Effect of Biofouling on the Performance of ... - ACS Publications

91 downloads 182655 Views 9MB Size Report
Jan 27, 2016 - pubs.acs.org/IECR. © 2016 American Chemical Society. 4328. DOI: 10.1021/acs.iecr.5b03457. Ind. Eng. ..... Current plans call for placing braids of adsorbent material in ..... Assessment Support Center 2002, 2002, 1−23.
Article pubs.acs.org/IECR

Effect of Biofouling on the Performance of Amidoxime-Based Polymeric Uranium Adsorbents Jiyeon Park,*,† Gary A. Gill,† Jonathan E. Strivens,† Li-Jung Kuo,† Robert T. Jeters,† Andrew Avila,† Jordana R. Wood,† Nicholas J. Schlafer,† Christopher J. Janke,‡ Erin A. Miller,§ Mathew Thomas,§ R. Shane Addleman,§ and George T. Bonheyo*,† †

Marine Sciences Laboratory, Pacific Northwest National Laboratory, 1529 West Sequim Bay Road, Sequim, Washington 98382, United States § Pacific Northwest National Laboratory, 902 Battelle Boulevard, Richland, Washington 99352, United States ‡ Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, Tennessee 37831, United States ABSTRACT: The Marine Science Laboratory at the Pacific Northwest National Laboratory evaluated the impact of biofouling on the performance or uranium adsorbents. A surface-modified polyethylene adsorbent fiber provided by Oak Ridge National Laboratory, AF adsorbent, was tested in either the presence or absence of light to simulate deployment in shallow or deep marine environments. Samples of the adsorbent fiber were exposed to seawater as loose fibers packed with glass beads in columns and as >10-cm-long braids of fiber placed in a flume that provided a continuous flow representative of natural ocean currents. Exposure tests (42 days) in column and flume settings showed that biofouling resulted in decreased uranium uptake by the adsorbent fiber. Uranium uptake was reduced by up to 30%, in the presence of simulated sunlight, which also increased biomass accumulation and altered the microbial community composition on the fibers. These results suggest that deployment below the photic zone would mitigate the effects of biofouling, resulting in greater yields of uranium extracted from seawater.



INTRODUCTION The Fuel Resources Program at the U.S. Department of Energy’s Office of Nuclear Energy is supporting the development of adsorbent technology to extract uranium from seawater. This technology is being developed to provide a sustainable and economically viable supply of uranium fuel for nuclear reactors.1 Uranium is present in seawater at a concentration of ∼3.3 ppb, amounting to a global marine resource of an estimated 4.5 billion metric tons.2 A major effort in the development of this technology at the Pacific Northwest National Laboratory (PNNL) is to test the performance of the uranium adsorption materials in natural seawater under realistic marine conditions. Briefly, the envisioned strategy of the program is to incorporate uranium adsorbent chemistry into a fibrous form (e.g., through grafting onto polyethylene fibers) and to create large (i.e., tens of meters long) braids of the material that may be anchored to the seafloor during the collection (adsorption) process.1 Preliminary testing and selection of amidoxime-based adsorbent fibers were performed at a number of participating laboratories using artificial seawater that lacked many of the chemical and biological properties of natural seawater. Additional evaluations were then performed at the PNNL Marine Science Laboratory (MSL) in Sequim, WA, using filtered seawater. Although artificial and filtered seawater was sufficient for an initial downselection of both support materials and ligand chemistry, the lack of cellular, biomolecular, and organic matter in these test waters was unrealistic. Design features of the adsorbent materials, such as large porous or textured surface areas, with micro- or nanoscaled features that are intended to maximize uranium uptake may also enhance the adsorption of dissolved and © 2016 American Chemical Society

