mar. Ecol. Prog. Ser. 104 (1e2), 153e161. Passow, U., Shipe, R.F., Murray, A., Pak, D.K., Brzezinski, M.A., Alldredge, A.L., 2001. Origin of transparent exopolymer particles (TEP) and their role in the sedimentation of particulate matter. Cont. Shelf Res. 21, 327e346. Ramus, J., 1977. Alcian blue: a quantitative aqueous assay for algal acid and sulfated polysaccharides. J. Phycol. 13, 345e348. Schurer, R., Tabatabai, A., Villacorte, L., Schippers, J.C., Kennedy, M.D., 2013. Three years operational experience with ultrafiltration as SWRO pre-treatment during algal bloom. Desalin. Water Treat. 51 (4e6), 1034e1042. Schurer, R., Janssen, A., Villacorte, L., Kennedy, M.D., 2012. Performance of ultrafiltration & coagulation in an UF-RO seawater desalination demonstration plant. Desalin. Water Treat. 42 (1e3), 57e64. Scott, J.E., Quintarelli, G., Dellovo, M.C., 1964. The chemical and histochemical properties of alcian blue e I. the mechanism of alcian blue staining. Histochemie 4 (2), 73e85. Thornton, D.C.O., Fejes, E.M., DiMarco, S.F., Clancy, K.M., 2007. Measurement of acid polysaccharides (APS) in marine and freshwater samples using alcian blue. Limnol. Oceanogr. Methods 5, 73e87. Verdugo, P., 2012. Marine microgels. Annu. Rev. Mar. Sci. 4, 375e400. Verdugo, P., Alldredge, A.L., Azam, F., Kirchman, D.L., Passow, U., Santschi, P.H., 2004. The oceanic gel phase: a bridge in the DOMePOM continuum. Mar. Chem. 92, 67e85. Villacorte, L.O., Kennedy, M.D., Amy, G.L., Schippers, J.C., 2009a. Measuring Transparent Exopolymer Particles (TEP) as indicator of the (bio)fouling potential of RO feed water. Desalin. Water Treat. 5, 207e212. Villacorte, L.O., Kennedy, M.D., Amy, G.L., Schippers, J.C., 2009b. The fate of Transparent Exopolymer Particles (TEP) in integrated membrane systems: removal through pretreatment processes and deposition on reverse osmosis membranes. Water Res. 43 (20), 5039e5052. Villacorte, L.O., Schurer, R., Kennedy, M., Amy, G., Schippers, J.C., 2010a. Removal and deposition of Transparent Exopolymer Particles (TEP) in seawater UF-RO system. IDA J. 2, 45e55. Villacorte, L.O., Schurer, R., Kennedy, M., Amy, G., Schippers, J.C., 2010b. The fate of transparent exopolymer particles in

312

w a t e r r e s e a r c h 7 0 ( 2 0 1 5 ) 3 0 0 e3 1 2

integrated membrane systems: a pilot plant study in Zeeland, The Netherlands. Desalin. Water Treat. 13, 109e119. Villacorte, L.O., Ekowati, Y., Winters, H., Amy, G.L., Schippers, J.C., Kennedy, M.D., 2013. Characterisation of transparent exopolymer particles (TEP) produced during algal bloom: a membrane treatment perspective. Desalin. Water Treat. 51 (4e6), 1021e1033.

Wardi, A.H., Allen, W.S., 1972. Alcian blue staining of glycoproteins. Anal. Biochem. 48 (2), 621e623. Zhou, J., Mopper, K., Passow, U., 1998. The role of surface-active carbohydrates in the formation of transparent exopolymer particles by bubble adsorption of seawater. Limnol. Oceanogr. 43 (8), 1860e1871.