particulate organic matter and colonization by marine microorganisms.3−6 Biofouling is the accumulation of microorganisms, algae, plants, or animals on wetted surfaces. In the marine environment, biofouling is generally a four-step process.7,8 During the first stage, surfaces are rapidly coated with an organic conditioning film beginning within 5−10 s after immersion.9 Then single bacterial cells and diatoms begin to settle, adhere, and colonize on the surface.10 During the third stage, microbial films develop, creating rough surfaces that trap more particles and organisms including the larval forms of macroorganisms such as barnacles. During the final stage, outgrowth of macroorganisms like mussels, barnacles, or algae occurs on the fouled surface.11,12 In marine industries, biofouling causes various problems including increased shipping costs due to ship hull fouling (greater fuel consumption), increased maintenance cost and time, the release of toxins in antifouling paints during cleaning, and the unintended transport of invasive foreign species.13−17 Concerns about the impact of biofouling on the performance of uranium adsorbents may be summarized as shown in Figure 1. In the absence of biofouling, the amount of uranium uptake should increase until the adsorbent is saturated, as shown by the blue curve. However, prolonged deployment of uranium adsorbent in the ocean will increase exposure to live and dead Special Issue: Uranium in Seawater Received: Revised: Accepted: Published: 4328

September 25, 2015 January 12, 2016 January 26, 2016 January 27, 2016 DOI: 10.1021/acs.iecr.5b03457 Ind. Eng. Chem. Res. 2016, 55, 4328−4338

Article

Industrial & Engineering Chemistry Research

and with or without lighting to simulate shallow or deep seawater settings. Some experiments were performed using an adsorbent material packed into columns to strictly control and measure the amount of water passing over the fibers and to ensure that nearly 100% of the fiber could be recovered at the conclusion of a test. This column setup has been previously described.2 A limitation of the columns was that these cannot be used with unfiltered seawater because of rapid clogging of the filter bed that holds the adsorbent. Therefore, additional experiments were conducted using braids of fiber in an openflume setting with flowing unfiltered seawater, with or without light. A limitation of these flume studies was the potential loss of loose adsorbent material and the ability to accurately measure the mass of adsorbent (excluding biofouling and water) in recovered samples. Measurements quantified the amount of biofouling material and the total amount of uranium adsorbed onto the fiber using a destructive method of analysis. A DNA-based approach was used to assess the composition and diversity of the microbial communities found on the adsorbent. An attempt was also made to visualize fouling on the surface of adsorbent fibers to better understand where fouling occurs relative to the surface structure of the fiber. These data could be used to gain insight into structural designs that might maximize the water-accessible surface area of the adsorbent while limiting the extent and impact of biofouling.

Figure 1. Environmental impact on uranium extraction from seawater. Conceptual figure showing the relationship of time with biomass accumulation (biofouling) and its effect on uranium uptake and recovery.

biogenic materials and allow for the growth of any cells that colonize the surface of the adsorbent material (green curve). An increase in biomass will have two primary effects that decrease uranium uptake and recovery (red curve): (1) limiting the accessibility of the ligands to ocean water during uptake and (2) interfering with the uranium recovery process by diluting the extraction solution and/or restricting its access to the ligands. A third potential detrimental effect that is not covered in this study is that fouling may lead to decreased reusability of the adsorbent because of biocorrosion, detrimental treatments necessary to remove fouling, damage from added weight and drag, or promotion of damage by fish and invertebrates grazing on organic matter found on the surface of the adsorbent. Therefore, in order to maximize uranium extraction in a real world environment, it is important to understand how biofouling impacts uranium uptake and recovery, the potential rates of buildup, the correlation of biomass to interference, and the kinetics of the adsorbents and their durability. This information will be highly valuable and extensively useful in the development of deployment and operation plans and identification of cost-effective and compatible methods to mitigate the effects of biofouling. To quantify the potential impact of fouling on uranium uptake, a set of time series experiments were conducted that examined biomass accumulation, cell growth, and uranium uptake on a representative high performance adsorbent material. One set of experiments was designed to assess the maximum potential impact fouling might have on the adsorbents. A second set of experiments considered how different deployment strategies, shallow water with light exposure or deep water with no light exposure, might affect the performance. Sunlight plays an important role in biofouling because in many of the early stage, surface-colonizing microorganisms are photosynthetic (e.g., diatoms, cyanobacteria). In this study, preweighed samples of the adsorbent material were exposed to either filtered or unfiltered seawater, with or without pretreatment with microorganisms (to maximize fouling and establish a theoretical upper bound on the effects)



EXPERIMENTAL METHODS Adsorbents. Oak Ridge National Laboratory (ORNL) developed and produced AF1 and AI8 adsorbents. Adsorbents were prepared using hollow-gear-shaped, high-surface-area polyethylene fibers and a radiation-induced graft polymerization method.2 AF1 and AI8 adsorbents were different in the grafting comonomer (itaconic acid and vinylphosphonic acid, respectively). Both used amidoxime as the uranium binding ligand. Column System. Marine testing was conducted using ambient seawater from Sequim Bay. The PNNL MSL seawater delivery system provided ambient seawater to a wet laboratory for scientific investigation. Ambient seawater was drawn continuously (day and night) by pump from a depth of ∼10 m from Sequim Bay; consequently, water was collected through all diurnal and tidal cycles. Input water was delivered through a plastic pipe and passed through an Arkal Spin Klin filter system (nominal pore size 40 μm) to remove large particles. The seawater was then fed into a large-volume outdoor reservoir tank and supplied to the laboratory research facilities at MSL by a gravity feed through poly(vinyl chloride) (PVC) piping. Seawater from the large reservoir tank was fed sequentially through 5 μm and then 1 μm cellulose filters and collected in a 180 L fiberglass reservoir tank (head tank). Seawater in the head tank could be heated to a desired temperature. Temperature-controlled seawater was drawn from the head tank with a pump (nonmetallic pump head) and passed through 0.35−0.45 μm poly(ether sufone) (Memtrex MP, GE Power and Water, ora cellulobromoperoxidase membrane cartridge filter and into a 24-port PVC manifold. Water that was not used to expose the adsorbent material passed through the manifold was returned to the head tank. The pressure in the manifold was controlled with a gate valve at the outlet of the manifold. MSL has four separate 24-port manifolds, linked to three separate head tanks, that permitted testing of 96

4329

DOI: 10.1021/acs.iecr.5b03457 Ind. Eng. Chem. Res. 2016, 55, 4328−4338

Article

Industrial & Engineering Chemistry Research adsorbent materials in flow-through columns simultaneously (Figure 2).

species that contribute to biofouling but removed larger marine plankton species. Temperature control for experimental purposes was achieved by feeding the 150 μm filtered seawater into a 180 L head tank that had a 10000 W titanium immersion heater. The ambient temperature of seawater as it entered the head tank was ∼10 °C, and the temperature in the head tank was maintained at 20 °C, with a variability of approximately ±1.5 °C. Temperature-controlled water was drawn out of the head tank using a pump and delivered to a multiport manifold on the wet-laboratory table for experimental distribution. PNNL developed recirculating flumes for conducting exposure tests with braided adsorbent materials under controlled temperature and flow-rate conditions. For exposure with 150 μm filtered seawater, the flumes were constructed of an acrylic material to allow ambient light to pass to the seawater in the flume. Allowing light penetration permitted the growth of photosynthetic organisms (biofouling), as would occur in the natural photic seawater environment. Two translucent flumes were built for the biofouling study. Flume Experiment. The braided ORNL AI8 adsorbent was tested in clear flumes in either the presence or absence of light. Diurnal daylight spectral light (5000 K) was provided for the light flume, and the dark flume was covered with a black tarp to prevent exposure to light (Figure 3). The intensity of light at

Figure 2. Column experimental setup in either the absence or presence of light: (A) Columns were covered with aluminum foil to prevent exposure to light. (B) Daylight spectral light was provided on a 12 h−12 h light−dark cycle. A fan was placed to prevent heating of the columns.

Prefouling with Navicula incerta. A stock culture of the marine benthic diatom N. incerta (UTEX 2044) was grown in a F/2 medium (NCMA, ME) using the conditions described below. The ORNL AF1 adsorbent was weighed (50 mg) and then conditioned in 2.5% KOH at 80 °C for 1 h. Conditioned fibers were pH-neutralized, then placed in a 50 mL plastic vial, and submerged in deionized (DI) water to remove most of the KOH solution. Fibers were transferred to sterile 250 mL glass flasks each containing 150 mL (6.7 × 104 cells/mL) of the N. incerta culture. Fibers and culture mixtures were then incubated at room temperature with shaking (135 rpm) and daylight spectral light (5000 K) on a 12 h−12 h light−dark cycle until fouling became visible. Prefouling was designed to ensure biological growth and test the maximum impact of biofouling on uranium uptake, particularly in the column experiments where minimal cellular material was expected to be present in the filtered seawater. Column Experiment in Either the Presence or Absence of Light. The preweighed ORNL AF1 adsorbent was conditioned in 2.5% KOH at 80 °C for 1 h. Conditioned adsorbents with or without prefouling were placed between packing materials (acid-cleaned 5 mm glass beads and glass fiber) in columns. Columns were mounted on manifolds, and 0.45 μm filtered seawater was passed through these columns for 42 days at 20 °C. To prevent exposure to light, the columns were wrapped in aluminum foil (Figure 2A). For column testing with light, daylight spectral light (5000 K) was provided on a 12 h−12 h light−dark cycle, and a fan was used to prevent overheating of the columns (Figure 2B). Flume System. Ambient, unfiltered seawater was drawn by pump from a depth of ∼10 m from Sequim Bay through a plastic pipe and delivered through a PVC piping system to the wet laboratory for use. Water was drawn continuously through all diurnal and tidal cycles. Furthermore, the continuous operation of the pumping system over several months prior to the experiments afforded species from multiple seasons and all diurnal and tidal cycles with an opportunity to colonize the plumbing and tanks, thus providing a potential secondary source of organisms. Gross filtration of the ambient seawater was conducted to remove large debris using a Big Bubba nonmetallic filter housing and a 150 μm filter. The 150 μm filtration allowed for the free passage of most phytoplankton

Figure 3. Flume setup. In order to test the impact of light on biofouling and subsequent uranium uptake, two clear flumes, a dark flume and a light flume, were set up. 150 μm filtered seawater (to remove large particles) was drawn to the flumes by pumps at 2−5 L/ min.

the surface of the water in the flume was measured to be approximately 50 μmol of photons/m2·s, a small fraction of the light intensity of a bright summer day at noon (∼2000 μmol of photons/m2·s) yet still strong enough to enhance photosynthetic growth inside the flume, as evidenced by recurring visible algae growth on the flume walls. The adsorbent braids were tested in two formats: “regular” braids (9 g) and premeasured “minibraids” (100 mg) that were prepared as follows. A braid of the adsorbent material was made by ORNL. This braid consisted of long (tens of meters) fibers that were braided together to form a tight central stem surrounded by loose loops of fiber in which each loop was 6− 10 cm in circumference. Shorter sections were cut at the stem from the primary preparation to create ∼10−15-cm-long (9 g) sections of braided material, herein referred to as regular braids. Minibraids were made by cutting loops off of the stem to create 4330

DOI: 10.1021/acs.iecr.5b03457 Ind. Eng. Chem. Res. 2016, 55, 4328−4338

Article

Industrial & Engineering Chemistry Research

that fiber sample inside was directly facing the end of the horn. Sonication was performed for 1 min (100% amplitude, 0.75 cycle), then the tube was rotated 180°, and sonication was repeated to ensure thorough treatment of the fiber surface. NPOC analysis was performed using a Shimadzu TOC-L (Shimadzu Scientific Instruments, Columbia, MD). A 30 mL sample of the hydrogen peroxide solution used to clean the fiber samples was placed in a 40 mL certified clean NPOC test vial (Fisher Scientific, Hampton, NH) for NPOC testing. Potassium hydrogen phthalate (50 and 5 ppm) and unused 3% hydrogen peroxide were used as test controls. All NPOC vials were loaded into the Shimadzu ASI-L auto sampler. NPOC settings were as follows: sparge time 2 min; acid addition 1.5% HCl trace grade (Fisher Scientific, Pittsburgh, PA); 3 of 5 injections per sample; CV max 2%. To subtract any carbon input from the solvents, samples of the 3% hydrogen peroxide solution and Milli-Q water were taken and tested in NPOC analysis as blank controls. DNA Extraction and Community Analysis. A 50 mL centrifuge tube containing the fiber sample was filled with 25 mL of nuclease free water (BioExpress Corp., Kaysville, UT) and vortexed to release biomass accumulation from the fiber into the water. The cell suspension, excluding adsorbent fibers, was carefully transferred by pipet to a sterile 50 mL centrifuge tube and then further split into two equivalent fractions. The fractions were centrifuged at 4150 rpm for 30 min at 4 °C. The supernatant of each fraction was decanted, resulting in two pellets. One pellet was resuspended in 480 μL of nuclease free water and placed in a sterile 2 mL microcentrifuge tube to be used for direct polymerase chain reaction (PCR) amplification. The other pellet was resuspended in 480 μL of 50 mM ethylenediaminetetraacetic acid and then transferred to a sterile 2 mL microcentrifuge tube to be further processed for total DNA extraction using the Wizard Genomic DNA Purification Kit (Promega Corp., Fitchburg, WI). DNA amplification of the bacterial 16S-23S internal transcribed spacer (ITS) and ITS15.8S0ITS2 regions of eukaryotes was performed. PCR was performed using OneTaq DNA polymerase (New England Biolabs, Ipswich, MA). The template for the PCR reaction consisted of either purified DNA from the total DNA extraction process or a raw suspension for direct amplification. The PCR products were run on a Labchip GXII bioanalyzer (PerkinElmer, Waltham, MA) using the High Sensitivity Lab-Chip and Reagent Kit in order to determine the amplicon sizes and relative amounts of amplified DNA. Advanced Photon Source (APS) Imaging. An attempt was made to examine where fouling organisms might grow and adhere on the adsorbent fibers and to investigate whether the growth of organisms on the adsorbent resulted in any physical damage to the fibers. X-ray microtomography was selected as a means of analysis because it provided a means for both crosssectional (2D) and three-dimensional (3D) visualization of the fibers and biofilms at appropriate scales and has been previously demonstrated in the analysis of biofilm structures.18,19 For these experiments, the model biofilm and fouling bacterium Pseudomonas f luorescens was grown in LB medium with kanamycin (50 μg/mL) overnight at 37 °C in an incubator shaker. A 9 mL sample of overnight culture was added to a sterile 50 mL centrifuge tube containing the ORNL AF1 adsorbent and incubated for 3 days. The fiber samples were then collected and fixed using 4% paraformaldehyde (16% stock solution diluted with 1 X PBS) for 5 min at room temperature followed by washing twice with 1X PBS for two

clusters of now separate, linearized fiber pieces. The minibraids were weighed and subsequently tied midlength using a fishing line prior to KOH conditioning (Figure 4A). The prepared

Figure 4. ORNL AI8 adsorbent: (A) Minibraids were preweighed (100 mg) and tied using fishing line prior to KOH conditioning. (B) Conditioned minibraids were secured inside the flume using cable ties. The red box indicates a single minibraid. (C) Braided AI8 before conditioning is shown. (D) The conditioned AI8 braid was secured inside the flume.

braids (which had also been KOH-conditioned) were secured inside the flume using cable ties and exposed to the coarsely filtered (150 μm) seawater at 20 °C (Figure 4B,D). The regular braids provided a format intended for actual deployment in the environment, but fouling on the surface appeared to be uneven, the size of the sample was deemed large enough to pose a radiation safety challenge, and smaller samples cut from the regular braid after exposure could not be weighed accurately because these now contained seawater, biofouling, and accumulated metals. The preweighed, smaller minibraids provided greater accuracy for measurements depending upon the starting mass of the material but may have introduced some differences because of the altered physical format. Samples were collected at six different time points: 0, 7, 14, 21, 28, and 42 days. Uranium and Trace Element Analysis. Following exposure to seawater, adsorbent samples were rinsed in DI water to remove accumulated salts. Each sample was then placed in a plastic vial and dried using a heating block (80 °C overnight). The dried and weighed adsorbent sample was placed in a plastic or Teflon container, and then a solution of 50% aqua regia was added for total digestion of the adsorbent. The digestion mix was incubated in a heating block at 80 °C for 2 h and then analyzed for uranium and trace elements via inductively coupled plasma optical emission spectrometry. Adsorption (uptake) was determined based on the mass of the recovered elements per mass of adsorbent (g of element adsorbed per kg of dry adsorbent). Microscopy. Adsorbent fiber samples were examined under a Leica DMIRB inverted fluorescence microscope (Bartels & Stout, Inc., Issaquah, WA) at 1000× magnification. Nonpurgeable Organic Carbon (NPOC) Analysis. The adsorbent samples (50 mg) were placed in a 50 mL centrifuge tube, to which 30 mL of a 3% hydrogen peroxide solution was added and then vortexed for 10 s to strip the accumulated biomass off of the adsorbent fiber. The 3% hydrogen peroxide solution was made by diluting 30% hydrogen peroxide (JT Baker, Easton, PA) with 18 MΩ Milli-Q water (1:10 ratio). The tube was then placed in a 55 °C water bath. After incubation for 30 min, the tube was removed, vortexed for an additional 10 s, and then placed back into the water bath for another 30 min incubation. Following these incubations, the 50 mL conical tube containing the fiber sample was fastened to the ultrasound horn of a VialTweeter ultrasound unit (Hielscher USA, Ringwood, NJ) by two large rubber bands and aligned such 4331

DOI: 10.1021/acs.iecr.5b03457 Ind. Eng. Chem. Res. 2016, 55, 4328−4338

Article

Industrial & Engineering Chemistry Research times. A control fiber sample that was only exposed to sterile culture medium was also prepared for imaging. The wet fibers were then shipped to the APS at Argonne National Laboratory in sterile seawater media. X-ray microtomography was performed at the APS, beamline 2-BM.20 The dry fibers were transferred into a plastic pipet tip, which provided stability with low attenuation. The fibers were imaged at a resolution of 0.7 μm per pixel; a series of images are acquired at different sample angles over the course of 5−10 min to form a 3-D data set, which is then reconstructed into a volume of 2048 × 2048 × 2048 pixels. The X-ray beam energy was 22.5 keV, with the detector 200 mm back from the sample. This distance allowed for edge enhancement due to refractive index effects, which increases contrast on low-atomic number materials. After reconstruction into a volume, the data set was then rendered into a 2D image using Matlab. The color scale in the image indicates the relative absence of intensity, due to either attenuation or refractive effects, and tends to be larger for higher atomic number materials or near large changes in the electron density.

adsorbents. The experiments considered conditions that may promote or limit the rate of fouling, developed a correlation between the amount of fouling and the degree of impact, and attempted to place an upper bound on the potential impact fouling may have on the rate of uranium uptake and total capacity of a model adsorbent material. A leading adsorbent material provided by ORNL, AF1 adsorbent, which is based upon a polyethylene trunk material and amidoxime ligands was used for the tests. Current plans call for placing braids of adsorbent material in water